diff --git "a/checkpoint-19025/trainer_state.json" "b/checkpoint-19025/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-19025/trainer_state.json" @@ -0,0 +1,133208 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.19025, + "eval_steps": 500, + "global_step": 19025, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 1e-05, + "grad_norm": 1.4574347149106512, + "learning_rate": 3e-06, + "loss": 10.8576, + "step": 1 + }, + { + "epoch": 2e-05, + "grad_norm": 1.4482443114713344, + "learning_rate": 6e-06, + "loss": 10.8575, + "step": 2 + }, + { + "epoch": 3e-05, + "grad_norm": 1.4642003089092823, + "learning_rate": 9e-06, + "loss": 10.8576, + "step": 3 + }, + { + "epoch": 4e-05, + "grad_norm": 1.4485584072048276, + "learning_rate": 1.2e-05, + "loss": 10.8569, + "step": 4 + }, + { + "epoch": 5e-05, + "grad_norm": 1.4574321657516995, + "learning_rate": 1.5e-05, + "loss": 10.8527, + "step": 5 + }, + { + "epoch": 6e-05, + "grad_norm": 1.4585537094370684, + "learning_rate": 1.8e-05, + "loss": 10.8518, + "step": 6 + }, + { + "epoch": 7e-05, + "grad_norm": 1.4183780170798466, + "learning_rate": 2.1000000000000002e-05, + "loss": 10.8383, + "step": 7 + }, + { + "epoch": 8e-05, + "grad_norm": 1.286272643239374, + "learning_rate": 2.4e-05, + "loss": 10.8119, + "step": 8 + }, + { + "epoch": 9e-05, + "grad_norm": 1.246364249616181, + "learning_rate": 2.7e-05, + "loss": 10.8063, + "step": 9 + }, + { + "epoch": 0.0001, + "grad_norm": 1.2300728857390288, + "learning_rate": 3e-05, + "loss": 10.7913, + "step": 10 + }, + { + "epoch": 0.00011, + "grad_norm": 1.1639358472437353, + "learning_rate": 3.2999999999999996e-05, + "loss": 10.7756, + "step": 11 + }, + { + "epoch": 0.00012, + "grad_norm": 1.1455069336623074, + "learning_rate": 3.6e-05, + "loss": 10.7622, + "step": 12 + }, + { + "epoch": 0.00013, + "grad_norm": 1.104711768149659, + "learning_rate": 3.9e-05, + "loss": 10.7415, + "step": 13 + }, + { + "epoch": 0.00014, + "grad_norm": 1.0916409179063882, + "learning_rate": 4.2000000000000004e-05, + "loss": 10.7296, + "step": 14 + }, + { + "epoch": 0.00015, + "grad_norm": 1.0723152562842644, + "learning_rate": 4.4999999999999996e-05, + "loss": 10.7183, + "step": 15 + }, + { + "epoch": 0.00016, + "grad_norm": 1.0297043497697835, + "learning_rate": 4.8e-05, + "loss": 10.7, + "step": 16 + }, + { + "epoch": 0.00017, + "grad_norm": 0.9882210863590011, + "learning_rate": 5.1000000000000006e-05, + "loss": 10.6838, + "step": 17 + }, + { + "epoch": 0.00018, + "grad_norm": 0.966791538351231, + "learning_rate": 5.4e-05, + "loss": 10.6659, + "step": 18 + }, + { + "epoch": 0.00019, + "grad_norm": 0.9358769725124266, + "learning_rate": 5.7e-05, + "loss": 10.6509, + "step": 19 + }, + { + "epoch": 0.0002, + "grad_norm": 0.9313245306892226, + "learning_rate": 6e-05, + "loss": 10.638, + "step": 20 + }, + { + "epoch": 0.00021, + "grad_norm": 0.9056933112957348, + "learning_rate": 6.3e-05, + "loss": 10.6235, + "step": 21 + }, + { + "epoch": 0.00022, + "grad_norm": 0.9013277325687112, + "learning_rate": 6.599999999999999e-05, + "loss": 10.6069, + "step": 22 + }, + { + "epoch": 0.00023, + "grad_norm": 0.8982860560109955, + "learning_rate": 6.9e-05, + "loss": 10.5935, + "step": 23 + }, + { + "epoch": 0.00024, + "grad_norm": 0.8941575476865218, + "learning_rate": 7.2e-05, + "loss": 10.5802, + "step": 24 + }, + { + "epoch": 0.00025, + "grad_norm": 0.8955811690787305, + "learning_rate": 7.500000000000001e-05, + "loss": 10.5652, + "step": 25 + }, + { + "epoch": 0.00026, + "grad_norm": 0.9005417636666915, + "learning_rate": 7.8e-05, + "loss": 10.5507, + "step": 26 + }, + { + "epoch": 0.00027, + "grad_norm": 0.8929394290279239, + "learning_rate": 8.1e-05, + "loss": 10.5379, + "step": 27 + }, + { + "epoch": 0.00028, + "grad_norm": 0.892893092343029, + "learning_rate": 8.400000000000001e-05, + "loss": 10.5234, + "step": 28 + }, + { + "epoch": 0.00029, + "grad_norm": 0.8951984198637418, + "learning_rate": 8.7e-05, + "loss": 10.5078, + "step": 29 + }, + { + "epoch": 0.0003, + "grad_norm": 0.901037080772758, + "learning_rate": 8.999999999999999e-05, + "loss": 10.4913, + "step": 30 + }, + { + "epoch": 0.00031, + "grad_norm": 0.899757879512845, + "learning_rate": 9.3e-05, + "loss": 10.4759, + "step": 31 + }, + { + "epoch": 0.00032, + "grad_norm": 0.8963415000423225, + "learning_rate": 9.6e-05, + "loss": 10.4612, + "step": 32 + }, + { + "epoch": 0.00033, + "grad_norm": 0.8971468155029705, + "learning_rate": 9.900000000000001e-05, + "loss": 10.4429, + "step": 33 + }, + { + "epoch": 0.00034, + "grad_norm": 0.8964368558533553, + "learning_rate": 0.00010200000000000001, + "loss": 10.426, + "step": 34 + }, + { + "epoch": 0.00035, + "grad_norm": 0.9007340868108898, + "learning_rate": 0.00010500000000000002, + "loss": 10.4083, + "step": 35 + }, + { + "epoch": 0.00036, + "grad_norm": 0.8977475222905443, + "learning_rate": 0.000108, + "loss": 10.3895, + "step": 36 + }, + { + "epoch": 0.00037, + "grad_norm": 0.8929896912849768, + "learning_rate": 0.000111, + "loss": 10.3721, + "step": 37 + }, + { + "epoch": 0.00038, + "grad_norm": 0.8925204062685723, + "learning_rate": 0.000114, + "loss": 10.3515, + "step": 38 + }, + { + "epoch": 0.00039, + "grad_norm": 0.8947925451707294, + "learning_rate": 0.000117, + "loss": 10.3314, + "step": 39 + }, + { + "epoch": 0.0004, + "grad_norm": 0.8990961452390619, + "learning_rate": 0.00012, + "loss": 10.3088, + "step": 40 + }, + { + "epoch": 0.00041, + "grad_norm": 0.8951984584897338, + "learning_rate": 0.000123, + "loss": 10.2891, + "step": 41 + }, + { + "epoch": 0.00042, + "grad_norm": 0.8942493459254965, + "learning_rate": 0.000126, + "loss": 10.2679, + "step": 42 + }, + { + "epoch": 0.00043, + "grad_norm": 0.8983684110980745, + "learning_rate": 0.000129, + "loss": 10.243, + "step": 43 + }, + { + "epoch": 0.00044, + "grad_norm": 0.8985869855625094, + "learning_rate": 0.00013199999999999998, + "loss": 10.2206, + "step": 44 + }, + { + "epoch": 0.00045, + "grad_norm": 0.899962207774676, + "learning_rate": 0.000135, + "loss": 10.1962, + "step": 45 + }, + { + "epoch": 0.00046, + "grad_norm": 0.8914153211826606, + "learning_rate": 0.000138, + "loss": 10.1735, + "step": 46 + }, + { + "epoch": 0.00047, + "grad_norm": 0.8980761529388176, + "learning_rate": 0.000141, + "loss": 10.146, + "step": 47 + }, + { + "epoch": 0.00048, + "grad_norm": 0.9007540011988344, + "learning_rate": 0.000144, + "loss": 10.1222, + "step": 48 + }, + { + "epoch": 0.00049, + "grad_norm": 0.896344611975273, + "learning_rate": 0.000147, + "loss": 10.0974, + "step": 49 + }, + { + "epoch": 0.0005, + "grad_norm": 0.9015948414141589, + "learning_rate": 0.00015000000000000001, + "loss": 10.0706, + "step": 50 + }, + { + "epoch": 0.00051, + "grad_norm": 0.9016102183936469, + "learning_rate": 0.000153, + "loss": 10.0433, + "step": 51 + }, + { + "epoch": 0.00052, + "grad_norm": 0.8993585803645682, + "learning_rate": 0.000156, + "loss": 10.0158, + "step": 52 + }, + { + "epoch": 0.00053, + "grad_norm": 0.8901736410301992, + "learning_rate": 0.000159, + "loss": 9.9917, + "step": 53 + }, + { + "epoch": 0.00054, + "grad_norm": 0.9083328832659304, + "learning_rate": 0.000162, + "loss": 9.9578, + "step": 54 + }, + { + "epoch": 0.00055, + "grad_norm": 0.8975570632788401, + "learning_rate": 0.000165, + "loss": 9.9337, + "step": 55 + }, + { + "epoch": 0.00056, + "grad_norm": 0.8924582131156604, + "learning_rate": 0.00016800000000000002, + "loss": 9.904, + "step": 56 + }, + { + "epoch": 0.00057, + "grad_norm": 0.8937117244788118, + "learning_rate": 0.000171, + "loss": 9.8765, + "step": 57 + }, + { + "epoch": 0.00058, + "grad_norm": 0.8912667557674427, + "learning_rate": 0.000174, + "loss": 9.8514, + "step": 58 + }, + { + "epoch": 0.00059, + "grad_norm": 0.899445317105929, + "learning_rate": 0.000177, + "loss": 9.819, + "step": 59 + }, + { + "epoch": 0.0006, + "grad_norm": 0.8827247644946434, + "learning_rate": 0.00017999999999999998, + "loss": 9.7927, + "step": 60 + }, + { + "epoch": 0.00061, + "grad_norm": 0.8856006114384688, + "learning_rate": 0.000183, + "loss": 9.7645, + "step": 61 + }, + { + "epoch": 0.00062, + "grad_norm": 0.8887861476206323, + "learning_rate": 0.000186, + "loss": 9.7329, + "step": 62 + }, + { + "epoch": 0.00063, + "grad_norm": 0.8870948257177573, + "learning_rate": 0.000189, + "loss": 9.7044, + "step": 63 + }, + { + "epoch": 0.00064, + "grad_norm": 0.8854465518366905, + "learning_rate": 0.000192, + "loss": 9.6741, + "step": 64 + }, + { + "epoch": 0.00065, + "grad_norm": 0.8913964558604793, + "learning_rate": 0.00019500000000000002, + "loss": 9.6417, + "step": 65 + }, + { + "epoch": 0.00066, + "grad_norm": 0.8860459843940278, + "learning_rate": 0.00019800000000000002, + "loss": 9.6181, + "step": 66 + }, + { + "epoch": 0.00067, + "grad_norm": 0.8855602986146246, + "learning_rate": 0.000201, + "loss": 9.5886, + "step": 67 + }, + { + "epoch": 0.00068, + "grad_norm": 0.89182613307124, + "learning_rate": 0.00020400000000000003, + "loss": 9.5554, + "step": 68 + }, + { + "epoch": 0.00069, + "grad_norm": 0.8788147221752699, + "learning_rate": 0.00020700000000000002, + "loss": 9.5276, + "step": 69 + }, + { + "epoch": 0.0007, + "grad_norm": 0.89013480384513, + "learning_rate": 0.00021000000000000004, + "loss": 9.4911, + "step": 70 + }, + { + "epoch": 0.00071, + "grad_norm": 0.8832608390392153, + "learning_rate": 0.00021299999999999997, + "loss": 9.4637, + "step": 71 + }, + { + "epoch": 0.00072, + "grad_norm": 0.8850551460752941, + "learning_rate": 0.000216, + "loss": 9.4371, + "step": 72 + }, + { + "epoch": 0.00073, + "grad_norm": 0.8854430272945265, + "learning_rate": 0.00021899999999999998, + "loss": 9.4029, + "step": 73 + }, + { + "epoch": 0.00074, + "grad_norm": 0.881388508551569, + "learning_rate": 0.000222, + "loss": 9.3766, + "step": 74 + }, + { + "epoch": 0.00075, + "grad_norm": 0.8804734413831777, + "learning_rate": 0.000225, + "loss": 9.3413, + "step": 75 + }, + { + "epoch": 0.00076, + "grad_norm": 0.891470306594604, + "learning_rate": 0.000228, + "loss": 9.3136, + "step": 76 + }, + { + "epoch": 0.00077, + "grad_norm": 0.8835806398314079, + "learning_rate": 0.000231, + "loss": 9.2843, + "step": 77 + }, + { + "epoch": 0.00078, + "grad_norm": 0.8778470434340718, + "learning_rate": 0.000234, + "loss": 9.2534, + "step": 78 + }, + { + "epoch": 0.00079, + "grad_norm": 0.8849142747864956, + "learning_rate": 0.00023700000000000001, + "loss": 9.2174, + "step": 79 + }, + { + "epoch": 0.0008, + "grad_norm": 0.8831859916332898, + "learning_rate": 0.00024, + "loss": 9.1835, + "step": 80 + }, + { + "epoch": 0.00081, + "grad_norm": 0.8883671155155716, + "learning_rate": 0.00024300000000000002, + "loss": 9.1516, + "step": 81 + }, + { + "epoch": 0.00082, + "grad_norm": 0.8919005998406658, + "learning_rate": 0.000246, + "loss": 9.1278, + "step": 82 + }, + { + "epoch": 0.00083, + "grad_norm": 0.8904109612127737, + "learning_rate": 0.00024900000000000004, + "loss": 9.0926, + "step": 83 + }, + { + "epoch": 0.00084, + "grad_norm": 0.8922919840670517, + "learning_rate": 0.000252, + "loss": 9.0627, + "step": 84 + }, + { + "epoch": 0.00085, + "grad_norm": 0.8851770394662277, + "learning_rate": 0.000255, + "loss": 9.0389, + "step": 85 + }, + { + "epoch": 0.00086, + "grad_norm": 0.892287423580011, + "learning_rate": 0.000258, + "loss": 9.0029, + "step": 86 + }, + { + "epoch": 0.00087, + "grad_norm": 0.8819038715566697, + "learning_rate": 0.000261, + "loss": 8.9819, + "step": 87 + }, + { + "epoch": 0.00088, + "grad_norm": 0.8825670273484024, + "learning_rate": 0.00026399999999999997, + "loss": 8.9452, + "step": 88 + }, + { + "epoch": 0.00089, + "grad_norm": 0.87734677010751, + "learning_rate": 0.000267, + "loss": 8.92, + "step": 89 + }, + { + "epoch": 0.0009, + "grad_norm": 0.8800982265590005, + "learning_rate": 0.00027, + "loss": 8.8935, + "step": 90 + }, + { + "epoch": 0.00091, + "grad_norm": 0.8751738333241339, + "learning_rate": 0.000273, + "loss": 8.8635, + "step": 91 + }, + { + "epoch": 0.00092, + "grad_norm": 0.8805112062017327, + "learning_rate": 0.000276, + "loss": 8.8313, + "step": 92 + }, + { + "epoch": 0.00093, + "grad_norm": 0.8782110873043902, + "learning_rate": 0.000279, + "loss": 8.8018, + "step": 93 + }, + { + "epoch": 0.00094, + "grad_norm": 0.8793675998346636, + "learning_rate": 0.000282, + "loss": 8.7769, + "step": 94 + }, + { + "epoch": 0.00095, + "grad_norm": 0.8739437022686155, + "learning_rate": 0.000285, + "loss": 8.7468, + "step": 95 + }, + { + "epoch": 0.00096, + "grad_norm": 0.8792093745975903, + "learning_rate": 0.000288, + "loss": 8.7137, + "step": 96 + }, + { + "epoch": 0.00097, + "grad_norm": 0.8733827683228169, + "learning_rate": 0.000291, + "loss": 8.6904, + "step": 97 + }, + { + "epoch": 0.00098, + "grad_norm": 0.8720585520116785, + "learning_rate": 0.000294, + "loss": 8.6531, + "step": 98 + }, + { + "epoch": 0.00099, + "grad_norm": 0.8706078195741852, + "learning_rate": 0.000297, + "loss": 8.6333, + "step": 99 + }, + { + "epoch": 0.001, + "grad_norm": 0.868963072522692, + "learning_rate": 0.00030000000000000003, + "loss": 8.6078, + "step": 100 + }, + { + "epoch": 0.00101, + "grad_norm": 0.8733428125714966, + "learning_rate": 0.00030300000000000005, + "loss": 8.5742, + "step": 101 + }, + { + "epoch": 0.00102, + "grad_norm": 0.8594222895452354, + "learning_rate": 0.000306, + "loss": 8.5555, + "step": 102 + }, + { + "epoch": 0.00103, + "grad_norm": 0.8663526773106932, + "learning_rate": 0.000309, + "loss": 8.5314, + "step": 103 + }, + { + "epoch": 0.00104, + "grad_norm": 0.8588890352637412, + "learning_rate": 0.000312, + "loss": 8.5113, + "step": 104 + }, + { + "epoch": 0.00105, + "grad_norm": 0.862875507176608, + "learning_rate": 0.000315, + "loss": 8.4848, + "step": 105 + }, + { + "epoch": 0.00106, + "grad_norm": 0.858707935395144, + "learning_rate": 0.000318, + "loss": 8.4562, + "step": 106 + }, + { + "epoch": 0.00107, + "grad_norm": 0.8477892944676333, + "learning_rate": 0.000321, + "loss": 8.4313, + "step": 107 + }, + { + "epoch": 0.00108, + "grad_norm": 0.8523289020593828, + "learning_rate": 0.000324, + "loss": 8.4028, + "step": 108 + }, + { + "epoch": 0.00109, + "grad_norm": 0.847730490341604, + "learning_rate": 0.000327, + "loss": 8.3837, + "step": 109 + }, + { + "epoch": 0.0011, + "grad_norm": 0.8528691839316361, + "learning_rate": 0.00033, + "loss": 8.3605, + "step": 110 + }, + { + "epoch": 0.00111, + "grad_norm": 0.8490568938398722, + "learning_rate": 0.000333, + "loss": 8.3302, + "step": 111 + }, + { + "epoch": 0.00112, + "grad_norm": 0.8687159568709311, + "learning_rate": 0.00033600000000000004, + "loss": 8.3067, + "step": 112 + }, + { + "epoch": 0.00113, + "grad_norm": 0.9514329048402193, + "learning_rate": 0.000339, + "loss": 8.2682, + "step": 113 + }, + { + "epoch": 0.00114, + "grad_norm": 1.0232734531890686, + "learning_rate": 0.000342, + "loss": 8.2639, + "step": 114 + }, + { + "epoch": 0.00115, + "grad_norm": 0.88782361437861, + "learning_rate": 0.00034500000000000004, + "loss": 8.2339, + "step": 115 + }, + { + "epoch": 0.00116, + "grad_norm": 0.8477580889036649, + "learning_rate": 0.000348, + "loss": 8.2115, + "step": 116 + }, + { + "epoch": 0.00117, + "grad_norm": 0.9025353672223668, + "learning_rate": 0.000351, + "loss": 8.182, + "step": 117 + }, + { + "epoch": 0.00118, + "grad_norm": 0.8194804043714383, + "learning_rate": 0.000354, + "loss": 8.1632, + "step": 118 + }, + { + "epoch": 0.00119, + "grad_norm": 0.8747187640572928, + "learning_rate": 0.000357, + "loss": 8.1526, + "step": 119 + }, + { + "epoch": 0.0012, + "grad_norm": 0.819302051650556, + "learning_rate": 0.00035999999999999997, + "loss": 8.125, + "step": 120 + }, + { + "epoch": 0.00121, + "grad_norm": 0.8357670289489353, + "learning_rate": 0.000363, + "loss": 8.0977, + "step": 121 + }, + { + "epoch": 0.00122, + "grad_norm": 0.8269622916799205, + "learning_rate": 0.000366, + "loss": 8.0806, + "step": 122 + }, + { + "epoch": 0.00123, + "grad_norm": 0.7923909888868983, + "learning_rate": 0.000369, + "loss": 8.0558, + "step": 123 + }, + { + "epoch": 0.00124, + "grad_norm": 0.8405159164509748, + "learning_rate": 0.000372, + "loss": 8.0354, + "step": 124 + }, + { + "epoch": 0.00125, + "grad_norm": 0.7957014508382675, + "learning_rate": 0.000375, + "loss": 8.0077, + "step": 125 + }, + { + "epoch": 0.00126, + "grad_norm": 0.8032762193906265, + "learning_rate": 0.000378, + "loss": 7.9954, + "step": 126 + }, + { + "epoch": 0.00127, + "grad_norm": 0.8534337354315669, + "learning_rate": 0.000381, + "loss": 7.9712, + "step": 127 + }, + { + "epoch": 0.00128, + "grad_norm": 0.7889902179927859, + "learning_rate": 0.000384, + "loss": 7.9489, + "step": 128 + }, + { + "epoch": 0.00129, + "grad_norm": 0.7667804338310586, + "learning_rate": 0.00038700000000000003, + "loss": 7.936, + "step": 129 + }, + { + "epoch": 0.0013, + "grad_norm": 0.9090077886672961, + "learning_rate": 0.00039000000000000005, + "loss": 7.9113, + "step": 130 + }, + { + "epoch": 0.00131, + "grad_norm": 1.1017726636538738, + "learning_rate": 0.000393, + "loss": 7.8813, + "step": 131 + }, + { + "epoch": 0.00132, + "grad_norm": 0.9363739073599702, + "learning_rate": 0.00039600000000000003, + "loss": 7.8715, + "step": 132 + }, + { + "epoch": 0.00133, + "grad_norm": 0.7265974712286192, + "learning_rate": 0.00039900000000000005, + "loss": 7.8436, + "step": 133 + }, + { + "epoch": 0.00134, + "grad_norm": 1.085264808454749, + "learning_rate": 0.000402, + "loss": 7.8358, + "step": 134 + }, + { + "epoch": 0.00135, + "grad_norm": 1.1024504993278263, + "learning_rate": 0.00040500000000000003, + "loss": 7.8094, + "step": 135 + }, + { + "epoch": 0.00136, + "grad_norm": 0.7406778288782102, + "learning_rate": 0.00040800000000000005, + "loss": 7.7862, + "step": 136 + }, + { + "epoch": 0.00137, + "grad_norm": 0.948337541042023, + "learning_rate": 0.000411, + "loss": 7.7688, + "step": 137 + }, + { + "epoch": 0.00138, + "grad_norm": 0.9084991981726261, + "learning_rate": 0.00041400000000000003, + "loss": 7.7417, + "step": 138 + }, + { + "epoch": 0.00139, + "grad_norm": 0.7074494223870315, + "learning_rate": 0.00041700000000000005, + "loss": 7.7227, + "step": 139 + }, + { + "epoch": 0.0014, + "grad_norm": 0.7974190203181878, + "learning_rate": 0.00042000000000000007, + "loss": 7.7086, + "step": 140 + }, + { + "epoch": 0.00141, + "grad_norm": 0.6861992297830735, + "learning_rate": 0.000423, + "loss": 7.6913, + "step": 141 + }, + { + "epoch": 0.00142, + "grad_norm": 0.7349980301116962, + "learning_rate": 0.00042599999999999995, + "loss": 7.677, + "step": 142 + }, + { + "epoch": 0.00143, + "grad_norm": 0.7684994731116016, + "learning_rate": 0.00042899999999999997, + "loss": 7.6484, + "step": 143 + }, + { + "epoch": 0.00144, + "grad_norm": 0.6719380878952201, + "learning_rate": 0.000432, + "loss": 7.629, + "step": 144 + }, + { + "epoch": 0.00145, + "grad_norm": 0.7001362890446781, + "learning_rate": 0.000435, + "loss": 7.6232, + "step": 145 + }, + { + "epoch": 0.00146, + "grad_norm": 0.7384188504614574, + "learning_rate": 0.00043799999999999997, + "loss": 7.5849, + "step": 146 + }, + { + "epoch": 0.00147, + "grad_norm": 0.6536906507985312, + "learning_rate": 0.000441, + "loss": 7.5755, + "step": 147 + }, + { + "epoch": 0.00148, + "grad_norm": 0.6570387947682329, + "learning_rate": 0.000444, + "loss": 7.5632, + "step": 148 + }, + { + "epoch": 0.00149, + "grad_norm": 0.7127206041028632, + "learning_rate": 0.00044699999999999997, + "loss": 7.5571, + "step": 149 + }, + { + "epoch": 0.0015, + "grad_norm": 0.7308289900379658, + "learning_rate": 0.00045, + "loss": 7.5253, + "step": 150 + }, + { + "epoch": 0.00151, + "grad_norm": 0.8664894863114535, + "learning_rate": 0.000453, + "loss": 7.505, + "step": 151 + }, + { + "epoch": 0.00152, + "grad_norm": 0.8895841422117233, + "learning_rate": 0.000456, + "loss": 7.5034, + "step": 152 + }, + { + "epoch": 0.00153, + "grad_norm": 0.662621907628313, + "learning_rate": 0.000459, + "loss": 7.4855, + "step": 153 + }, + { + "epoch": 0.00154, + "grad_norm": 0.7176449004695503, + "learning_rate": 0.000462, + "loss": 7.4655, + "step": 154 + }, + { + "epoch": 0.00155, + "grad_norm": 0.7407325792933065, + "learning_rate": 0.000465, + "loss": 7.4508, + "step": 155 + }, + { + "epoch": 0.00156, + "grad_norm": 0.7037717820949844, + "learning_rate": 0.000468, + "loss": 7.4265, + "step": 156 + }, + { + "epoch": 0.00157, + "grad_norm": 0.5688508632010355, + "learning_rate": 0.000471, + "loss": 7.4159, + "step": 157 + }, + { + "epoch": 0.00158, + "grad_norm": 0.812094320370736, + "learning_rate": 0.00047400000000000003, + "loss": 7.4015, + "step": 158 + }, + { + "epoch": 0.00159, + "grad_norm": 0.6851465337071334, + "learning_rate": 0.000477, + "loss": 7.3879, + "step": 159 + }, + { + "epoch": 0.0016, + "grad_norm": 0.9553203809874583, + "learning_rate": 0.00048, + "loss": 7.3739, + "step": 160 + }, + { + "epoch": 0.00161, + "grad_norm": 0.7328141993841643, + "learning_rate": 0.00048300000000000003, + "loss": 7.3497, + "step": 161 + }, + { + "epoch": 0.00162, + "grad_norm": 0.6781537584586433, + "learning_rate": 0.00048600000000000005, + "loss": 7.3535, + "step": 162 + }, + { + "epoch": 0.00163, + "grad_norm": 0.5807738383940755, + "learning_rate": 0.0004890000000000001, + "loss": 7.3306, + "step": 163 + }, + { + "epoch": 0.00164, + "grad_norm": 0.6466438270814104, + "learning_rate": 0.000492, + "loss": 7.3371, + "step": 164 + }, + { + "epoch": 0.00165, + "grad_norm": 0.4365334647721832, + "learning_rate": 0.000495, + "loss": 7.3038, + "step": 165 + }, + { + "epoch": 0.00166, + "grad_norm": 0.7387692677149859, + "learning_rate": 0.0004980000000000001, + "loss": 7.2849, + "step": 166 + }, + { + "epoch": 0.00167, + "grad_norm": 0.6489238814400348, + "learning_rate": 0.000501, + "loss": 7.2765, + "step": 167 + }, + { + "epoch": 0.00168, + "grad_norm": 0.4943076416069544, + "learning_rate": 0.000504, + "loss": 7.2512, + "step": 168 + }, + { + "epoch": 0.00169, + "grad_norm": 0.6623047646201438, + "learning_rate": 0.0005070000000000001, + "loss": 7.2485, + "step": 169 + }, + { + "epoch": 0.0017, + "grad_norm": 0.5841690806336316, + "learning_rate": 0.00051, + "loss": 7.2331, + "step": 170 + }, + { + "epoch": 0.00171, + "grad_norm": 0.455313434659774, + "learning_rate": 0.000513, + "loss": 7.2132, + "step": 171 + }, + { + "epoch": 0.00172, + "grad_norm": 0.6012704724036779, + "learning_rate": 0.000516, + "loss": 7.214, + "step": 172 + }, + { + "epoch": 0.00173, + "grad_norm": 0.5433283029678282, + "learning_rate": 0.0005189999999999999, + "loss": 7.1993, + "step": 173 + }, + { + "epoch": 0.00174, + "grad_norm": 0.38251135313099477, + "learning_rate": 0.000522, + "loss": 7.2121, + "step": 174 + }, + { + "epoch": 0.00175, + "grad_norm": 0.6483519243845758, + "learning_rate": 0.000525, + "loss": 7.1796, + "step": 175 + }, + { + "epoch": 0.00176, + "grad_norm": 0.5106681215905619, + "learning_rate": 0.0005279999999999999, + "loss": 7.1531, + "step": 176 + }, + { + "epoch": 0.00177, + "grad_norm": 0.3867365673342131, + "learning_rate": 0.000531, + "loss": 7.1566, + "step": 177 + }, + { + "epoch": 0.00178, + "grad_norm": 0.6803005272577738, + "learning_rate": 0.000534, + "loss": 7.1525, + "step": 178 + }, + { + "epoch": 0.00179, + "grad_norm": 0.44918531151657043, + "learning_rate": 0.000537, + "loss": 7.1407, + "step": 179 + }, + { + "epoch": 0.0018, + "grad_norm": 0.39944612038572996, + "learning_rate": 0.00054, + "loss": 7.1121, + "step": 180 + }, + { + "epoch": 0.00181, + "grad_norm": 0.5937704312371144, + "learning_rate": 0.000543, + "loss": 7.1037, + "step": 181 + }, + { + "epoch": 0.00182, + "grad_norm": 0.3956001398507633, + "learning_rate": 0.000546, + "loss": 7.0976, + "step": 182 + }, + { + "epoch": 0.00183, + "grad_norm": 0.36202282333499824, + "learning_rate": 0.000549, + "loss": 7.0864, + "step": 183 + }, + { + "epoch": 0.00184, + "grad_norm": 0.564760920006622, + "learning_rate": 0.000552, + "loss": 7.0785, + "step": 184 + }, + { + "epoch": 0.00185, + "grad_norm": 0.36333201524042624, + "learning_rate": 0.000555, + "loss": 7.0515, + "step": 185 + }, + { + "epoch": 0.00186, + "grad_norm": 0.3517854453126823, + "learning_rate": 0.000558, + "loss": 7.0614, + "step": 186 + }, + { + "epoch": 0.00187, + "grad_norm": 0.5413049623390372, + "learning_rate": 0.000561, + "loss": 7.0394, + "step": 187 + }, + { + "epoch": 0.00188, + "grad_norm": 0.4452202297805183, + "learning_rate": 0.000564, + "loss": 7.0168, + "step": 188 + }, + { + "epoch": 0.00189, + "grad_norm": 0.2888404630990515, + "learning_rate": 0.000567, + "loss": 7.0108, + "step": 189 + }, + { + "epoch": 0.0019, + "grad_norm": 0.5069984229642092, + "learning_rate": 0.00057, + "loss": 6.9967, + "step": 190 + }, + { + "epoch": 0.00191, + "grad_norm": 0.5049159939798347, + "learning_rate": 0.000573, + "loss": 6.9897, + "step": 191 + }, + { + "epoch": 0.00192, + "grad_norm": 0.353459237780404, + "learning_rate": 0.000576, + "loss": 6.9784, + "step": 192 + }, + { + "epoch": 0.00193, + "grad_norm": 0.4583730098712965, + "learning_rate": 0.000579, + "loss": 6.9798, + "step": 193 + }, + { + "epoch": 0.00194, + "grad_norm": 0.5934016188318005, + "learning_rate": 0.000582, + "loss": 6.9757, + "step": 194 + }, + { + "epoch": 0.00195, + "grad_norm": 0.6611510755360627, + "learning_rate": 0.000585, + "loss": 6.9556, + "step": 195 + }, + { + "epoch": 0.00196, + "grad_norm": 0.6794801564658052, + "learning_rate": 0.000588, + "loss": 6.9455, + "step": 196 + }, + { + "epoch": 0.00197, + "grad_norm": 0.6965025324131076, + "learning_rate": 0.000591, + "loss": 6.9424, + "step": 197 + }, + { + "epoch": 0.00198, + "grad_norm": 1.0010424094015926, + "learning_rate": 0.000594, + "loss": 6.9428, + "step": 198 + }, + { + "epoch": 0.00199, + "grad_norm": 0.8473959375492113, + "learning_rate": 0.0005970000000000001, + "loss": 6.9257, + "step": 199 + }, + { + "epoch": 0.002, + "grad_norm": 0.4674164352666973, + "learning_rate": 0.0006000000000000001, + "loss": 6.9135, + "step": 200 + }, + { + "epoch": 0.00201, + "grad_norm": 0.6130101305251902, + "learning_rate": 0.000603, + "loss": 6.8907, + "step": 201 + }, + { + "epoch": 0.00202, + "grad_norm": 0.5766204915712545, + "learning_rate": 0.0006060000000000001, + "loss": 6.8786, + "step": 202 + }, + { + "epoch": 0.00203, + "grad_norm": 0.42726729176690026, + "learning_rate": 0.0006090000000000001, + "loss": 6.8819, + "step": 203 + }, + { + "epoch": 0.00204, + "grad_norm": 0.5161441207128264, + "learning_rate": 0.000612, + "loss": 6.8675, + "step": 204 + }, + { + "epoch": 0.00205, + "grad_norm": 0.44005998133884483, + "learning_rate": 0.000615, + "loss": 6.852, + "step": 205 + }, + { + "epoch": 0.00206, + "grad_norm": 0.4214710856814876, + "learning_rate": 0.000618, + "loss": 6.8455, + "step": 206 + }, + { + "epoch": 0.00207, + "grad_norm": 0.3835330802523119, + "learning_rate": 0.000621, + "loss": 6.8358, + "step": 207 + }, + { + "epoch": 0.00208, + "grad_norm": 0.4118282878888594, + "learning_rate": 0.000624, + "loss": 6.8325, + "step": 208 + }, + { + "epoch": 0.00209, + "grad_norm": 0.35207765117623135, + "learning_rate": 0.000627, + "loss": 6.8108, + "step": 209 + }, + { + "epoch": 0.0021, + "grad_norm": 0.35165038308098223, + "learning_rate": 0.00063, + "loss": 6.8001, + "step": 210 + }, + { + "epoch": 0.00211, + "grad_norm": 0.2907998960376576, + "learning_rate": 0.000633, + "loss": 6.794, + "step": 211 + }, + { + "epoch": 0.00212, + "grad_norm": 0.34183459291083, + "learning_rate": 0.000636, + "loss": 6.7864, + "step": 212 + }, + { + "epoch": 0.00213, + "grad_norm": 0.31625023606106545, + "learning_rate": 0.000639, + "loss": 6.7664, + "step": 213 + }, + { + "epoch": 0.00214, + "grad_norm": 0.3389167462098675, + "learning_rate": 0.000642, + "loss": 6.7645, + "step": 214 + }, + { + "epoch": 0.00215, + "grad_norm": 0.4689783367804683, + "learning_rate": 0.000645, + "loss": 6.7617, + "step": 215 + }, + { + "epoch": 0.00216, + "grad_norm": 0.7554849005001304, + "learning_rate": 0.000648, + "loss": 6.763, + "step": 216 + }, + { + "epoch": 0.00217, + "grad_norm": 1.3852957930548755, + "learning_rate": 0.000651, + "loss": 6.7541, + "step": 217 + }, + { + "epoch": 0.00218, + "grad_norm": 0.6607509792225358, + "learning_rate": 0.000654, + "loss": 6.7405, + "step": 218 + }, + { + "epoch": 0.00219, + "grad_norm": 0.6480761266032634, + "learning_rate": 0.000657, + "loss": 6.7122, + "step": 219 + }, + { + "epoch": 0.0022, + "grad_norm": 1.2227203355321279, + "learning_rate": 0.00066, + "loss": 6.7471, + "step": 220 + }, + { + "epoch": 0.00221, + "grad_norm": 1.1390371544106608, + "learning_rate": 0.0006630000000000001, + "loss": 6.7248, + "step": 221 + }, + { + "epoch": 0.00222, + "grad_norm": 0.7354052143767025, + "learning_rate": 0.000666, + "loss": 6.7013, + "step": 222 + }, + { + "epoch": 0.00223, + "grad_norm": 0.6134943537271429, + "learning_rate": 0.000669, + "loss": 6.6937, + "step": 223 + }, + { + "epoch": 0.00224, + "grad_norm": 0.5502679770385809, + "learning_rate": 0.0006720000000000001, + "loss": 6.6934, + "step": 224 + }, + { + "epoch": 0.00225, + "grad_norm": 0.47738132510350817, + "learning_rate": 0.000675, + "loss": 6.6778, + "step": 225 + }, + { + "epoch": 0.00226, + "grad_norm": 0.507114065421453, + "learning_rate": 0.000678, + "loss": 6.6712, + "step": 226 + }, + { + "epoch": 0.00227, + "grad_norm": 0.4403172481189424, + "learning_rate": 0.0006810000000000001, + "loss": 6.646, + "step": 227 + }, + { + "epoch": 0.00228, + "grad_norm": 0.34083753361811386, + "learning_rate": 0.000684, + "loss": 6.6465, + "step": 228 + }, + { + "epoch": 0.00229, + "grad_norm": 0.3502366193988861, + "learning_rate": 0.000687, + "loss": 6.6418, + "step": 229 + }, + { + "epoch": 0.0023, + "grad_norm": 0.40867670062411404, + "learning_rate": 0.0006900000000000001, + "loss": 6.6218, + "step": 230 + }, + { + "epoch": 0.00231, + "grad_norm": 0.3512396474179769, + "learning_rate": 0.000693, + "loss": 6.605, + "step": 231 + }, + { + "epoch": 0.00232, + "grad_norm": 0.37061772246720376, + "learning_rate": 0.000696, + "loss": 6.6164, + "step": 232 + }, + { + "epoch": 0.00233, + "grad_norm": 0.30975845835339183, + "learning_rate": 0.0006990000000000001, + "loss": 6.6026, + "step": 233 + }, + { + "epoch": 0.00234, + "grad_norm": 0.31056463949410484, + "learning_rate": 0.000702, + "loss": 6.5934, + "step": 234 + }, + { + "epoch": 0.00235, + "grad_norm": 0.3416829938754332, + "learning_rate": 0.000705, + "loss": 6.5798, + "step": 235 + }, + { + "epoch": 0.00236, + "grad_norm": 0.3580449551775802, + "learning_rate": 0.000708, + "loss": 6.5618, + "step": 236 + }, + { + "epoch": 0.00237, + "grad_norm": 0.31186166958910727, + "learning_rate": 0.0007109999999999999, + "loss": 6.5709, + "step": 237 + }, + { + "epoch": 0.00238, + "grad_norm": 0.35712039334009765, + "learning_rate": 0.000714, + "loss": 6.5601, + "step": 238 + }, + { + "epoch": 0.00239, + "grad_norm": 0.4689874670209286, + "learning_rate": 0.000717, + "loss": 6.5431, + "step": 239 + }, + { + "epoch": 0.0024, + "grad_norm": 0.8385556457838154, + "learning_rate": 0.0007199999999999999, + "loss": 6.5652, + "step": 240 + }, + { + "epoch": 0.00241, + "grad_norm": 1.2597764356570753, + "learning_rate": 0.000723, + "loss": 6.5721, + "step": 241 + }, + { + "epoch": 0.00242, + "grad_norm": 0.7018547016449656, + "learning_rate": 0.000726, + "loss": 6.5258, + "step": 242 + }, + { + "epoch": 0.00243, + "grad_norm": 0.9679840331331718, + "learning_rate": 0.000729, + "loss": 6.5235, + "step": 243 + }, + { + "epoch": 0.00244, + "grad_norm": 0.8151244219176279, + "learning_rate": 0.000732, + "loss": 6.5289, + "step": 244 + }, + { + "epoch": 0.00245, + "grad_norm": 0.7408239035858821, + "learning_rate": 0.000735, + "loss": 6.5159, + "step": 245 + }, + { + "epoch": 0.00246, + "grad_norm": 0.7269888029650857, + "learning_rate": 0.000738, + "loss": 6.5006, + "step": 246 + }, + { + "epoch": 0.00247, + "grad_norm": 0.6715573633422388, + "learning_rate": 0.000741, + "loss": 6.4916, + "step": 247 + }, + { + "epoch": 0.00248, + "grad_norm": 0.5355440416613977, + "learning_rate": 0.000744, + "loss": 6.4965, + "step": 248 + }, + { + "epoch": 0.00249, + "grad_norm": 0.47837466185087324, + "learning_rate": 0.000747, + "loss": 6.474, + "step": 249 + }, + { + "epoch": 0.0025, + "grad_norm": 0.5421674013984196, + "learning_rate": 0.00075, + "loss": 6.4781, + "step": 250 + }, + { + "epoch": 0.00251, + "grad_norm": 0.5046285559759315, + "learning_rate": 0.000753, + "loss": 6.4394, + "step": 251 + }, + { + "epoch": 0.00252, + "grad_norm": 0.4902399046559828, + "learning_rate": 0.000756, + "loss": 6.4478, + "step": 252 + }, + { + "epoch": 0.00253, + "grad_norm": 0.6026763597870729, + "learning_rate": 0.000759, + "loss": 6.4417, + "step": 253 + }, + { + "epoch": 0.00254, + "grad_norm": 0.6534629043130656, + "learning_rate": 0.000762, + "loss": 6.4357, + "step": 254 + }, + { + "epoch": 0.00255, + "grad_norm": 0.5879568865224861, + "learning_rate": 0.0007650000000000001, + "loss": 6.4331, + "step": 255 + }, + { + "epoch": 0.00256, + "grad_norm": 0.587549512889128, + "learning_rate": 0.000768, + "loss": 6.4121, + "step": 256 + }, + { + "epoch": 0.00257, + "grad_norm": 0.4948417519702818, + "learning_rate": 0.000771, + "loss": 6.4144, + "step": 257 + }, + { + "epoch": 0.00258, + "grad_norm": 0.500640645286767, + "learning_rate": 0.0007740000000000001, + "loss": 6.3915, + "step": 258 + }, + { + "epoch": 0.00259, + "grad_norm": 0.46096026590523803, + "learning_rate": 0.000777, + "loss": 6.3904, + "step": 259 + }, + { + "epoch": 0.0026, + "grad_norm": 0.36710644781296853, + "learning_rate": 0.0007800000000000001, + "loss": 6.3738, + "step": 260 + }, + { + "epoch": 0.00261, + "grad_norm": 0.3228962429333065, + "learning_rate": 0.0007830000000000001, + "loss": 6.3839, + "step": 261 + }, + { + "epoch": 0.00262, + "grad_norm": 0.33723485694460476, + "learning_rate": 0.000786, + "loss": 6.3601, + "step": 262 + }, + { + "epoch": 0.00263, + "grad_norm": 0.3792844357509868, + "learning_rate": 0.0007890000000000001, + "loss": 6.3664, + "step": 263 + }, + { + "epoch": 0.00264, + "grad_norm": 0.41999467297545995, + "learning_rate": 0.0007920000000000001, + "loss": 6.3505, + "step": 264 + }, + { + "epoch": 0.00265, + "grad_norm": 0.5058327981670172, + "learning_rate": 0.000795, + "loss": 6.3427, + "step": 265 + }, + { + "epoch": 0.00266, + "grad_norm": 0.728666013298328, + "learning_rate": 0.0007980000000000001, + "loss": 6.3445, + "step": 266 + }, + { + "epoch": 0.00267, + "grad_norm": 1.051597167429085, + "learning_rate": 0.0008010000000000001, + "loss": 6.3504, + "step": 267 + }, + { + "epoch": 0.00268, + "grad_norm": 0.8184600443662068, + "learning_rate": 0.000804, + "loss": 6.3425, + "step": 268 + }, + { + "epoch": 0.00269, + "grad_norm": 0.9044150906220578, + "learning_rate": 0.0008070000000000001, + "loss": 6.3249, + "step": 269 + }, + { + "epoch": 0.0027, + "grad_norm": 1.128589514160359, + "learning_rate": 0.0008100000000000001, + "loss": 6.3418, + "step": 270 + }, + { + "epoch": 0.00271, + "grad_norm": 0.8269681270889044, + "learning_rate": 0.000813, + "loss": 6.3324, + "step": 271 + }, + { + "epoch": 0.00272, + "grad_norm": 0.9605488749490486, + "learning_rate": 0.0008160000000000001, + "loss": 6.3229, + "step": 272 + }, + { + "epoch": 0.00273, + "grad_norm": 1.4256959145221717, + "learning_rate": 0.0008190000000000001, + "loss": 6.3332, + "step": 273 + }, + { + "epoch": 0.00274, + "grad_norm": 0.8746481700595169, + "learning_rate": 0.000822, + "loss": 6.301, + "step": 274 + }, + { + "epoch": 0.00275, + "grad_norm": 0.7483839590071554, + "learning_rate": 0.0008250000000000001, + "loss": 6.3047, + "step": 275 + }, + { + "epoch": 0.00276, + "grad_norm": 0.9683696395132253, + "learning_rate": 0.0008280000000000001, + "loss": 6.2965, + "step": 276 + }, + { + "epoch": 0.00277, + "grad_norm": 0.9445364353770574, + "learning_rate": 0.0008310000000000001, + "loss": 6.3021, + "step": 277 + }, + { + "epoch": 0.00278, + "grad_norm": 0.5637393663463164, + "learning_rate": 0.0008340000000000001, + "loss": 6.2711, + "step": 278 + }, + { + "epoch": 0.00279, + "grad_norm": 0.6474632709262491, + "learning_rate": 0.0008370000000000001, + "loss": 6.2722, + "step": 279 + }, + { + "epoch": 0.0028, + "grad_norm": 0.5477535984985658, + "learning_rate": 0.0008400000000000001, + "loss": 6.2548, + "step": 280 + }, + { + "epoch": 0.00281, + "grad_norm": 0.5235139271800053, + "learning_rate": 0.0008430000000000001, + "loss": 6.2428, + "step": 281 + }, + { + "epoch": 0.00282, + "grad_norm": 0.4309641347762167, + "learning_rate": 0.000846, + "loss": 6.2382, + "step": 282 + }, + { + "epoch": 0.00283, + "grad_norm": 0.47598469419389666, + "learning_rate": 0.0008489999999999999, + "loss": 6.2321, + "step": 283 + }, + { + "epoch": 0.00284, + "grad_norm": 0.41791767912211497, + "learning_rate": 0.0008519999999999999, + "loss": 6.2199, + "step": 284 + }, + { + "epoch": 0.00285, + "grad_norm": 0.4068034669968041, + "learning_rate": 0.000855, + "loss": 6.2226, + "step": 285 + }, + { + "epoch": 0.00286, + "grad_norm": 0.461449105404332, + "learning_rate": 0.0008579999999999999, + "loss": 6.2004, + "step": 286 + }, + { + "epoch": 0.00287, + "grad_norm": 0.3833700707849029, + "learning_rate": 0.000861, + "loss": 6.2013, + "step": 287 + }, + { + "epoch": 0.00288, + "grad_norm": 0.3960645710165419, + "learning_rate": 0.000864, + "loss": 6.1818, + "step": 288 + }, + { + "epoch": 0.00289, + "grad_norm": 0.44431212101949524, + "learning_rate": 0.0008669999999999999, + "loss": 6.1961, + "step": 289 + }, + { + "epoch": 0.0029, + "grad_norm": 0.5861155579490576, + "learning_rate": 0.00087, + "loss": 6.1786, + "step": 290 + }, + { + "epoch": 0.00291, + "grad_norm": 0.6639450218120423, + "learning_rate": 0.000873, + "loss": 6.1739, + "step": 291 + }, + { + "epoch": 0.00292, + "grad_norm": 0.8465664162859037, + "learning_rate": 0.0008759999999999999, + "loss": 6.1756, + "step": 292 + }, + { + "epoch": 0.00293, + "grad_norm": 0.9291009677838105, + "learning_rate": 0.000879, + "loss": 6.1618, + "step": 293 + }, + { + "epoch": 0.00294, + "grad_norm": 1.3082966000676708, + "learning_rate": 0.000882, + "loss": 6.1833, + "step": 294 + }, + { + "epoch": 0.00295, + "grad_norm": 1.1893342417992165, + "learning_rate": 0.0008849999999999999, + "loss": 6.1586, + "step": 295 + }, + { + "epoch": 0.00296, + "grad_norm": 0.6546841726146672, + "learning_rate": 0.000888, + "loss": 6.1551, + "step": 296 + }, + { + "epoch": 0.00297, + "grad_norm": 0.8416245966488266, + "learning_rate": 0.000891, + "loss": 6.1315, + "step": 297 + }, + { + "epoch": 0.00298, + "grad_norm": 0.806301942306892, + "learning_rate": 0.0008939999999999999, + "loss": 6.1452, + "step": 298 + }, + { + "epoch": 0.00299, + "grad_norm": 1.0994469943740992, + "learning_rate": 0.000897, + "loss": 6.1288, + "step": 299 + }, + { + "epoch": 0.003, + "grad_norm": 0.7425213774845364, + "learning_rate": 0.0009, + "loss": 6.123, + "step": 300 + }, + { + "epoch": 0.00301, + "grad_norm": 0.6433032250398036, + "learning_rate": 0.0009029999999999999, + "loss": 6.1149, + "step": 301 + }, + { + "epoch": 0.00302, + "grad_norm": 0.5449990789285414, + "learning_rate": 0.000906, + "loss": 6.0952, + "step": 302 + }, + { + "epoch": 0.00303, + "grad_norm": 0.5356674479383384, + "learning_rate": 0.000909, + "loss": 6.1067, + "step": 303 + }, + { + "epoch": 0.00304, + "grad_norm": 0.5049891146020773, + "learning_rate": 0.000912, + "loss": 6.1013, + "step": 304 + }, + { + "epoch": 0.00305, + "grad_norm": 0.48068374440179407, + "learning_rate": 0.000915, + "loss": 6.0985, + "step": 305 + }, + { + "epoch": 0.00306, + "grad_norm": 0.39305660615027715, + "learning_rate": 0.000918, + "loss": 6.0835, + "step": 306 + }, + { + "epoch": 0.00307, + "grad_norm": 0.4044411514506039, + "learning_rate": 0.000921, + "loss": 6.0774, + "step": 307 + }, + { + "epoch": 0.00308, + "grad_norm": 0.3317383414417003, + "learning_rate": 0.000924, + "loss": 6.0578, + "step": 308 + }, + { + "epoch": 0.00309, + "grad_norm": 0.4170081186564571, + "learning_rate": 0.000927, + "loss": 6.0741, + "step": 309 + }, + { + "epoch": 0.0031, + "grad_norm": 0.3537372632819937, + "learning_rate": 0.00093, + "loss": 6.0544, + "step": 310 + }, + { + "epoch": 0.00311, + "grad_norm": 0.3807795877969161, + "learning_rate": 0.000933, + "loss": 6.0455, + "step": 311 + }, + { + "epoch": 0.00312, + "grad_norm": 0.36301395919450175, + "learning_rate": 0.000936, + "loss": 6.025, + "step": 312 + }, + { + "epoch": 0.00313, + "grad_norm": 0.456154645430038, + "learning_rate": 0.0009390000000000001, + "loss": 6.0399, + "step": 313 + }, + { + "epoch": 0.00314, + "grad_norm": 0.6252307526263491, + "learning_rate": 0.000942, + "loss": 6.048, + "step": 314 + }, + { + "epoch": 0.00315, + "grad_norm": 1.0054536709858048, + "learning_rate": 0.000945, + "loss": 6.0332, + "step": 315 + }, + { + "epoch": 0.00316, + "grad_norm": 1.2671801420052267, + "learning_rate": 0.0009480000000000001, + "loss": 6.0322, + "step": 316 + }, + { + "epoch": 0.00317, + "grad_norm": 0.5267566088823273, + "learning_rate": 0.000951, + "loss": 6.0198, + "step": 317 + }, + { + "epoch": 0.00318, + "grad_norm": 1.2005145600955376, + "learning_rate": 0.000954, + "loss": 6.041, + "step": 318 + }, + { + "epoch": 0.00319, + "grad_norm": 1.004603435172909, + "learning_rate": 0.0009570000000000001, + "loss": 6.025, + "step": 319 + }, + { + "epoch": 0.0032, + "grad_norm": 1.1201777687880106, + "learning_rate": 0.00096, + "loss": 6.0356, + "step": 320 + }, + { + "epoch": 0.00321, + "grad_norm": 1.0167944206203925, + "learning_rate": 0.000963, + "loss": 6.012, + "step": 321 + }, + { + "epoch": 0.00322, + "grad_norm": 1.374478470085335, + "learning_rate": 0.0009660000000000001, + "loss": 6.025, + "step": 322 + }, + { + "epoch": 0.00323, + "grad_norm": 0.7765654772896999, + "learning_rate": 0.000969, + "loss": 6.0057, + "step": 323 + }, + { + "epoch": 0.00324, + "grad_norm": 0.6987857474960232, + "learning_rate": 0.0009720000000000001, + "loss": 6.0035, + "step": 324 + }, + { + "epoch": 0.00325, + "grad_norm": 0.7564069291902213, + "learning_rate": 0.0009750000000000001, + "loss": 5.9742, + "step": 325 + }, + { + "epoch": 0.00326, + "grad_norm": 0.7275905653375951, + "learning_rate": 0.0009780000000000001, + "loss": 5.9877, + "step": 326 + }, + { + "epoch": 0.00327, + "grad_norm": 0.6984989438722892, + "learning_rate": 0.000981, + "loss": 5.9857, + "step": 327 + }, + { + "epoch": 0.00328, + "grad_norm": 0.721382136086564, + "learning_rate": 0.000984, + "loss": 5.9627, + "step": 328 + }, + { + "epoch": 0.00329, + "grad_norm": 0.7841656285031653, + "learning_rate": 0.000987, + "loss": 5.9649, + "step": 329 + }, + { + "epoch": 0.0033, + "grad_norm": 0.8334397806028965, + "learning_rate": 0.00099, + "loss": 5.9568, + "step": 330 + }, + { + "epoch": 0.00331, + "grad_norm": 1.070828806370832, + "learning_rate": 0.0009930000000000002, + "loss": 5.9642, + "step": 331 + }, + { + "epoch": 0.00332, + "grad_norm": 1.0826703064607723, + "learning_rate": 0.0009960000000000001, + "loss": 5.9538, + "step": 332 + }, + { + "epoch": 0.00333, + "grad_norm": 0.9458285226462658, + "learning_rate": 0.000999, + "loss": 5.9361, + "step": 333 + }, + { + "epoch": 0.00334, + "grad_norm": 0.8783572499928525, + "learning_rate": 0.001002, + "loss": 5.9443, + "step": 334 + }, + { + "epoch": 0.00335, + "grad_norm": 0.5654549281403035, + "learning_rate": 0.001005, + "loss": 5.9354, + "step": 335 + }, + { + "epoch": 0.00336, + "grad_norm": 0.6516350129184051, + "learning_rate": 0.001008, + "loss": 5.9285, + "step": 336 + }, + { + "epoch": 0.00337, + "grad_norm": 0.5946351510133966, + "learning_rate": 0.0010110000000000002, + "loss": 5.9026, + "step": 337 + }, + { + "epoch": 0.00338, + "grad_norm": 0.46824568572142555, + "learning_rate": 0.0010140000000000001, + "loss": 5.9165, + "step": 338 + }, + { + "epoch": 0.00339, + "grad_norm": 0.5098867793571775, + "learning_rate": 0.0010170000000000001, + "loss": 5.9215, + "step": 339 + }, + { + "epoch": 0.0034, + "grad_norm": 0.5202026348670868, + "learning_rate": 0.00102, + "loss": 5.8881, + "step": 340 + }, + { + "epoch": 0.00341, + "grad_norm": 0.4748523902202147, + "learning_rate": 0.001023, + "loss": 5.8705, + "step": 341 + }, + { + "epoch": 0.00342, + "grad_norm": 0.42065780247182966, + "learning_rate": 0.001026, + "loss": 5.8732, + "step": 342 + }, + { + "epoch": 0.00343, + "grad_norm": 0.4385189981740614, + "learning_rate": 0.0010290000000000002, + "loss": 5.8762, + "step": 343 + }, + { + "epoch": 0.00344, + "grad_norm": 0.44131626809067537, + "learning_rate": 0.001032, + "loss": 5.8646, + "step": 344 + }, + { + "epoch": 0.00345, + "grad_norm": 0.5602404326473228, + "learning_rate": 0.001035, + "loss": 5.8576, + "step": 345 + }, + { + "epoch": 0.00346, + "grad_norm": 0.7338782884674983, + "learning_rate": 0.0010379999999999999, + "loss": 5.8684, + "step": 346 + }, + { + "epoch": 0.00347, + "grad_norm": 0.8871217127911775, + "learning_rate": 0.001041, + "loss": 5.8429, + "step": 347 + }, + { + "epoch": 0.00348, + "grad_norm": 0.7294005585257834, + "learning_rate": 0.001044, + "loss": 5.8498, + "step": 348 + }, + { + "epoch": 0.00349, + "grad_norm": 0.672488440407315, + "learning_rate": 0.001047, + "loss": 5.8268, + "step": 349 + }, + { + "epoch": 0.0035, + "grad_norm": 0.944274771030096, + "learning_rate": 0.00105, + "loss": 5.8574, + "step": 350 + }, + { + "epoch": 0.00351, + "grad_norm": 0.8431353927439175, + "learning_rate": 0.001053, + "loss": 5.8586, + "step": 351 + }, + { + "epoch": 0.00352, + "grad_norm": 0.8223595555288745, + "learning_rate": 0.0010559999999999999, + "loss": 5.8326, + "step": 352 + }, + { + "epoch": 0.00353, + "grad_norm": 0.7412573085023916, + "learning_rate": 0.001059, + "loss": 5.8336, + "step": 353 + }, + { + "epoch": 0.00354, + "grad_norm": 0.7219192917744488, + "learning_rate": 0.001062, + "loss": 5.8047, + "step": 354 + }, + { + "epoch": 0.00355, + "grad_norm": 0.7332482461779326, + "learning_rate": 0.001065, + "loss": 5.8213, + "step": 355 + }, + { + "epoch": 0.00356, + "grad_norm": 0.6300796792438244, + "learning_rate": 0.001068, + "loss": 5.8096, + "step": 356 + }, + { + "epoch": 0.00357, + "grad_norm": 0.7796926368866208, + "learning_rate": 0.001071, + "loss": 5.802, + "step": 357 + }, + { + "epoch": 0.00358, + "grad_norm": 1.048118676098695, + "learning_rate": 0.001074, + "loss": 5.8089, + "step": 358 + }, + { + "epoch": 0.00359, + "grad_norm": 0.8469283052653666, + "learning_rate": 0.001077, + "loss": 5.7908, + "step": 359 + }, + { + "epoch": 0.0036, + "grad_norm": 0.824969051981288, + "learning_rate": 0.00108, + "loss": 5.7948, + "step": 360 + }, + { + "epoch": 0.00361, + "grad_norm": 0.8007424119602606, + "learning_rate": 0.001083, + "loss": 5.7876, + "step": 361 + }, + { + "epoch": 0.00362, + "grad_norm": 0.7740442530973947, + "learning_rate": 0.001086, + "loss": 5.7713, + "step": 362 + }, + { + "epoch": 0.00363, + "grad_norm": 0.8217790100779416, + "learning_rate": 0.001089, + "loss": 5.7673, + "step": 363 + }, + { + "epoch": 0.00364, + "grad_norm": 0.9222078053619298, + "learning_rate": 0.001092, + "loss": 5.7813, + "step": 364 + }, + { + "epoch": 0.00365, + "grad_norm": 1.3725693861639678, + "learning_rate": 0.001095, + "loss": 5.801, + "step": 365 + }, + { + "epoch": 0.00366, + "grad_norm": 0.7809021294214551, + "learning_rate": 0.001098, + "loss": 5.7587, + "step": 366 + }, + { + "epoch": 0.00367, + "grad_norm": 1.1843193874441642, + "learning_rate": 0.001101, + "loss": 5.7731, + "step": 367 + }, + { + "epoch": 0.00368, + "grad_norm": 0.7875897064576857, + "learning_rate": 0.001104, + "loss": 5.7636, + "step": 368 + }, + { + "epoch": 0.00369, + "grad_norm": 0.6858717859751121, + "learning_rate": 0.001107, + "loss": 5.7637, + "step": 369 + }, + { + "epoch": 0.0037, + "grad_norm": 0.6252688126791845, + "learning_rate": 0.00111, + "loss": 5.7428, + "step": 370 + }, + { + "epoch": 0.00371, + "grad_norm": 0.6219765776345385, + "learning_rate": 0.001113, + "loss": 5.7116, + "step": 371 + }, + { + "epoch": 0.00372, + "grad_norm": 0.5998751494739524, + "learning_rate": 0.001116, + "loss": 5.7331, + "step": 372 + }, + { + "epoch": 0.00373, + "grad_norm": 0.5714053922949759, + "learning_rate": 0.001119, + "loss": 5.72, + "step": 373 + }, + { + "epoch": 0.00374, + "grad_norm": 0.628102864915935, + "learning_rate": 0.001122, + "loss": 5.7262, + "step": 374 + }, + { + "epoch": 0.00375, + "grad_norm": 0.5909991857975568, + "learning_rate": 0.0011250000000000001, + "loss": 5.7276, + "step": 375 + }, + { + "epoch": 0.00376, + "grad_norm": 0.5216592986380127, + "learning_rate": 0.001128, + "loss": 5.72, + "step": 376 + }, + { + "epoch": 0.00377, + "grad_norm": 0.5469626358369448, + "learning_rate": 0.001131, + "loss": 5.7145, + "step": 377 + }, + { + "epoch": 0.00378, + "grad_norm": 0.6291207985788981, + "learning_rate": 0.001134, + "loss": 5.7123, + "step": 378 + }, + { + "epoch": 0.00379, + "grad_norm": 0.6445779804368097, + "learning_rate": 0.001137, + "loss": 5.698, + "step": 379 + }, + { + "epoch": 0.0038, + "grad_norm": 0.7034810438425685, + "learning_rate": 0.00114, + "loss": 5.7177, + "step": 380 + }, + { + "epoch": 0.00381, + "grad_norm": 0.9330449756794958, + "learning_rate": 0.0011430000000000001, + "loss": 5.6924, + "step": 381 + }, + { + "epoch": 0.00382, + "grad_norm": 0.9449582673655755, + "learning_rate": 0.001146, + "loss": 5.6863, + "step": 382 + }, + { + "epoch": 0.00383, + "grad_norm": 0.6715143882341864, + "learning_rate": 0.001149, + "loss": 5.6846, + "step": 383 + }, + { + "epoch": 0.00384, + "grad_norm": 0.8662082921806001, + "learning_rate": 0.001152, + "loss": 5.699, + "step": 384 + }, + { + "epoch": 0.00385, + "grad_norm": 0.8482542645994808, + "learning_rate": 0.001155, + "loss": 5.6818, + "step": 385 + }, + { + "epoch": 0.00386, + "grad_norm": 0.8694120663581818, + "learning_rate": 0.001158, + "loss": 5.6914, + "step": 386 + }, + { + "epoch": 0.00387, + "grad_norm": 1.382803938865885, + "learning_rate": 0.0011610000000000001, + "loss": 5.6813, + "step": 387 + }, + { + "epoch": 0.00388, + "grad_norm": 0.9644228724095026, + "learning_rate": 0.001164, + "loss": 5.6689, + "step": 388 + }, + { + "epoch": 0.00389, + "grad_norm": 0.7396617808819147, + "learning_rate": 0.001167, + "loss": 5.6709, + "step": 389 + }, + { + "epoch": 0.0039, + "grad_norm": 0.5609185173455993, + "learning_rate": 0.00117, + "loss": 5.648, + "step": 390 + }, + { + "epoch": 0.00391, + "grad_norm": 0.735416794156057, + "learning_rate": 0.001173, + "loss": 5.6589, + "step": 391 + }, + { + "epoch": 0.00392, + "grad_norm": 0.7564046131732848, + "learning_rate": 0.001176, + "loss": 5.6662, + "step": 392 + }, + { + "epoch": 0.00393, + "grad_norm": 0.7233997204719024, + "learning_rate": 0.0011790000000000001, + "loss": 5.6377, + "step": 393 + }, + { + "epoch": 0.00394, + "grad_norm": 0.6288246845478384, + "learning_rate": 0.001182, + "loss": 5.6239, + "step": 394 + }, + { + "epoch": 0.00395, + "grad_norm": 0.6696319293361586, + "learning_rate": 0.001185, + "loss": 5.6102, + "step": 395 + }, + { + "epoch": 0.00396, + "grad_norm": 0.7214553632515296, + "learning_rate": 0.001188, + "loss": 5.6276, + "step": 396 + }, + { + "epoch": 0.00397, + "grad_norm": 0.7104651338358826, + "learning_rate": 0.001191, + "loss": 5.619, + "step": 397 + }, + { + "epoch": 0.00398, + "grad_norm": 0.5955487966253655, + "learning_rate": 0.0011940000000000002, + "loss": 5.6195, + "step": 398 + }, + { + "epoch": 0.00399, + "grad_norm": 0.6061151049974988, + "learning_rate": 0.0011970000000000001, + "loss": 5.6211, + "step": 399 + }, + { + "epoch": 0.004, + "grad_norm": 0.747975845260549, + "learning_rate": 0.0012000000000000001, + "loss": 5.6119, + "step": 400 + }, + { + "epoch": 0.00401, + "grad_norm": 1.066378470212407, + "learning_rate": 0.001203, + "loss": 5.6055, + "step": 401 + }, + { + "epoch": 0.00402, + "grad_norm": 1.1521859370487337, + "learning_rate": 0.001206, + "loss": 5.6144, + "step": 402 + }, + { + "epoch": 0.00403, + "grad_norm": 1.0137206411467539, + "learning_rate": 0.001209, + "loss": 5.6121, + "step": 403 + }, + { + "epoch": 0.00404, + "grad_norm": 1.1663989264954397, + "learning_rate": 0.0012120000000000002, + "loss": 5.6177, + "step": 404 + }, + { + "epoch": 0.00405, + "grad_norm": 1.1363050593119097, + "learning_rate": 0.0012150000000000002, + "loss": 5.6098, + "step": 405 + }, + { + "epoch": 0.00406, + "grad_norm": 1.0087910495777368, + "learning_rate": 0.0012180000000000001, + "loss": 5.5952, + "step": 406 + }, + { + "epoch": 0.00407, + "grad_norm": 0.9520738669869687, + "learning_rate": 0.0012209999999999999, + "loss": 5.5915, + "step": 407 + }, + { + "epoch": 0.00408, + "grad_norm": 1.0377478362348644, + "learning_rate": 0.001224, + "loss": 5.6129, + "step": 408 + }, + { + "epoch": 0.00409, + "grad_norm": 1.1016437762309168, + "learning_rate": 0.001227, + "loss": 5.5927, + "step": 409 + }, + { + "epoch": 0.0041, + "grad_norm": 0.9709841842795465, + "learning_rate": 0.00123, + "loss": 5.5883, + "step": 410 + }, + { + "epoch": 0.00411, + "grad_norm": 0.8240534275873757, + "learning_rate": 0.001233, + "loss": 5.5977, + "step": 411 + }, + { + "epoch": 0.00412, + "grad_norm": 0.8077831494606339, + "learning_rate": 0.001236, + "loss": 5.5894, + "step": 412 + }, + { + "epoch": 0.00413, + "grad_norm": 0.7433003641205871, + "learning_rate": 0.0012389999999999999, + "loss": 5.5815, + "step": 413 + }, + { + "epoch": 0.00414, + "grad_norm": 0.615571577509929, + "learning_rate": 0.001242, + "loss": 5.5582, + "step": 414 + }, + { + "epoch": 0.00415, + "grad_norm": 0.656841702591939, + "learning_rate": 0.001245, + "loss": 5.5606, + "step": 415 + }, + { + "epoch": 0.00416, + "grad_norm": 0.613196865035469, + "learning_rate": 0.001248, + "loss": 5.5418, + "step": 416 + }, + { + "epoch": 0.00417, + "grad_norm": 0.48631699887544017, + "learning_rate": 0.001251, + "loss": 5.5683, + "step": 417 + }, + { + "epoch": 0.00418, + "grad_norm": 0.4611320175617472, + "learning_rate": 0.001254, + "loss": 5.5401, + "step": 418 + }, + { + "epoch": 0.00419, + "grad_norm": 0.5824396705507784, + "learning_rate": 0.0012569999999999999, + "loss": 5.5305, + "step": 419 + }, + { + "epoch": 0.0042, + "grad_norm": 0.8055087238093066, + "learning_rate": 0.00126, + "loss": 5.5467, + "step": 420 + }, + { + "epoch": 0.00421, + "grad_norm": 0.9831917562206494, + "learning_rate": 0.001263, + "loss": 5.5503, + "step": 421 + }, + { + "epoch": 0.00422, + "grad_norm": 0.8124592037579013, + "learning_rate": 0.001266, + "loss": 5.5357, + "step": 422 + }, + { + "epoch": 0.00423, + "grad_norm": 1.0116811247233115, + "learning_rate": 0.001269, + "loss": 5.5337, + "step": 423 + }, + { + "epoch": 0.00424, + "grad_norm": 1.368200107021336, + "learning_rate": 0.001272, + "loss": 5.5595, + "step": 424 + }, + { + "epoch": 0.00425, + "grad_norm": 0.9353292964204464, + "learning_rate": 0.001275, + "loss": 5.5445, + "step": 425 + }, + { + "epoch": 0.00426, + "grad_norm": 0.9604600896045714, + "learning_rate": 0.001278, + "loss": 5.5493, + "step": 426 + }, + { + "epoch": 0.00427, + "grad_norm": 0.9684632680550875, + "learning_rate": 0.001281, + "loss": 5.5328, + "step": 427 + }, + { + "epoch": 0.00428, + "grad_norm": 0.9533788870858234, + "learning_rate": 0.001284, + "loss": 5.5058, + "step": 428 + }, + { + "epoch": 0.00429, + "grad_norm": 1.113271451738206, + "learning_rate": 0.001287, + "loss": 5.5322, + "step": 429 + }, + { + "epoch": 0.0043, + "grad_norm": 0.9492520989573756, + "learning_rate": 0.00129, + "loss": 5.5226, + "step": 430 + }, + { + "epoch": 0.00431, + "grad_norm": 1.1309185966069144, + "learning_rate": 0.001293, + "loss": 5.5118, + "step": 431 + }, + { + "epoch": 0.00432, + "grad_norm": 1.2851571092183034, + "learning_rate": 0.001296, + "loss": 5.5367, + "step": 432 + }, + { + "epoch": 0.00433, + "grad_norm": 0.7219415163940603, + "learning_rate": 0.001299, + "loss": 5.5106, + "step": 433 + }, + { + "epoch": 0.00434, + "grad_norm": 0.7943786167833424, + "learning_rate": 0.001302, + "loss": 5.5109, + "step": 434 + }, + { + "epoch": 0.00435, + "grad_norm": 0.7514833526497924, + "learning_rate": 0.001305, + "loss": 5.5051, + "step": 435 + }, + { + "epoch": 0.00436, + "grad_norm": 0.5499552719714149, + "learning_rate": 0.001308, + "loss": 5.4758, + "step": 436 + }, + { + "epoch": 0.00437, + "grad_norm": 0.5803132978630802, + "learning_rate": 0.001311, + "loss": 5.494, + "step": 437 + }, + { + "epoch": 0.00438, + "grad_norm": 0.4343754614294673, + "learning_rate": 0.001314, + "loss": 5.4701, + "step": 438 + }, + { + "epoch": 0.00439, + "grad_norm": 0.49450539118744047, + "learning_rate": 0.001317, + "loss": 5.4997, + "step": 439 + }, + { + "epoch": 0.0044, + "grad_norm": 0.3670496641121159, + "learning_rate": 0.00132, + "loss": 5.4633, + "step": 440 + }, + { + "epoch": 0.00441, + "grad_norm": 0.44023069828614625, + "learning_rate": 0.001323, + "loss": 5.4624, + "step": 441 + }, + { + "epoch": 0.00442, + "grad_norm": 0.45124022563783683, + "learning_rate": 0.0013260000000000001, + "loss": 5.4541, + "step": 442 + }, + { + "epoch": 0.00443, + "grad_norm": 0.5170657075032803, + "learning_rate": 0.001329, + "loss": 5.4572, + "step": 443 + }, + { + "epoch": 0.00444, + "grad_norm": 0.6406930762525409, + "learning_rate": 0.001332, + "loss": 5.4608, + "step": 444 + }, + { + "epoch": 0.00445, + "grad_norm": 0.7431205120059805, + "learning_rate": 0.001335, + "loss": 5.4462, + "step": 445 + }, + { + "epoch": 0.00446, + "grad_norm": 0.9849435734941417, + "learning_rate": 0.001338, + "loss": 5.4578, + "step": 446 + }, + { + "epoch": 0.00447, + "grad_norm": 1.1031854356255555, + "learning_rate": 0.001341, + "loss": 5.469, + "step": 447 + }, + { + "epoch": 0.00448, + "grad_norm": 0.8550264287932019, + "learning_rate": 0.0013440000000000001, + "loss": 5.4411, + "step": 448 + }, + { + "epoch": 0.00449, + "grad_norm": 0.8569858280927948, + "learning_rate": 0.001347, + "loss": 5.4491, + "step": 449 + }, + { + "epoch": 0.0045, + "grad_norm": 0.766625402040655, + "learning_rate": 0.00135, + "loss": 5.4461, + "step": 450 + }, + { + "epoch": 0.00451, + "grad_norm": 0.8992860281272513, + "learning_rate": 0.001353, + "loss": 5.4411, + "step": 451 + }, + { + "epoch": 0.00452, + "grad_norm": 0.9587820940422671, + "learning_rate": 0.001356, + "loss": 5.4273, + "step": 452 + }, + { + "epoch": 0.00453, + "grad_norm": 0.9706881368414635, + "learning_rate": 0.001359, + "loss": 5.4204, + "step": 453 + }, + { + "epoch": 0.00454, + "grad_norm": 1.1942708782025127, + "learning_rate": 0.0013620000000000001, + "loss": 5.4481, + "step": 454 + }, + { + "epoch": 0.00455, + "grad_norm": 1.0651260215062406, + "learning_rate": 0.0013650000000000001, + "loss": 5.4483, + "step": 455 + }, + { + "epoch": 0.00456, + "grad_norm": 1.0894309440916736, + "learning_rate": 0.001368, + "loss": 5.4405, + "step": 456 + }, + { + "epoch": 0.00457, + "grad_norm": 1.3632436316792311, + "learning_rate": 0.001371, + "loss": 5.442, + "step": 457 + }, + { + "epoch": 0.00458, + "grad_norm": 0.9802910013598249, + "learning_rate": 0.001374, + "loss": 5.4423, + "step": 458 + }, + { + "epoch": 0.00459, + "grad_norm": 0.9681577835390196, + "learning_rate": 0.0013770000000000002, + "loss": 5.4314, + "step": 459 + }, + { + "epoch": 0.0046, + "grad_norm": 0.955881976726698, + "learning_rate": 0.0013800000000000002, + "loss": 5.4146, + "step": 460 + }, + { + "epoch": 0.00461, + "grad_norm": 1.072138092727722, + "learning_rate": 0.0013830000000000001, + "loss": 5.4186, + "step": 461 + }, + { + "epoch": 0.00462, + "grad_norm": 0.8672451154734739, + "learning_rate": 0.001386, + "loss": 5.4193, + "step": 462 + }, + { + "epoch": 0.00463, + "grad_norm": 0.9908998654710571, + "learning_rate": 0.001389, + "loss": 5.4011, + "step": 463 + }, + { + "epoch": 0.00464, + "grad_norm": 1.1599842863140275, + "learning_rate": 0.001392, + "loss": 5.4321, + "step": 464 + }, + { + "epoch": 0.00465, + "grad_norm": 0.698212811337685, + "learning_rate": 0.0013950000000000002, + "loss": 5.3861, + "step": 465 + }, + { + "epoch": 0.00466, + "grad_norm": 0.6209828526031017, + "learning_rate": 0.0013980000000000002, + "loss": 5.4004, + "step": 466 + }, + { + "epoch": 0.00467, + "grad_norm": 0.626908730655191, + "learning_rate": 0.0014010000000000001, + "loss": 5.3894, + "step": 467 + }, + { + "epoch": 0.00468, + "grad_norm": 0.6046317726003113, + "learning_rate": 0.001404, + "loss": 5.3787, + "step": 468 + }, + { + "epoch": 0.00469, + "grad_norm": 0.5283854127913247, + "learning_rate": 0.001407, + "loss": 5.3767, + "step": 469 + }, + { + "epoch": 0.0047, + "grad_norm": 0.5494342704278756, + "learning_rate": 0.00141, + "loss": 5.3859, + "step": 470 + }, + { + "epoch": 0.00471, + "grad_norm": 0.6548172150181142, + "learning_rate": 0.001413, + "loss": 5.3653, + "step": 471 + }, + { + "epoch": 0.00472, + "grad_norm": 0.8352552132751145, + "learning_rate": 0.001416, + "loss": 5.3799, + "step": 472 + }, + { + "epoch": 0.00473, + "grad_norm": 1.04775859687285, + "learning_rate": 0.001419, + "loss": 5.3985, + "step": 473 + }, + { + "epoch": 0.00474, + "grad_norm": 0.9808295146566943, + "learning_rate": 0.0014219999999999999, + "loss": 5.3787, + "step": 474 + }, + { + "epoch": 0.00475, + "grad_norm": 1.0065096729700158, + "learning_rate": 0.001425, + "loss": 5.3725, + "step": 475 + }, + { + "epoch": 0.00476, + "grad_norm": 1.0604383614131039, + "learning_rate": 0.001428, + "loss": 5.4172, + "step": 476 + }, + { + "epoch": 0.00477, + "grad_norm": 0.5551620314639825, + "learning_rate": 0.001431, + "loss": 5.3484, + "step": 477 + }, + { + "epoch": 0.00478, + "grad_norm": 0.6446758952809472, + "learning_rate": 0.001434, + "loss": 5.3819, + "step": 478 + }, + { + "epoch": 0.00479, + "grad_norm": 0.6811842486215005, + "learning_rate": 0.001437, + "loss": 5.3551, + "step": 479 + }, + { + "epoch": 0.0048, + "grad_norm": 0.6985415352827566, + "learning_rate": 0.0014399999999999999, + "loss": 5.3502, + "step": 480 + }, + { + "epoch": 0.00481, + "grad_norm": 0.7430626707063542, + "learning_rate": 0.001443, + "loss": 5.3537, + "step": 481 + }, + { + "epoch": 0.00482, + "grad_norm": 0.7883555304691023, + "learning_rate": 0.001446, + "loss": 5.3281, + "step": 482 + }, + { + "epoch": 0.00483, + "grad_norm": 0.9604272864715826, + "learning_rate": 0.001449, + "loss": 5.3455, + "step": 483 + }, + { + "epoch": 0.00484, + "grad_norm": 0.9435245256880889, + "learning_rate": 0.001452, + "loss": 5.3494, + "step": 484 + }, + { + "epoch": 0.00485, + "grad_norm": 0.8042883678590446, + "learning_rate": 0.001455, + "loss": 5.3511, + "step": 485 + }, + { + "epoch": 0.00486, + "grad_norm": 0.7958734942371152, + "learning_rate": 0.001458, + "loss": 5.3247, + "step": 486 + }, + { + "epoch": 0.00487, + "grad_norm": 0.7735191187819994, + "learning_rate": 0.001461, + "loss": 5.3397, + "step": 487 + }, + { + "epoch": 0.00488, + "grad_norm": 1.012081783792506, + "learning_rate": 0.001464, + "loss": 5.3265, + "step": 488 + }, + { + "epoch": 0.00489, + "grad_norm": 0.8905045147335564, + "learning_rate": 0.001467, + "loss": 5.3131, + "step": 489 + }, + { + "epoch": 0.0049, + "grad_norm": 0.8898357576674187, + "learning_rate": 0.00147, + "loss": 5.3259, + "step": 490 + }, + { + "epoch": 0.00491, + "grad_norm": 0.9415043482581119, + "learning_rate": 0.001473, + "loss": 5.3269, + "step": 491 + }, + { + "epoch": 0.00492, + "grad_norm": 0.8320755719970162, + "learning_rate": 0.001476, + "loss": 5.3196, + "step": 492 + }, + { + "epoch": 0.00493, + "grad_norm": 0.8445852574519659, + "learning_rate": 0.001479, + "loss": 5.3145, + "step": 493 + }, + { + "epoch": 0.00494, + "grad_norm": 1.055556747775556, + "learning_rate": 0.001482, + "loss": 5.3274, + "step": 494 + }, + { + "epoch": 0.00495, + "grad_norm": 0.8469107680360277, + "learning_rate": 0.001485, + "loss": 5.3268, + "step": 495 + }, + { + "epoch": 0.00496, + "grad_norm": 0.7635320378283988, + "learning_rate": 0.001488, + "loss": 5.3083, + "step": 496 + }, + { + "epoch": 0.00497, + "grad_norm": 0.8508130432624411, + "learning_rate": 0.001491, + "loss": 5.2927, + "step": 497 + }, + { + "epoch": 0.00498, + "grad_norm": 0.6847944056515368, + "learning_rate": 0.001494, + "loss": 5.3134, + "step": 498 + }, + { + "epoch": 0.00499, + "grad_norm": 0.73441383592788, + "learning_rate": 0.001497, + "loss": 5.2894, + "step": 499 + }, + { + "epoch": 0.005, + "grad_norm": 0.6599139353534871, + "learning_rate": 0.0015, + "loss": 5.2898, + "step": 500 + }, + { + "epoch": 0.00501, + "grad_norm": 0.5852764030687648, + "learning_rate": 0.001503, + "loss": 5.2881, + "step": 501 + }, + { + "epoch": 0.00502, + "grad_norm": 0.6853021422001514, + "learning_rate": 0.001506, + "loss": 5.2946, + "step": 502 + }, + { + "epoch": 0.00503, + "grad_norm": 0.720141729806561, + "learning_rate": 0.0015090000000000001, + "loss": 5.2988, + "step": 503 + }, + { + "epoch": 0.00504, + "grad_norm": 0.7571759373525393, + "learning_rate": 0.001512, + "loss": 5.2829, + "step": 504 + }, + { + "epoch": 0.00505, + "grad_norm": 0.7892633543044816, + "learning_rate": 0.001515, + "loss": 5.2795, + "step": 505 + }, + { + "epoch": 0.00506, + "grad_norm": 0.7628738296895545, + "learning_rate": 0.001518, + "loss": 5.2778, + "step": 506 + }, + { + "epoch": 0.00507, + "grad_norm": 0.7334519380566521, + "learning_rate": 0.001521, + "loss": 5.2722, + "step": 507 + }, + { + "epoch": 0.00508, + "grad_norm": 0.7350882061617702, + "learning_rate": 0.001524, + "loss": 5.2693, + "step": 508 + }, + { + "epoch": 0.00509, + "grad_norm": 0.7407389039658853, + "learning_rate": 0.0015270000000000001, + "loss": 5.2562, + "step": 509 + }, + { + "epoch": 0.0051, + "grad_norm": 0.7104400166661097, + "learning_rate": 0.0015300000000000001, + "loss": 5.2696, + "step": 510 + }, + { + "epoch": 0.00511, + "grad_norm": 0.8484400275698938, + "learning_rate": 0.001533, + "loss": 5.2541, + "step": 511 + }, + { + "epoch": 0.00512, + "grad_norm": 0.7909533536283434, + "learning_rate": 0.001536, + "loss": 5.2475, + "step": 512 + }, + { + "epoch": 0.00513, + "grad_norm": 0.9481517521047255, + "learning_rate": 0.001539, + "loss": 5.2684, + "step": 513 + }, + { + "epoch": 0.00514, + "grad_norm": 1.405665610624422, + "learning_rate": 0.001542, + "loss": 5.2742, + "step": 514 + }, + { + "epoch": 0.00515, + "grad_norm": 0.9237050984729354, + "learning_rate": 0.0015450000000000001, + "loss": 5.2636, + "step": 515 + }, + { + "epoch": 0.00516, + "grad_norm": 1.1634719669204763, + "learning_rate": 0.0015480000000000001, + "loss": 5.2755, + "step": 516 + }, + { + "epoch": 0.00517, + "grad_norm": 0.8585760471561034, + "learning_rate": 0.001551, + "loss": 5.2577, + "step": 517 + }, + { + "epoch": 0.00518, + "grad_norm": 0.728905665920263, + "learning_rate": 0.001554, + "loss": 5.2422, + "step": 518 + }, + { + "epoch": 0.00519, + "grad_norm": 0.841003643450708, + "learning_rate": 0.001557, + "loss": 5.2472, + "step": 519 + }, + { + "epoch": 0.0052, + "grad_norm": 0.8542329179478354, + "learning_rate": 0.0015600000000000002, + "loss": 5.2573, + "step": 520 + }, + { + "epoch": 0.00521, + "grad_norm": 0.9136344705581151, + "learning_rate": 0.0015630000000000002, + "loss": 5.2327, + "step": 521 + }, + { + "epoch": 0.00522, + "grad_norm": 1.1142541660381047, + "learning_rate": 0.0015660000000000001, + "loss": 5.2435, + "step": 522 + }, + { + "epoch": 0.00523, + "grad_norm": 1.2424782927197504, + "learning_rate": 0.001569, + "loss": 5.2642, + "step": 523 + }, + { + "epoch": 0.00524, + "grad_norm": 0.9564855927553343, + "learning_rate": 0.001572, + "loss": 5.264, + "step": 524 + }, + { + "epoch": 0.00525, + "grad_norm": 0.8662515649518988, + "learning_rate": 0.001575, + "loss": 5.2537, + "step": 525 + }, + { + "epoch": 0.00526, + "grad_norm": 0.72079723918141, + "learning_rate": 0.0015780000000000002, + "loss": 5.2224, + "step": 526 + }, + { + "epoch": 0.00527, + "grad_norm": 0.6822789926865414, + "learning_rate": 0.0015810000000000002, + "loss": 5.2353, + "step": 527 + }, + { + "epoch": 0.00528, + "grad_norm": 0.6986034229049856, + "learning_rate": 0.0015840000000000001, + "loss": 5.2128, + "step": 528 + }, + { + "epoch": 0.00529, + "grad_norm": 0.6935526974121139, + "learning_rate": 0.001587, + "loss": 5.2236, + "step": 529 + }, + { + "epoch": 0.0053, + "grad_norm": 0.6635846542220551, + "learning_rate": 0.00159, + "loss": 5.2302, + "step": 530 + }, + { + "epoch": 0.00531, + "grad_norm": 0.7197123783695031, + "learning_rate": 0.001593, + "loss": 5.2167, + "step": 531 + }, + { + "epoch": 0.00532, + "grad_norm": 0.6862921713046577, + "learning_rate": 0.0015960000000000002, + "loss": 5.2205, + "step": 532 + }, + { + "epoch": 0.00533, + "grad_norm": 0.672254585940336, + "learning_rate": 0.0015990000000000002, + "loss": 5.2165, + "step": 533 + }, + { + "epoch": 0.00534, + "grad_norm": 0.6180247915886188, + "learning_rate": 0.0016020000000000001, + "loss": 5.2018, + "step": 534 + }, + { + "epoch": 0.00535, + "grad_norm": 0.7076887136252737, + "learning_rate": 0.001605, + "loss": 5.2099, + "step": 535 + }, + { + "epoch": 0.00536, + "grad_norm": 0.8627381010586813, + "learning_rate": 0.001608, + "loss": 5.2158, + "step": 536 + }, + { + "epoch": 0.00537, + "grad_norm": 0.9890089503230703, + "learning_rate": 0.0016110000000000002, + "loss": 5.2125, + "step": 537 + }, + { + "epoch": 0.00538, + "grad_norm": 1.0890684861329838, + "learning_rate": 0.0016140000000000002, + "loss": 5.1997, + "step": 538 + }, + { + "epoch": 0.00539, + "grad_norm": 0.7898695514456295, + "learning_rate": 0.0016170000000000002, + "loss": 5.1885, + "step": 539 + }, + { + "epoch": 0.0054, + "grad_norm": 0.7877858658336557, + "learning_rate": 0.0016200000000000001, + "loss": 5.1979, + "step": 540 + }, + { + "epoch": 0.00541, + "grad_norm": 0.8613625325852945, + "learning_rate": 0.001623, + "loss": 5.1884, + "step": 541 + }, + { + "epoch": 0.00542, + "grad_norm": 0.9310959295325639, + "learning_rate": 0.001626, + "loss": 5.2135, + "step": 542 + }, + { + "epoch": 0.00543, + "grad_norm": 1.0650490028229627, + "learning_rate": 0.0016290000000000002, + "loss": 5.1821, + "step": 543 + }, + { + "epoch": 0.00544, + "grad_norm": 1.058155288535174, + "learning_rate": 0.0016320000000000002, + "loss": 5.2106, + "step": 544 + }, + { + "epoch": 0.00545, + "grad_norm": 0.7888748411290551, + "learning_rate": 0.0016350000000000002, + "loss": 5.2093, + "step": 545 + }, + { + "epoch": 0.00546, + "grad_norm": 0.8327623405474364, + "learning_rate": 0.0016380000000000001, + "loss": 5.1821, + "step": 546 + }, + { + "epoch": 0.00547, + "grad_norm": 0.9359889846839972, + "learning_rate": 0.001641, + "loss": 5.1867, + "step": 547 + }, + { + "epoch": 0.00548, + "grad_norm": 0.8250520922796094, + "learning_rate": 0.001644, + "loss": 5.1781, + "step": 548 + }, + { + "epoch": 0.00549, + "grad_norm": 0.8346002681162655, + "learning_rate": 0.0016470000000000002, + "loss": 5.1786, + "step": 549 + }, + { + "epoch": 0.0055, + "grad_norm": 0.8307775357404767, + "learning_rate": 0.0016500000000000002, + "loss": 5.1901, + "step": 550 + }, + { + "epoch": 0.00551, + "grad_norm": 0.8868587049067305, + "learning_rate": 0.0016530000000000002, + "loss": 5.1882, + "step": 551 + }, + { + "epoch": 0.00552, + "grad_norm": 1.0021721179116465, + "learning_rate": 0.0016560000000000001, + "loss": 5.1887, + "step": 552 + }, + { + "epoch": 0.00553, + "grad_norm": 1.0435640829012027, + "learning_rate": 0.001659, + "loss": 5.1828, + "step": 553 + }, + { + "epoch": 0.00554, + "grad_norm": 1.091952614784726, + "learning_rate": 0.0016620000000000003, + "loss": 5.1843, + "step": 554 + }, + { + "epoch": 0.00555, + "grad_norm": 0.7530351004511261, + "learning_rate": 0.0016650000000000002, + "loss": 5.1804, + "step": 555 + }, + { + "epoch": 0.00556, + "grad_norm": 0.7840106393058217, + "learning_rate": 0.0016680000000000002, + "loss": 5.1664, + "step": 556 + }, + { + "epoch": 0.00557, + "grad_norm": 0.6999816254444311, + "learning_rate": 0.0016710000000000002, + "loss": 5.1437, + "step": 557 + }, + { + "epoch": 0.00558, + "grad_norm": 0.7845980119871422, + "learning_rate": 0.0016740000000000001, + "loss": 5.1792, + "step": 558 + }, + { + "epoch": 0.00559, + "grad_norm": 1.002237738003299, + "learning_rate": 0.001677, + "loss": 5.165, + "step": 559 + }, + { + "epoch": 0.0056, + "grad_norm": 1.1590360602458978, + "learning_rate": 0.0016800000000000003, + "loss": 5.15, + "step": 560 + }, + { + "epoch": 0.00561, + "grad_norm": 0.9693350319936842, + "learning_rate": 0.0016830000000000003, + "loss": 5.1673, + "step": 561 + }, + { + "epoch": 0.00562, + "grad_norm": 0.9337806686381701, + "learning_rate": 0.0016860000000000002, + "loss": 5.1656, + "step": 562 + }, + { + "epoch": 0.00563, + "grad_norm": 0.7021371152666548, + "learning_rate": 0.001689, + "loss": 5.1554, + "step": 563 + }, + { + "epoch": 0.00564, + "grad_norm": 0.761762323285238, + "learning_rate": 0.001692, + "loss": 5.1364, + "step": 564 + }, + { + "epoch": 0.00565, + "grad_norm": 0.9962272803832337, + "learning_rate": 0.001695, + "loss": 5.1465, + "step": 565 + }, + { + "epoch": 0.00566, + "grad_norm": 0.9270663941212276, + "learning_rate": 0.0016979999999999999, + "loss": 5.1454, + "step": 566 + }, + { + "epoch": 0.00567, + "grad_norm": 1.0296751705367089, + "learning_rate": 0.0017009999999999998, + "loss": 5.1403, + "step": 567 + }, + { + "epoch": 0.00568, + "grad_norm": 1.0921308967550072, + "learning_rate": 0.0017039999999999998, + "loss": 5.1647, + "step": 568 + }, + { + "epoch": 0.00569, + "grad_norm": 0.9680118526613578, + "learning_rate": 0.001707, + "loss": 5.1376, + "step": 569 + }, + { + "epoch": 0.0057, + "grad_norm": 1.1373894554942883, + "learning_rate": 0.00171, + "loss": 5.1667, + "step": 570 + }, + { + "epoch": 0.00571, + "grad_norm": 0.9491793536820188, + "learning_rate": 0.001713, + "loss": 5.1618, + "step": 571 + }, + { + "epoch": 0.00572, + "grad_norm": 1.0938883440367575, + "learning_rate": 0.0017159999999999999, + "loss": 5.1499, + "step": 572 + }, + { + "epoch": 0.00573, + "grad_norm": 0.8953824806877001, + "learning_rate": 0.0017189999999999998, + "loss": 5.1459, + "step": 573 + }, + { + "epoch": 0.00574, + "grad_norm": 0.8777825611555061, + "learning_rate": 0.001722, + "loss": 5.1467, + "step": 574 + }, + { + "epoch": 0.00575, + "grad_norm": 0.9427806830230203, + "learning_rate": 0.001725, + "loss": 5.132, + "step": 575 + }, + { + "epoch": 0.00576, + "grad_norm": 0.8846520959631657, + "learning_rate": 0.001728, + "loss": 5.1269, + "step": 576 + }, + { + "epoch": 0.00577, + "grad_norm": 0.6344552018416748, + "learning_rate": 0.001731, + "loss": 5.1395, + "step": 577 + }, + { + "epoch": 0.00578, + "grad_norm": 0.6771922101340231, + "learning_rate": 0.0017339999999999999, + "loss": 5.1199, + "step": 578 + }, + { + "epoch": 0.00579, + "grad_norm": 0.5381676919488529, + "learning_rate": 0.0017369999999999998, + "loss": 5.1001, + "step": 579 + }, + { + "epoch": 0.0058, + "grad_norm": 0.5416857271346956, + "learning_rate": 0.00174, + "loss": 5.1179, + "step": 580 + }, + { + "epoch": 0.00581, + "grad_norm": 0.4886409554358658, + "learning_rate": 0.001743, + "loss": 5.1131, + "step": 581 + }, + { + "epoch": 0.00582, + "grad_norm": 0.3953796837190132, + "learning_rate": 0.001746, + "loss": 5.0842, + "step": 582 + }, + { + "epoch": 0.00583, + "grad_norm": 0.3816231764982968, + "learning_rate": 0.001749, + "loss": 5.1076, + "step": 583 + }, + { + "epoch": 0.00584, + "grad_norm": 0.39062356918599367, + "learning_rate": 0.0017519999999999999, + "loss": 5.1003, + "step": 584 + }, + { + "epoch": 0.00585, + "grad_norm": 0.3915858898605007, + "learning_rate": 0.0017549999999999998, + "loss": 5.0837, + "step": 585 + }, + { + "epoch": 0.00586, + "grad_norm": 0.4063811781565053, + "learning_rate": 0.001758, + "loss": 5.0866, + "step": 586 + }, + { + "epoch": 0.00587, + "grad_norm": 0.4274699383917251, + "learning_rate": 0.001761, + "loss": 5.0709, + "step": 587 + }, + { + "epoch": 0.00588, + "grad_norm": 0.5008029306674315, + "learning_rate": 0.001764, + "loss": 5.0767, + "step": 588 + }, + { + "epoch": 0.00589, + "grad_norm": 0.7229655060897183, + "learning_rate": 0.001767, + "loss": 5.084, + "step": 589 + }, + { + "epoch": 0.0059, + "grad_norm": 0.9281114393382421, + "learning_rate": 0.0017699999999999999, + "loss": 5.0779, + "step": 590 + }, + { + "epoch": 0.00591, + "grad_norm": 0.9689787780419555, + "learning_rate": 0.001773, + "loss": 5.0756, + "step": 591 + }, + { + "epoch": 0.00592, + "grad_norm": 0.8435580831880815, + "learning_rate": 0.001776, + "loss": 5.0975, + "step": 592 + }, + { + "epoch": 0.00593, + "grad_norm": 0.9986679613531879, + "learning_rate": 0.001779, + "loss": 5.098, + "step": 593 + }, + { + "epoch": 0.00594, + "grad_norm": 1.0996840470415932, + "learning_rate": 0.001782, + "loss": 5.1014, + "step": 594 + }, + { + "epoch": 0.00595, + "grad_norm": 0.9601369241773534, + "learning_rate": 0.001785, + "loss": 5.0957, + "step": 595 + }, + { + "epoch": 0.00596, + "grad_norm": 0.8606352953891906, + "learning_rate": 0.0017879999999999999, + "loss": 5.0932, + "step": 596 + }, + { + "epoch": 0.00597, + "grad_norm": 0.8716680293105395, + "learning_rate": 0.001791, + "loss": 5.066, + "step": 597 + }, + { + "epoch": 0.00598, + "grad_norm": 0.9954805607751368, + "learning_rate": 0.001794, + "loss": 5.0837, + "step": 598 + }, + { + "epoch": 0.00599, + "grad_norm": 1.0139742121604893, + "learning_rate": 0.001797, + "loss": 5.0892, + "step": 599 + }, + { + "epoch": 0.006, + "grad_norm": 0.8994983668637855, + "learning_rate": 0.0018, + "loss": 5.083, + "step": 600 + }, + { + "epoch": 0.00601, + "grad_norm": 1.012817382509115, + "learning_rate": 0.001803, + "loss": 5.0872, + "step": 601 + }, + { + "epoch": 0.00602, + "grad_norm": 0.829792766990954, + "learning_rate": 0.0018059999999999999, + "loss": 5.0655, + "step": 602 + }, + { + "epoch": 0.00603, + "grad_norm": 0.8978641621726422, + "learning_rate": 0.001809, + "loss": 5.0896, + "step": 603 + }, + { + "epoch": 0.00604, + "grad_norm": 1.006526980400353, + "learning_rate": 0.001812, + "loss": 5.0733, + "step": 604 + }, + { + "epoch": 0.00605, + "grad_norm": 0.9664384705752951, + "learning_rate": 0.001815, + "loss": 5.0828, + "step": 605 + }, + { + "epoch": 0.00606, + "grad_norm": 1.0962791606856168, + "learning_rate": 0.001818, + "loss": 5.0888, + "step": 606 + }, + { + "epoch": 0.00607, + "grad_norm": 0.9313257227556363, + "learning_rate": 0.001821, + "loss": 5.0876, + "step": 607 + }, + { + "epoch": 0.00608, + "grad_norm": 0.9207616431206566, + "learning_rate": 0.001824, + "loss": 5.068, + "step": 608 + }, + { + "epoch": 0.00609, + "grad_norm": 0.9552257534317795, + "learning_rate": 0.001827, + "loss": 5.0751, + "step": 609 + }, + { + "epoch": 0.0061, + "grad_norm": 1.1338881088580717, + "learning_rate": 0.00183, + "loss": 5.062, + "step": 610 + }, + { + "epoch": 0.00611, + "grad_norm": 0.907481169345242, + "learning_rate": 0.001833, + "loss": 5.0576, + "step": 611 + }, + { + "epoch": 0.00612, + "grad_norm": 1.04757168088542, + "learning_rate": 0.001836, + "loss": 5.0686, + "step": 612 + }, + { + "epoch": 0.00613, + "grad_norm": 0.9021638869008188, + "learning_rate": 0.001839, + "loss": 5.0559, + "step": 613 + }, + { + "epoch": 0.00614, + "grad_norm": 0.7877209937196055, + "learning_rate": 0.001842, + "loss": 5.0768, + "step": 614 + }, + { + "epoch": 0.00615, + "grad_norm": 0.7456491069500266, + "learning_rate": 0.001845, + "loss": 5.0572, + "step": 615 + }, + { + "epoch": 0.00616, + "grad_norm": 0.7643816652567319, + "learning_rate": 0.001848, + "loss": 5.043, + "step": 616 + }, + { + "epoch": 0.00617, + "grad_norm": 0.8920948602141958, + "learning_rate": 0.001851, + "loss": 5.0555, + "step": 617 + }, + { + "epoch": 0.00618, + "grad_norm": 1.0257178323795717, + "learning_rate": 0.001854, + "loss": 5.0507, + "step": 618 + }, + { + "epoch": 0.00619, + "grad_norm": 0.7092119630036949, + "learning_rate": 0.001857, + "loss": 5.0429, + "step": 619 + }, + { + "epoch": 0.0062, + "grad_norm": 0.6092469798525071, + "learning_rate": 0.00186, + "loss": 5.0262, + "step": 620 + }, + { + "epoch": 0.00621, + "grad_norm": 0.7552706683211996, + "learning_rate": 0.001863, + "loss": 5.032, + "step": 621 + }, + { + "epoch": 0.00622, + "grad_norm": 0.6988650940233329, + "learning_rate": 0.001866, + "loss": 5.0245, + "step": 622 + }, + { + "epoch": 0.00623, + "grad_norm": 0.6062593525325608, + "learning_rate": 0.001869, + "loss": 5.0091, + "step": 623 + }, + { + "epoch": 0.00624, + "grad_norm": 0.555479795278756, + "learning_rate": 0.001872, + "loss": 5.0161, + "step": 624 + }, + { + "epoch": 0.00625, + "grad_norm": 0.6339519418385815, + "learning_rate": 0.001875, + "loss": 5.0276, + "step": 625 + }, + { + "epoch": 0.00626, + "grad_norm": 0.7820017260467441, + "learning_rate": 0.0018780000000000001, + "loss": 5.0084, + "step": 626 + }, + { + "epoch": 0.00627, + "grad_norm": 0.8263278266088275, + "learning_rate": 0.001881, + "loss": 5.0003, + "step": 627 + }, + { + "epoch": 0.00628, + "grad_norm": 0.5717634931424201, + "learning_rate": 0.001884, + "loss": 5.0204, + "step": 628 + }, + { + "epoch": 0.00629, + "grad_norm": 0.5505525631139665, + "learning_rate": 0.001887, + "loss": 4.9928, + "step": 629 + }, + { + "epoch": 0.0063, + "grad_norm": 0.47331922258372455, + "learning_rate": 0.00189, + "loss": 4.9837, + "step": 630 + }, + { + "epoch": 0.00631, + "grad_norm": 0.4173073508747504, + "learning_rate": 0.0018930000000000002, + "loss": 4.9757, + "step": 631 + }, + { + "epoch": 0.00632, + "grad_norm": 0.42808966750972455, + "learning_rate": 0.0018960000000000001, + "loss": 4.9928, + "step": 632 + }, + { + "epoch": 0.00633, + "grad_norm": 0.46750467774391197, + "learning_rate": 0.001899, + "loss": 4.9879, + "step": 633 + }, + { + "epoch": 0.00634, + "grad_norm": 0.5696215613123979, + "learning_rate": 0.001902, + "loss": 4.9763, + "step": 634 + }, + { + "epoch": 0.00635, + "grad_norm": 0.7138673203190975, + "learning_rate": 0.001905, + "loss": 4.9686, + "step": 635 + }, + { + "epoch": 0.00636, + "grad_norm": 0.8836310721952346, + "learning_rate": 0.001908, + "loss": 4.9828, + "step": 636 + }, + { + "epoch": 0.00637, + "grad_norm": 0.9755328249694639, + "learning_rate": 0.0019110000000000002, + "loss": 5.002, + "step": 637 + }, + { + "epoch": 0.00638, + "grad_norm": 1.3272264449165134, + "learning_rate": 0.0019140000000000001, + "loss": 5.0051, + "step": 638 + }, + { + "epoch": 0.00639, + "grad_norm": 0.9140856571338883, + "learning_rate": 0.001917, + "loss": 4.9726, + "step": 639 + }, + { + "epoch": 0.0064, + "grad_norm": 1.0466736994412218, + "learning_rate": 0.00192, + "loss": 4.9869, + "step": 640 + }, + { + "epoch": 0.00641, + "grad_norm": 1.1161814084223103, + "learning_rate": 0.001923, + "loss": 5.0003, + "step": 641 + }, + { + "epoch": 0.00642, + "grad_norm": 1.1352153221992676, + "learning_rate": 0.001926, + "loss": 4.992, + "step": 642 + }, + { + "epoch": 0.00643, + "grad_norm": 1.2268384891507862, + "learning_rate": 0.0019290000000000002, + "loss": 5.008, + "step": 643 + }, + { + "epoch": 0.00644, + "grad_norm": 0.9564122465750423, + "learning_rate": 0.0019320000000000001, + "loss": 4.9857, + "step": 644 + }, + { + "epoch": 0.00645, + "grad_norm": 0.9066234784688915, + "learning_rate": 0.001935, + "loss": 4.9889, + "step": 645 + }, + { + "epoch": 0.00646, + "grad_norm": 0.894776309426942, + "learning_rate": 0.001938, + "loss": 4.995, + "step": 646 + }, + { + "epoch": 0.00647, + "grad_norm": 1.036514260058091, + "learning_rate": 0.001941, + "loss": 5.0081, + "step": 647 + }, + { + "epoch": 0.00648, + "grad_norm": 1.0669688261896326, + "learning_rate": 0.0019440000000000002, + "loss": 4.9991, + "step": 648 + }, + { + "epoch": 0.00649, + "grad_norm": 1.0027602192299327, + "learning_rate": 0.0019470000000000002, + "loss": 4.9834, + "step": 649 + }, + { + "epoch": 0.0065, + "grad_norm": 1.2453243505592877, + "learning_rate": 0.0019500000000000001, + "loss": 5.0183, + "step": 650 + }, + { + "epoch": 0.00651, + "grad_norm": 0.9810465710779535, + "learning_rate": 0.001953, + "loss": 4.9899, + "step": 651 + }, + { + "epoch": 0.00652, + "grad_norm": 1.0529607937914427, + "learning_rate": 0.0019560000000000003, + "loss": 4.9855, + "step": 652 + }, + { + "epoch": 0.00653, + "grad_norm": 0.8367844014470341, + "learning_rate": 0.0019590000000000002, + "loss": 4.9719, + "step": 653 + }, + { + "epoch": 0.00654, + "grad_norm": 0.9556352679930769, + "learning_rate": 0.001962, + "loss": 4.96, + "step": 654 + }, + { + "epoch": 0.00655, + "grad_norm": 1.0712859822115042, + "learning_rate": 0.001965, + "loss": 4.9811, + "step": 655 + }, + { + "epoch": 0.00656, + "grad_norm": 0.8381525542412449, + "learning_rate": 0.001968, + "loss": 4.9628, + "step": 656 + }, + { + "epoch": 0.00657, + "grad_norm": 0.853522104902103, + "learning_rate": 0.001971, + "loss": 4.9486, + "step": 657 + }, + { + "epoch": 0.00658, + "grad_norm": 0.9276507218825019, + "learning_rate": 0.001974, + "loss": 4.9746, + "step": 658 + }, + { + "epoch": 0.00659, + "grad_norm": 1.0041122671115763, + "learning_rate": 0.001977, + "loss": 4.9552, + "step": 659 + }, + { + "epoch": 0.0066, + "grad_norm": 0.925609681342879, + "learning_rate": 0.00198, + "loss": 4.9516, + "step": 660 + }, + { + "epoch": 0.00661, + "grad_norm": 0.9340897694356347, + "learning_rate": 0.001983, + "loss": 4.9301, + "step": 661 + }, + { + "epoch": 0.00662, + "grad_norm": 1.1039924911609775, + "learning_rate": 0.0019860000000000004, + "loss": 4.9353, + "step": 662 + }, + { + "epoch": 0.00663, + "grad_norm": 0.842271096957419, + "learning_rate": 0.0019890000000000003, + "loss": 4.942, + "step": 663 + }, + { + "epoch": 0.00664, + "grad_norm": 0.9255021185692508, + "learning_rate": 0.0019920000000000003, + "loss": 4.9294, + "step": 664 + }, + { + "epoch": 0.00665, + "grad_norm": 0.9624840652606003, + "learning_rate": 0.0019950000000000002, + "loss": 4.9033, + "step": 665 + }, + { + "epoch": 0.00666, + "grad_norm": 0.8706912476713129, + "learning_rate": 0.001998, + "loss": 4.9275, + "step": 666 + }, + { + "epoch": 0.00667, + "grad_norm": 0.9355886153293348, + "learning_rate": 0.002001, + "loss": 4.9186, + "step": 667 + }, + { + "epoch": 0.00668, + "grad_norm": 0.9525697321085718, + "learning_rate": 0.002004, + "loss": 4.9422, + "step": 668 + }, + { + "epoch": 0.00669, + "grad_norm": 0.9357519892866003, + "learning_rate": 0.002007, + "loss": 4.9423, + "step": 669 + }, + { + "epoch": 0.0067, + "grad_norm": 0.8227655724451596, + "learning_rate": 0.00201, + "loss": 4.9287, + "step": 670 + }, + { + "epoch": 0.00671, + "grad_norm": 0.6363703099565691, + "learning_rate": 0.002013, + "loss": 4.8975, + "step": 671 + }, + { + "epoch": 0.00672, + "grad_norm": 0.5774852286623988, + "learning_rate": 0.002016, + "loss": 4.894, + "step": 672 + }, + { + "epoch": 0.00673, + "grad_norm": 0.5472152449551199, + "learning_rate": 0.002019, + "loss": 4.8881, + "step": 673 + }, + { + "epoch": 0.00674, + "grad_norm": 0.5267919584843693, + "learning_rate": 0.0020220000000000004, + "loss": 4.8857, + "step": 674 + }, + { + "epoch": 0.00675, + "grad_norm": 0.4823533534454447, + "learning_rate": 0.0020250000000000003, + "loss": 4.8922, + "step": 675 + }, + { + "epoch": 0.00676, + "grad_norm": 0.5135326218761617, + "learning_rate": 0.0020280000000000003, + "loss": 4.8739, + "step": 676 + }, + { + "epoch": 0.00677, + "grad_norm": 0.5629070673817519, + "learning_rate": 0.0020310000000000003, + "loss": 4.8753, + "step": 677 + }, + { + "epoch": 0.00678, + "grad_norm": 0.6636518437686225, + "learning_rate": 0.0020340000000000002, + "loss": 4.8844, + "step": 678 + }, + { + "epoch": 0.00679, + "grad_norm": 0.7939150938071041, + "learning_rate": 0.002037, + "loss": 4.8694, + "step": 679 + }, + { + "epoch": 0.0068, + "grad_norm": 0.6961050241971515, + "learning_rate": 0.00204, + "loss": 4.8681, + "step": 680 + }, + { + "epoch": 0.00681, + "grad_norm": 0.55281231514776, + "learning_rate": 0.002043, + "loss": 4.854, + "step": 681 + }, + { + "epoch": 0.00682, + "grad_norm": 0.6966375314025457, + "learning_rate": 0.002046, + "loss": 4.8606, + "step": 682 + }, + { + "epoch": 0.00683, + "grad_norm": 0.7919918537287608, + "learning_rate": 0.002049, + "loss": 4.8437, + "step": 683 + }, + { + "epoch": 0.00684, + "grad_norm": 1.014505658611748, + "learning_rate": 0.002052, + "loss": 4.868, + "step": 684 + }, + { + "epoch": 0.00685, + "grad_norm": 1.1979460789898289, + "learning_rate": 0.0020550000000000004, + "loss": 4.8824, + "step": 685 + }, + { + "epoch": 0.00686, + "grad_norm": 1.0209197434565493, + "learning_rate": 0.0020580000000000004, + "loss": 4.8555, + "step": 686 + }, + { + "epoch": 0.00687, + "grad_norm": 0.8666785819902827, + "learning_rate": 0.0020610000000000003, + "loss": 4.8457, + "step": 687 + }, + { + "epoch": 0.00688, + "grad_norm": 1.066277682185938, + "learning_rate": 0.002064, + "loss": 4.8576, + "step": 688 + }, + { + "epoch": 0.00689, + "grad_norm": 0.9564373481813865, + "learning_rate": 0.002067, + "loss": 4.8511, + "step": 689 + }, + { + "epoch": 0.0069, + "grad_norm": 1.0921039434762347, + "learning_rate": 0.00207, + "loss": 4.854, + "step": 690 + }, + { + "epoch": 0.00691, + "grad_norm": 0.9682767243250711, + "learning_rate": 0.0020729999999999998, + "loss": 4.8265, + "step": 691 + }, + { + "epoch": 0.00692, + "grad_norm": 1.0182149348151321, + "learning_rate": 0.0020759999999999997, + "loss": 4.8523, + "step": 692 + }, + { + "epoch": 0.00693, + "grad_norm": 1.1559582715656997, + "learning_rate": 0.0020789999999999997, + "loss": 4.8626, + "step": 693 + }, + { + "epoch": 0.00694, + "grad_norm": 0.8731253854999043, + "learning_rate": 0.002082, + "loss": 4.8314, + "step": 694 + }, + { + "epoch": 0.00695, + "grad_norm": 1.02924772696324, + "learning_rate": 0.002085, + "loss": 4.8388, + "step": 695 + }, + { + "epoch": 0.00696, + "grad_norm": 1.1148665982097032, + "learning_rate": 0.002088, + "loss": 4.8532, + "step": 696 + }, + { + "epoch": 0.00697, + "grad_norm": 1.1306884538795905, + "learning_rate": 0.002091, + "loss": 4.8556, + "step": 697 + }, + { + "epoch": 0.00698, + "grad_norm": 1.267565149046305, + "learning_rate": 0.002094, + "loss": 4.8474, + "step": 698 + }, + { + "epoch": 0.00699, + "grad_norm": 0.9067921371830103, + "learning_rate": 0.002097, + "loss": 4.8312, + "step": 699 + }, + { + "epoch": 0.007, + "grad_norm": 0.8797095491461238, + "learning_rate": 0.0021, + "loss": 4.8378, + "step": 700 + }, + { + "epoch": 0.00701, + "grad_norm": 0.9764684557107473, + "learning_rate": 0.002103, + "loss": 4.8231, + "step": 701 + }, + { + "epoch": 0.00702, + "grad_norm": 0.9975561372265659, + "learning_rate": 0.002106, + "loss": 4.8269, + "step": 702 + }, + { + "epoch": 0.00703, + "grad_norm": 0.8409706105195134, + "learning_rate": 0.0021089999999999998, + "loss": 4.8046, + "step": 703 + }, + { + "epoch": 0.00704, + "grad_norm": 0.8602349489288845, + "learning_rate": 0.0021119999999999997, + "loss": 4.8112, + "step": 704 + }, + { + "epoch": 0.00705, + "grad_norm": 0.810283974291816, + "learning_rate": 0.002115, + "loss": 4.8189, + "step": 705 + }, + { + "epoch": 0.00706, + "grad_norm": 0.7904979262360543, + "learning_rate": 0.002118, + "loss": 4.8127, + "step": 706 + }, + { + "epoch": 0.00707, + "grad_norm": 0.8638007527609218, + "learning_rate": 0.002121, + "loss": 4.7902, + "step": 707 + }, + { + "epoch": 0.00708, + "grad_norm": 0.9274642295068019, + "learning_rate": 0.002124, + "loss": 4.7754, + "step": 708 + }, + { + "epoch": 0.00709, + "grad_norm": 0.9265048264631317, + "learning_rate": 0.002127, + "loss": 4.8051, + "step": 709 + }, + { + "epoch": 0.0071, + "grad_norm": 1.0021235226233682, + "learning_rate": 0.00213, + "loss": 4.8021, + "step": 710 + }, + { + "epoch": 0.00711, + "grad_norm": 0.8616225643918476, + "learning_rate": 0.002133, + "loss": 4.7687, + "step": 711 + }, + { + "epoch": 0.00712, + "grad_norm": 1.1509118831082872, + "learning_rate": 0.002136, + "loss": 4.8063, + "step": 712 + }, + { + "epoch": 0.00713, + "grad_norm": 0.8979386810595619, + "learning_rate": 0.002139, + "loss": 4.8085, + "step": 713 + }, + { + "epoch": 0.00714, + "grad_norm": 0.9672478380991173, + "learning_rate": 0.002142, + "loss": 4.7999, + "step": 714 + }, + { + "epoch": 0.00715, + "grad_norm": 1.0283238664767786, + "learning_rate": 0.0021449999999999998, + "loss": 4.7746, + "step": 715 + }, + { + "epoch": 0.00716, + "grad_norm": 0.8031836486660412, + "learning_rate": 0.002148, + "loss": 4.7611, + "step": 716 + }, + { + "epoch": 0.00717, + "grad_norm": 0.8079202930068127, + "learning_rate": 0.002151, + "loss": 4.7802, + "step": 717 + }, + { + "epoch": 0.00718, + "grad_norm": 0.7625878769693025, + "learning_rate": 0.002154, + "loss": 4.7607, + "step": 718 + }, + { + "epoch": 0.00719, + "grad_norm": 0.8833410128202536, + "learning_rate": 0.002157, + "loss": 4.7767, + "step": 719 + }, + { + "epoch": 0.0072, + "grad_norm": 1.0099718860880083, + "learning_rate": 0.00216, + "loss": 4.7733, + "step": 720 + }, + { + "epoch": 0.00721, + "grad_norm": 1.0037155368349988, + "learning_rate": 0.002163, + "loss": 4.78, + "step": 721 + }, + { + "epoch": 0.00722, + "grad_norm": 0.8678480066188063, + "learning_rate": 0.002166, + "loss": 4.7474, + "step": 722 + }, + { + "epoch": 0.00723, + "grad_norm": 0.9399210002572385, + "learning_rate": 0.002169, + "loss": 4.7657, + "step": 723 + }, + { + "epoch": 0.00724, + "grad_norm": 0.8184829699796181, + "learning_rate": 0.002172, + "loss": 4.7533, + "step": 724 + }, + { + "epoch": 0.00725, + "grad_norm": 0.8676839571587074, + "learning_rate": 0.002175, + "loss": 4.7513, + "step": 725 + }, + { + "epoch": 0.00726, + "grad_norm": 0.9799992988904748, + "learning_rate": 0.002178, + "loss": 4.7626, + "step": 726 + }, + { + "epoch": 0.00727, + "grad_norm": 1.2281779129682024, + "learning_rate": 0.0021809999999999998, + "loss": 4.7581, + "step": 727 + }, + { + "epoch": 0.00728, + "grad_norm": 1.082945747060172, + "learning_rate": 0.002184, + "loss": 4.7657, + "step": 728 + }, + { + "epoch": 0.00729, + "grad_norm": 1.0915510364818644, + "learning_rate": 0.002187, + "loss": 4.7617, + "step": 729 + }, + { + "epoch": 0.0073, + "grad_norm": 1.0738468909531949, + "learning_rate": 0.00219, + "loss": 4.7676, + "step": 730 + }, + { + "epoch": 0.00731, + "grad_norm": 1.0774407965183543, + "learning_rate": 0.002193, + "loss": 4.7572, + "step": 731 + }, + { + "epoch": 0.00732, + "grad_norm": 0.9732910355796593, + "learning_rate": 0.002196, + "loss": 4.7638, + "step": 732 + }, + { + "epoch": 0.00733, + "grad_norm": 1.02567545008427, + "learning_rate": 0.002199, + "loss": 4.7467, + "step": 733 + }, + { + "epoch": 0.00734, + "grad_norm": 0.9058765241181546, + "learning_rate": 0.002202, + "loss": 4.7686, + "step": 734 + }, + { + "epoch": 0.00735, + "grad_norm": 0.9682392169542167, + "learning_rate": 0.002205, + "loss": 4.7759, + "step": 735 + }, + { + "epoch": 0.00736, + "grad_norm": 0.9938811207200824, + "learning_rate": 0.002208, + "loss": 4.7615, + "step": 736 + }, + { + "epoch": 0.00737, + "grad_norm": 1.0964297344539389, + "learning_rate": 0.002211, + "loss": 4.7347, + "step": 737 + }, + { + "epoch": 0.00738, + "grad_norm": 0.8707349769325928, + "learning_rate": 0.002214, + "loss": 4.7342, + "step": 738 + }, + { + "epoch": 0.00739, + "grad_norm": 0.7399818233744658, + "learning_rate": 0.0022170000000000002, + "loss": 4.717, + "step": 739 + }, + { + "epoch": 0.0074, + "grad_norm": 0.9377539124718652, + "learning_rate": 0.00222, + "loss": 4.7301, + "step": 740 + }, + { + "epoch": 0.00741, + "grad_norm": 0.8596400455739317, + "learning_rate": 0.002223, + "loss": 4.699, + "step": 741 + }, + { + "epoch": 0.00742, + "grad_norm": 0.6578432901740889, + "learning_rate": 0.002226, + "loss": 4.7249, + "step": 742 + }, + { + "epoch": 0.00743, + "grad_norm": 0.6175884285032084, + "learning_rate": 0.002229, + "loss": 4.6843, + "step": 743 + }, + { + "epoch": 0.00744, + "grad_norm": 0.6615203369086972, + "learning_rate": 0.002232, + "loss": 4.6918, + "step": 744 + }, + { + "epoch": 0.00745, + "grad_norm": 0.6999197355703424, + "learning_rate": 0.002235, + "loss": 4.7005, + "step": 745 + }, + { + "epoch": 0.00746, + "grad_norm": 0.7056349857734648, + "learning_rate": 0.002238, + "loss": 4.6964, + "step": 746 + }, + { + "epoch": 0.00747, + "grad_norm": 0.6678161149510893, + "learning_rate": 0.002241, + "loss": 4.6817, + "step": 747 + }, + { + "epoch": 0.00748, + "grad_norm": 0.6725287147155753, + "learning_rate": 0.002244, + "loss": 4.6915, + "step": 748 + }, + { + "epoch": 0.00749, + "grad_norm": 0.7368138262221237, + "learning_rate": 0.002247, + "loss": 4.6725, + "step": 749 + }, + { + "epoch": 0.0075, + "grad_norm": 0.7550252977049275, + "learning_rate": 0.0022500000000000003, + "loss": 4.7124, + "step": 750 + }, + { + "epoch": 0.00751, + "grad_norm": 0.6461697616177359, + "learning_rate": 0.0022530000000000002, + "loss": 4.6948, + "step": 751 + }, + { + "epoch": 0.00752, + "grad_norm": 0.6473605328229959, + "learning_rate": 0.002256, + "loss": 4.648, + "step": 752 + }, + { + "epoch": 0.00753, + "grad_norm": 0.8172272904356894, + "learning_rate": 0.002259, + "loss": 4.6929, + "step": 753 + }, + { + "epoch": 0.00754, + "grad_norm": 0.8400684755887758, + "learning_rate": 0.002262, + "loss": 4.7068, + "step": 754 + }, + { + "epoch": 0.00755, + "grad_norm": 0.7078832518794317, + "learning_rate": 0.002265, + "loss": 4.6656, + "step": 755 + }, + { + "epoch": 0.00756, + "grad_norm": 0.6076060117836831, + "learning_rate": 0.002268, + "loss": 4.6484, + "step": 756 + }, + { + "epoch": 0.00757, + "grad_norm": 0.7133093934008413, + "learning_rate": 0.002271, + "loss": 4.6658, + "step": 757 + }, + { + "epoch": 0.00758, + "grad_norm": 0.7661771348142844, + "learning_rate": 0.002274, + "loss": 4.6521, + "step": 758 + }, + { + "epoch": 0.00759, + "grad_norm": 0.9250288948777622, + "learning_rate": 0.002277, + "loss": 4.6753, + "step": 759 + }, + { + "epoch": 0.0076, + "grad_norm": 1.037174236565274, + "learning_rate": 0.00228, + "loss": 4.669, + "step": 760 + }, + { + "epoch": 0.00761, + "grad_norm": 0.9678315157211191, + "learning_rate": 0.002283, + "loss": 4.6392, + "step": 761 + }, + { + "epoch": 0.00762, + "grad_norm": 1.3728001530688312, + "learning_rate": 0.0022860000000000003, + "loss": 4.6453, + "step": 762 + }, + { + "epoch": 0.00763, + "grad_norm": 1.0284727877786697, + "learning_rate": 0.0022890000000000002, + "loss": 4.6793, + "step": 763 + }, + { + "epoch": 0.00764, + "grad_norm": 0.9914794664489192, + "learning_rate": 0.002292, + "loss": 4.6942, + "step": 764 + }, + { + "epoch": 0.00765, + "grad_norm": 0.984322504117537, + "learning_rate": 0.002295, + "loss": 4.6765, + "step": 765 + }, + { + "epoch": 0.00766, + "grad_norm": 0.9320893698991433, + "learning_rate": 0.002298, + "loss": 4.6792, + "step": 766 + }, + { + "epoch": 0.00767, + "grad_norm": 1.0651442494276249, + "learning_rate": 0.002301, + "loss": 4.6823, + "step": 767 + }, + { + "epoch": 0.00768, + "grad_norm": 0.9179111624711317, + "learning_rate": 0.002304, + "loss": 4.6817, + "step": 768 + }, + { + "epoch": 0.00769, + "grad_norm": 1.026483766110404, + "learning_rate": 0.002307, + "loss": 4.6987, + "step": 769 + }, + { + "epoch": 0.0077, + "grad_norm": 1.1653119571960542, + "learning_rate": 0.00231, + "loss": 4.6621, + "step": 770 + }, + { + "epoch": 0.00771, + "grad_norm": 0.8477238808348645, + "learning_rate": 0.002313, + "loss": 4.689, + "step": 771 + }, + { + "epoch": 0.00772, + "grad_norm": 0.6519421622488206, + "learning_rate": 0.002316, + "loss": 4.6631, + "step": 772 + }, + { + "epoch": 0.00773, + "grad_norm": 0.6177861857364649, + "learning_rate": 0.0023190000000000003, + "loss": 4.6627, + "step": 773 + }, + { + "epoch": 0.00774, + "grad_norm": 0.6901665734497584, + "learning_rate": 0.0023220000000000003, + "loss": 4.6775, + "step": 774 + }, + { + "epoch": 0.00775, + "grad_norm": 0.7356087164350898, + "learning_rate": 0.0023250000000000002, + "loss": 4.6725, + "step": 775 + }, + { + "epoch": 0.00776, + "grad_norm": 0.8693432194982287, + "learning_rate": 0.002328, + "loss": 4.6722, + "step": 776 + }, + { + "epoch": 0.00777, + "grad_norm": 1.099570573598906, + "learning_rate": 0.002331, + "loss": 4.6596, + "step": 777 + }, + { + "epoch": 0.00778, + "grad_norm": 1.0694357671416344, + "learning_rate": 0.002334, + "loss": 4.6725, + "step": 778 + }, + { + "epoch": 0.00779, + "grad_norm": 1.0144407661707453, + "learning_rate": 0.002337, + "loss": 4.6345, + "step": 779 + }, + { + "epoch": 0.0078, + "grad_norm": 0.9392788681741788, + "learning_rate": 0.00234, + "loss": 4.6579, + "step": 780 + }, + { + "epoch": 0.00781, + "grad_norm": 0.9039044158767507, + "learning_rate": 0.002343, + "loss": 4.6528, + "step": 781 + }, + { + "epoch": 0.00782, + "grad_norm": 0.9671545635863801, + "learning_rate": 0.002346, + "loss": 4.6552, + "step": 782 + }, + { + "epoch": 0.00783, + "grad_norm": 0.9789758043675277, + "learning_rate": 0.002349, + "loss": 4.6404, + "step": 783 + }, + { + "epoch": 0.00784, + "grad_norm": 0.7674708275760124, + "learning_rate": 0.002352, + "loss": 4.6449, + "step": 784 + }, + { + "epoch": 0.00785, + "grad_norm": 0.6509064102772842, + "learning_rate": 0.0023550000000000003, + "loss": 4.6391, + "step": 785 + }, + { + "epoch": 0.00786, + "grad_norm": 0.6282839162170526, + "learning_rate": 0.0023580000000000003, + "loss": 4.6356, + "step": 786 + }, + { + "epoch": 0.00787, + "grad_norm": 0.6520823986249177, + "learning_rate": 0.0023610000000000003, + "loss": 4.6384, + "step": 787 + }, + { + "epoch": 0.00788, + "grad_norm": 0.6693545084259133, + "learning_rate": 0.002364, + "loss": 4.6342, + "step": 788 + }, + { + "epoch": 0.00789, + "grad_norm": 0.7061838607079715, + "learning_rate": 0.002367, + "loss": 4.6597, + "step": 789 + }, + { + "epoch": 0.0079, + "grad_norm": 0.7001074671969121, + "learning_rate": 0.00237, + "loss": 4.6333, + "step": 790 + }, + { + "epoch": 0.00791, + "grad_norm": 0.6934841557036142, + "learning_rate": 0.002373, + "loss": 4.5873, + "step": 791 + }, + { + "epoch": 0.00792, + "grad_norm": 0.6555126375785874, + "learning_rate": 0.002376, + "loss": 4.6158, + "step": 792 + }, + { + "epoch": 0.00793, + "grad_norm": 0.6198619755446345, + "learning_rate": 0.002379, + "loss": 4.6082, + "step": 793 + }, + { + "epoch": 0.00794, + "grad_norm": 0.5540734864838481, + "learning_rate": 0.002382, + "loss": 4.615, + "step": 794 + }, + { + "epoch": 0.00795, + "grad_norm": 0.5756469670820633, + "learning_rate": 0.002385, + "loss": 4.5927, + "step": 795 + }, + { + "epoch": 0.00796, + "grad_norm": 0.5912602735893169, + "learning_rate": 0.0023880000000000004, + "loss": 4.5839, + "step": 796 + }, + { + "epoch": 0.00797, + "grad_norm": 0.6477479187436139, + "learning_rate": 0.0023910000000000003, + "loss": 4.628, + "step": 797 + }, + { + "epoch": 0.00798, + "grad_norm": 0.6568610896012951, + "learning_rate": 0.0023940000000000003, + "loss": 4.5975, + "step": 798 + }, + { + "epoch": 0.00799, + "grad_norm": 0.8964867071559416, + "learning_rate": 0.0023970000000000003, + "loss": 4.6327, + "step": 799 + }, + { + "epoch": 0.008, + "grad_norm": 1.1130272400690795, + "learning_rate": 0.0024000000000000002, + "loss": 4.6127, + "step": 800 + }, + { + "epoch": 0.00801, + "grad_norm": 0.8936330318178004, + "learning_rate": 0.002403, + "loss": 4.624, + "step": 801 + }, + { + "epoch": 0.00802, + "grad_norm": 1.279259611288336, + "learning_rate": 0.002406, + "loss": 4.6431, + "step": 802 + }, + { + "epoch": 0.00803, + "grad_norm": 0.7588568023085343, + "learning_rate": 0.002409, + "loss": 4.5967, + "step": 803 + }, + { + "epoch": 0.00804, + "grad_norm": 0.9911172738466991, + "learning_rate": 0.002412, + "loss": 4.6083, + "step": 804 + }, + { + "epoch": 0.00805, + "grad_norm": 1.0297855025384943, + "learning_rate": 0.002415, + "loss": 4.6095, + "step": 805 + }, + { + "epoch": 0.00806, + "grad_norm": 0.8656546102408385, + "learning_rate": 0.002418, + "loss": 4.6231, + "step": 806 + }, + { + "epoch": 0.00807, + "grad_norm": 0.7922998530369505, + "learning_rate": 0.0024210000000000004, + "loss": 4.6144, + "step": 807 + }, + { + "epoch": 0.00808, + "grad_norm": 0.7119238411669042, + "learning_rate": 0.0024240000000000004, + "loss": 4.5933, + "step": 808 + }, + { + "epoch": 0.00809, + "grad_norm": 0.6923611180056076, + "learning_rate": 0.0024270000000000003, + "loss": 4.5855, + "step": 809 + }, + { + "epoch": 0.0081, + "grad_norm": 0.759382946117684, + "learning_rate": 0.0024300000000000003, + "loss": 4.6149, + "step": 810 + }, + { + "epoch": 0.00811, + "grad_norm": 0.9629068291697402, + "learning_rate": 0.0024330000000000003, + "loss": 4.5955, + "step": 811 + }, + { + "epoch": 0.00812, + "grad_norm": 0.9897740547487952, + "learning_rate": 0.0024360000000000002, + "loss": 4.5793, + "step": 812 + }, + { + "epoch": 0.00813, + "grad_norm": 1.0343029805479964, + "learning_rate": 0.0024389999999999998, + "loss": 4.6192, + "step": 813 + }, + { + "epoch": 0.00814, + "grad_norm": 1.188458156645309, + "learning_rate": 0.0024419999999999997, + "loss": 4.6409, + "step": 814 + }, + { + "epoch": 0.00815, + "grad_norm": 0.960728927074141, + "learning_rate": 0.0024449999999999997, + "loss": 4.6319, + "step": 815 + }, + { + "epoch": 0.00816, + "grad_norm": 1.0099133377105225, + "learning_rate": 0.002448, + "loss": 4.6186, + "step": 816 + }, + { + "epoch": 0.00817, + "grad_norm": 0.9311887346236664, + "learning_rate": 0.002451, + "loss": 4.6246, + "step": 817 + }, + { + "epoch": 0.00818, + "grad_norm": 1.2609797944179002, + "learning_rate": 0.002454, + "loss": 4.625, + "step": 818 + }, + { + "epoch": 0.00819, + "grad_norm": 1.0447566978787928, + "learning_rate": 0.002457, + "loss": 4.6401, + "step": 819 + }, + { + "epoch": 0.0082, + "grad_norm": 1.1000322233283122, + "learning_rate": 0.00246, + "loss": 4.6239, + "step": 820 + }, + { + "epoch": 0.00821, + "grad_norm": 0.8928816997822231, + "learning_rate": 0.002463, + "loss": 4.5932, + "step": 821 + }, + { + "epoch": 0.00822, + "grad_norm": 0.9105337296182261, + "learning_rate": 0.002466, + "loss": 4.5996, + "step": 822 + }, + { + "epoch": 0.00823, + "grad_norm": 0.6537296194510775, + "learning_rate": 0.002469, + "loss": 4.5905, + "step": 823 + }, + { + "epoch": 0.00824, + "grad_norm": 0.6615102644633621, + "learning_rate": 0.002472, + "loss": 4.61, + "step": 824 + }, + { + "epoch": 0.00825, + "grad_norm": 0.7605760557018463, + "learning_rate": 0.0024749999999999998, + "loss": 4.6148, + "step": 825 + }, + { + "epoch": 0.00826, + "grad_norm": 0.9288808111624368, + "learning_rate": 0.0024779999999999997, + "loss": 4.5806, + "step": 826 + }, + { + "epoch": 0.00827, + "grad_norm": 0.8099205132023622, + "learning_rate": 0.002481, + "loss": 4.6084, + "step": 827 + }, + { + "epoch": 0.00828, + "grad_norm": 0.7444084275861881, + "learning_rate": 0.002484, + "loss": 4.5965, + "step": 828 + }, + { + "epoch": 0.00829, + "grad_norm": 0.6438605280155706, + "learning_rate": 0.002487, + "loss": 4.5891, + "step": 829 + }, + { + "epoch": 0.0083, + "grad_norm": 0.6242869251575957, + "learning_rate": 0.00249, + "loss": 4.5955, + "step": 830 + }, + { + "epoch": 0.00831, + "grad_norm": 0.6705073637500226, + "learning_rate": 0.002493, + "loss": 4.576, + "step": 831 + }, + { + "epoch": 0.00832, + "grad_norm": 0.7603459804613621, + "learning_rate": 0.002496, + "loss": 4.5759, + "step": 832 + }, + { + "epoch": 0.00833, + "grad_norm": 0.7662808437283888, + "learning_rate": 0.002499, + "loss": 4.5696, + "step": 833 + }, + { + "epoch": 0.00834, + "grad_norm": 0.8012380275176963, + "learning_rate": 0.002502, + "loss": 4.5743, + "step": 834 + }, + { + "epoch": 0.00835, + "grad_norm": 1.0075156101089233, + "learning_rate": 0.002505, + "loss": 4.5783, + "step": 835 + }, + { + "epoch": 0.00836, + "grad_norm": 1.0847073328294785, + "learning_rate": 0.002508, + "loss": 4.6155, + "step": 836 + }, + { + "epoch": 0.00837, + "grad_norm": 0.8849400794535106, + "learning_rate": 0.0025109999999999998, + "loss": 4.5603, + "step": 837 + }, + { + "epoch": 0.00838, + "grad_norm": 0.8832989831172253, + "learning_rate": 0.0025139999999999997, + "loss": 4.5816, + "step": 838 + }, + { + "epoch": 0.00839, + "grad_norm": 0.9511730781538094, + "learning_rate": 0.002517, + "loss": 4.5843, + "step": 839 + }, + { + "epoch": 0.0084, + "grad_norm": 1.037817208386755, + "learning_rate": 0.00252, + "loss": 4.5831, + "step": 840 + }, + { + "epoch": 0.00841, + "grad_norm": 0.8472466492390959, + "learning_rate": 0.002523, + "loss": 4.5624, + "step": 841 + }, + { + "epoch": 0.00842, + "grad_norm": 0.9269217528832409, + "learning_rate": 0.002526, + "loss": 4.5661, + "step": 842 + }, + { + "epoch": 0.00843, + "grad_norm": 0.8716247137682855, + "learning_rate": 0.002529, + "loss": 4.5688, + "step": 843 + }, + { + "epoch": 0.00844, + "grad_norm": 0.7416972066179122, + "learning_rate": 0.002532, + "loss": 4.585, + "step": 844 + }, + { + "epoch": 0.00845, + "grad_norm": 0.6177824387251759, + "learning_rate": 0.002535, + "loss": 4.5442, + "step": 845 + }, + { + "epoch": 0.00846, + "grad_norm": 0.5865348849804463, + "learning_rate": 0.002538, + "loss": 4.5656, + "step": 846 + }, + { + "epoch": 0.00847, + "grad_norm": 0.5324615561823309, + "learning_rate": 0.002541, + "loss": 4.5442, + "step": 847 + }, + { + "epoch": 0.00848, + "grad_norm": 0.5568583953268653, + "learning_rate": 0.002544, + "loss": 4.5407, + "step": 848 + }, + { + "epoch": 0.00849, + "grad_norm": 0.600307108588631, + "learning_rate": 0.002547, + "loss": 4.5371, + "step": 849 + }, + { + "epoch": 0.0085, + "grad_norm": 0.5343909127282478, + "learning_rate": 0.00255, + "loss": 4.5435, + "step": 850 + }, + { + "epoch": 0.00851, + "grad_norm": 0.5790732432599173, + "learning_rate": 0.002553, + "loss": 4.5588, + "step": 851 + }, + { + "epoch": 0.00852, + "grad_norm": 0.6406298666409433, + "learning_rate": 0.002556, + "loss": 4.5503, + "step": 852 + }, + { + "epoch": 0.00853, + "grad_norm": 0.6711876835719848, + "learning_rate": 0.002559, + "loss": 4.5298, + "step": 853 + }, + { + "epoch": 0.00854, + "grad_norm": 0.643558144142947, + "learning_rate": 0.002562, + "loss": 4.5219, + "step": 854 + }, + { + "epoch": 0.00855, + "grad_norm": 0.621905866188772, + "learning_rate": 0.002565, + "loss": 4.5026, + "step": 855 + }, + { + "epoch": 0.00856, + "grad_norm": 0.7503391003054042, + "learning_rate": 0.002568, + "loss": 4.5375, + "step": 856 + }, + { + "epoch": 0.00857, + "grad_norm": 0.90802719077466, + "learning_rate": 0.002571, + "loss": 4.5344, + "step": 857 + }, + { + "epoch": 0.00858, + "grad_norm": 0.9157789056738207, + "learning_rate": 0.002574, + "loss": 4.5308, + "step": 858 + }, + { + "epoch": 0.00859, + "grad_norm": 0.8455467899517649, + "learning_rate": 0.002577, + "loss": 4.5208, + "step": 859 + }, + { + "epoch": 0.0086, + "grad_norm": 0.929986585155377, + "learning_rate": 0.00258, + "loss": 4.5557, + "step": 860 + }, + { + "epoch": 0.00861, + "grad_norm": 1.062443351155347, + "learning_rate": 0.0025830000000000002, + "loss": 4.5815, + "step": 861 + }, + { + "epoch": 0.00862, + "grad_norm": 1.0476479971551458, + "learning_rate": 0.002586, + "loss": 4.5382, + "step": 862 + }, + { + "epoch": 0.00863, + "grad_norm": 0.9516272804079478, + "learning_rate": 0.002589, + "loss": 4.5686, + "step": 863 + }, + { + "epoch": 0.00864, + "grad_norm": 1.0775847362135182, + "learning_rate": 0.002592, + "loss": 4.5222, + "step": 864 + }, + { + "epoch": 0.00865, + "grad_norm": 0.902924351244226, + "learning_rate": 0.002595, + "loss": 4.5357, + "step": 865 + }, + { + "epoch": 0.00866, + "grad_norm": 0.889409265471235, + "learning_rate": 0.002598, + "loss": 4.5623, + "step": 866 + }, + { + "epoch": 0.00867, + "grad_norm": 1.1312843257696636, + "learning_rate": 0.002601, + "loss": 4.5565, + "step": 867 + }, + { + "epoch": 0.00868, + "grad_norm": 1.0195609890673947, + "learning_rate": 0.002604, + "loss": 4.5479, + "step": 868 + }, + { + "epoch": 0.00869, + "grad_norm": 0.8378311515658349, + "learning_rate": 0.002607, + "loss": 4.5492, + "step": 869 + }, + { + "epoch": 0.0087, + "grad_norm": 0.93569973985326, + "learning_rate": 0.00261, + "loss": 4.5413, + "step": 870 + }, + { + "epoch": 0.00871, + "grad_norm": 0.9947247250751194, + "learning_rate": 0.002613, + "loss": 4.5608, + "step": 871 + }, + { + "epoch": 0.00872, + "grad_norm": 0.9456334146879876, + "learning_rate": 0.002616, + "loss": 4.5489, + "step": 872 + }, + { + "epoch": 0.00873, + "grad_norm": 0.9088952463307589, + "learning_rate": 0.0026190000000000002, + "loss": 4.5587, + "step": 873 + }, + { + "epoch": 0.00874, + "grad_norm": 0.7636008668853458, + "learning_rate": 0.002622, + "loss": 4.5702, + "step": 874 + }, + { + "epoch": 0.00875, + "grad_norm": 0.8446171091890929, + "learning_rate": 0.002625, + "loss": 4.5191, + "step": 875 + }, + { + "epoch": 0.00876, + "grad_norm": 0.8731048122579586, + "learning_rate": 0.002628, + "loss": 4.551, + "step": 876 + }, + { + "epoch": 0.00877, + "grad_norm": 0.9444127454444267, + "learning_rate": 0.002631, + "loss": 4.5701, + "step": 877 + }, + { + "epoch": 0.00878, + "grad_norm": 0.9863621265162925, + "learning_rate": 0.002634, + "loss": 4.5108, + "step": 878 + }, + { + "epoch": 0.00879, + "grad_norm": 0.7746522972030878, + "learning_rate": 0.002637, + "loss": 4.5293, + "step": 879 + }, + { + "epoch": 0.0088, + "grad_norm": 0.7150869014843059, + "learning_rate": 0.00264, + "loss": 4.5368, + "step": 880 + }, + { + "epoch": 0.00881, + "grad_norm": 0.866142425879982, + "learning_rate": 0.002643, + "loss": 4.5448, + "step": 881 + }, + { + "epoch": 0.00882, + "grad_norm": 1.0565004714930601, + "learning_rate": 0.002646, + "loss": 4.5522, + "step": 882 + }, + { + "epoch": 0.00883, + "grad_norm": 0.88323512650563, + "learning_rate": 0.002649, + "loss": 4.5382, + "step": 883 + }, + { + "epoch": 0.00884, + "grad_norm": 0.8377440854137777, + "learning_rate": 0.0026520000000000003, + "loss": 4.5227, + "step": 884 + }, + { + "epoch": 0.00885, + "grad_norm": 0.8587569363150891, + "learning_rate": 0.0026550000000000002, + "loss": 4.5189, + "step": 885 + }, + { + "epoch": 0.00886, + "grad_norm": 0.9455034420832737, + "learning_rate": 0.002658, + "loss": 4.5539, + "step": 886 + }, + { + "epoch": 0.00887, + "grad_norm": 0.8258368480900744, + "learning_rate": 0.002661, + "loss": 4.536, + "step": 887 + }, + { + "epoch": 0.00888, + "grad_norm": 0.927023832062946, + "learning_rate": 0.002664, + "loss": 4.5392, + "step": 888 + }, + { + "epoch": 0.00889, + "grad_norm": 0.9905400773231482, + "learning_rate": 0.002667, + "loss": 4.5077, + "step": 889 + }, + { + "epoch": 0.0089, + "grad_norm": 1.0181625806478707, + "learning_rate": 0.00267, + "loss": 4.551, + "step": 890 + }, + { + "epoch": 0.00891, + "grad_norm": 1.0618776306697646, + "learning_rate": 0.002673, + "loss": 4.5446, + "step": 891 + }, + { + "epoch": 0.00892, + "grad_norm": 0.9464629097549706, + "learning_rate": 0.002676, + "loss": 4.5493, + "step": 892 + }, + { + "epoch": 0.00893, + "grad_norm": 1.02959831042168, + "learning_rate": 0.002679, + "loss": 4.5321, + "step": 893 + }, + { + "epoch": 0.00894, + "grad_norm": 0.8717588229222071, + "learning_rate": 0.002682, + "loss": 4.5126, + "step": 894 + }, + { + "epoch": 0.00895, + "grad_norm": 0.7766302283006379, + "learning_rate": 0.0026850000000000003, + "loss": 4.5191, + "step": 895 + }, + { + "epoch": 0.00896, + "grad_norm": 0.7210349653989065, + "learning_rate": 0.0026880000000000003, + "loss": 4.5061, + "step": 896 + }, + { + "epoch": 0.00897, + "grad_norm": 0.7840425119747009, + "learning_rate": 0.0026910000000000002, + "loss": 4.5249, + "step": 897 + }, + { + "epoch": 0.00898, + "grad_norm": 0.7907811575022647, + "learning_rate": 0.002694, + "loss": 4.518, + "step": 898 + }, + { + "epoch": 0.00899, + "grad_norm": 0.7866169386193732, + "learning_rate": 0.002697, + "loss": 4.5195, + "step": 899 + }, + { + "epoch": 0.009, + "grad_norm": 0.7303478296484299, + "learning_rate": 0.0027, + "loss": 4.5449, + "step": 900 + }, + { + "epoch": 0.00901, + "grad_norm": 0.6444049592088867, + "learning_rate": 0.002703, + "loss": 4.4943, + "step": 901 + }, + { + "epoch": 0.00902, + "grad_norm": 0.5222361304775145, + "learning_rate": 0.002706, + "loss": 4.498, + "step": 902 + }, + { + "epoch": 0.00903, + "grad_norm": 0.6152635474063327, + "learning_rate": 0.002709, + "loss": 4.4967, + "step": 903 + }, + { + "epoch": 0.00904, + "grad_norm": 0.6931791624762859, + "learning_rate": 0.002712, + "loss": 4.4818, + "step": 904 + }, + { + "epoch": 0.00905, + "grad_norm": 0.6467753244335184, + "learning_rate": 0.002715, + "loss": 4.4929, + "step": 905 + }, + { + "epoch": 0.00906, + "grad_norm": 0.6638188976320862, + "learning_rate": 0.002718, + "loss": 4.5049, + "step": 906 + }, + { + "epoch": 0.00907, + "grad_norm": 0.503631484987646, + "learning_rate": 0.0027210000000000003, + "loss": 4.5121, + "step": 907 + }, + { + "epoch": 0.00908, + "grad_norm": 0.4745734069368428, + "learning_rate": 0.0027240000000000003, + "loss": 4.481, + "step": 908 + }, + { + "epoch": 0.00909, + "grad_norm": 0.5648420216556126, + "learning_rate": 0.0027270000000000003, + "loss": 4.5013, + "step": 909 + }, + { + "epoch": 0.0091, + "grad_norm": 0.5471966544704259, + "learning_rate": 0.0027300000000000002, + "loss": 4.5027, + "step": 910 + }, + { + "epoch": 0.00911, + "grad_norm": 0.6580432739675335, + "learning_rate": 0.002733, + "loss": 4.4619, + "step": 911 + }, + { + "epoch": 0.00912, + "grad_norm": 0.652804172410743, + "learning_rate": 0.002736, + "loss": 4.4678, + "step": 912 + }, + { + "epoch": 0.00913, + "grad_norm": 0.6108886726833864, + "learning_rate": 0.002739, + "loss": 4.4601, + "step": 913 + }, + { + "epoch": 0.00914, + "grad_norm": 0.5246452629450868, + "learning_rate": 0.002742, + "loss": 4.4865, + "step": 914 + }, + { + "epoch": 0.00915, + "grad_norm": 0.6046562570851716, + "learning_rate": 0.002745, + "loss": 4.4501, + "step": 915 + }, + { + "epoch": 0.00916, + "grad_norm": 0.6470672223067736, + "learning_rate": 0.002748, + "loss": 4.4304, + "step": 916 + }, + { + "epoch": 0.00917, + "grad_norm": 0.6458845430146066, + "learning_rate": 0.002751, + "loss": 4.4431, + "step": 917 + }, + { + "epoch": 0.00918, + "grad_norm": 0.6107710101855935, + "learning_rate": 0.0027540000000000004, + "loss": 4.4622, + "step": 918 + }, + { + "epoch": 0.00919, + "grad_norm": 0.6056157888916294, + "learning_rate": 0.0027570000000000003, + "loss": 4.4436, + "step": 919 + }, + { + "epoch": 0.0092, + "grad_norm": 0.6858065732895877, + "learning_rate": 0.0027600000000000003, + "loss": 4.462, + "step": 920 + }, + { + "epoch": 0.00921, + "grad_norm": 0.9391926745722488, + "learning_rate": 0.0027630000000000003, + "loss": 4.4556, + "step": 921 + }, + { + "epoch": 0.00922, + "grad_norm": 1.1348542218598812, + "learning_rate": 0.0027660000000000002, + "loss": 4.5063, + "step": 922 + }, + { + "epoch": 0.00923, + "grad_norm": 1.0421503736233508, + "learning_rate": 0.002769, + "loss": 4.4713, + "step": 923 + }, + { + "epoch": 0.00924, + "grad_norm": 1.0588478925879097, + "learning_rate": 0.002772, + "loss": 4.4959, + "step": 924 + }, + { + "epoch": 0.00925, + "grad_norm": 0.9781832929515508, + "learning_rate": 0.002775, + "loss": 4.4806, + "step": 925 + }, + { + "epoch": 0.00926, + "grad_norm": 0.7776497655560727, + "learning_rate": 0.002778, + "loss": 4.488, + "step": 926 + }, + { + "epoch": 0.00927, + "grad_norm": 0.7034519817521436, + "learning_rate": 0.002781, + "loss": 4.4448, + "step": 927 + }, + { + "epoch": 0.00928, + "grad_norm": 0.9270286412247504, + "learning_rate": 0.002784, + "loss": 4.4965, + "step": 928 + }, + { + "epoch": 0.00929, + "grad_norm": 1.0728170784278697, + "learning_rate": 0.0027870000000000004, + "loss": 4.4869, + "step": 929 + }, + { + "epoch": 0.0093, + "grad_norm": 0.9492127377122095, + "learning_rate": 0.0027900000000000004, + "loss": 4.4732, + "step": 930 + }, + { + "epoch": 0.00931, + "grad_norm": 0.9513900088751025, + "learning_rate": 0.0027930000000000003, + "loss": 4.4833, + "step": 931 + }, + { + "epoch": 0.00932, + "grad_norm": 1.4536440294804005, + "learning_rate": 0.0027960000000000003, + "loss": 4.5168, + "step": 932 + }, + { + "epoch": 0.00933, + "grad_norm": 1.132862147568844, + "learning_rate": 0.0027990000000000003, + "loss": 4.4769, + "step": 933 + }, + { + "epoch": 0.00934, + "grad_norm": 0.8513409648274727, + "learning_rate": 0.0028020000000000002, + "loss": 4.4941, + "step": 934 + }, + { + "epoch": 0.00935, + "grad_norm": 0.8677938211616196, + "learning_rate": 0.002805, + "loss": 4.5057, + "step": 935 + }, + { + "epoch": 0.00936, + "grad_norm": 0.8298971693271944, + "learning_rate": 0.002808, + "loss": 4.5081, + "step": 936 + }, + { + "epoch": 0.00937, + "grad_norm": 0.8363829119527492, + "learning_rate": 0.002811, + "loss": 4.521, + "step": 937 + }, + { + "epoch": 0.00938, + "grad_norm": 1.0036953395826609, + "learning_rate": 0.002814, + "loss": 4.5006, + "step": 938 + }, + { + "epoch": 0.00939, + "grad_norm": 1.0054078613176451, + "learning_rate": 0.002817, + "loss": 4.5073, + "step": 939 + }, + { + "epoch": 0.0094, + "grad_norm": 0.8667374621688471, + "learning_rate": 0.00282, + "loss": 4.4958, + "step": 940 + }, + { + "epoch": 0.00941, + "grad_norm": 0.9205808599892458, + "learning_rate": 0.002823, + "loss": 4.4733, + "step": 941 + }, + { + "epoch": 0.00942, + "grad_norm": 0.9778408651584425, + "learning_rate": 0.002826, + "loss": 4.5008, + "step": 942 + }, + { + "epoch": 0.00943, + "grad_norm": 1.082550194860624, + "learning_rate": 0.002829, + "loss": 4.4958, + "step": 943 + }, + { + "epoch": 0.00944, + "grad_norm": 1.355725245571646, + "learning_rate": 0.002832, + "loss": 4.5179, + "step": 944 + }, + { + "epoch": 0.00945, + "grad_norm": 0.8345488861943434, + "learning_rate": 0.002835, + "loss": 4.5015, + "step": 945 + }, + { + "epoch": 0.00946, + "grad_norm": 1.0009931674850154, + "learning_rate": 0.002838, + "loss": 4.5193, + "step": 946 + }, + { + "epoch": 0.00947, + "grad_norm": 1.0176650405493206, + "learning_rate": 0.0028409999999999998, + "loss": 4.4912, + "step": 947 + }, + { + "epoch": 0.00948, + "grad_norm": 0.836896216552642, + "learning_rate": 0.0028439999999999997, + "loss": 4.472, + "step": 948 + }, + { + "epoch": 0.00949, + "grad_norm": 0.6881971993105954, + "learning_rate": 0.002847, + "loss": 4.5057, + "step": 949 + }, + { + "epoch": 0.0095, + "grad_norm": 0.8926787293989068, + "learning_rate": 0.00285, + "loss": 4.5027, + "step": 950 + }, + { + "epoch": 0.00951, + "grad_norm": 1.0437772730376889, + "learning_rate": 0.002853, + "loss": 4.4861, + "step": 951 + }, + { + "epoch": 0.00952, + "grad_norm": 0.8745743913439339, + "learning_rate": 0.002856, + "loss": 4.5041, + "step": 952 + }, + { + "epoch": 0.00953, + "grad_norm": 0.7847706213592531, + "learning_rate": 0.002859, + "loss": 4.4446, + "step": 953 + }, + { + "epoch": 0.00954, + "grad_norm": 0.6236105730880978, + "learning_rate": 0.002862, + "loss": 4.4945, + "step": 954 + }, + { + "epoch": 0.00955, + "grad_norm": 0.5696186051972435, + "learning_rate": 0.002865, + "loss": 4.475, + "step": 955 + }, + { + "epoch": 0.00956, + "grad_norm": 0.5459272314199634, + "learning_rate": 0.002868, + "loss": 4.4518, + "step": 956 + }, + { + "epoch": 0.00957, + "grad_norm": 0.5100325019322003, + "learning_rate": 0.002871, + "loss": 4.4721, + "step": 957 + }, + { + "epoch": 0.00958, + "grad_norm": 0.6617022302690957, + "learning_rate": 0.002874, + "loss": 4.4567, + "step": 958 + }, + { + "epoch": 0.00959, + "grad_norm": 0.7948420381771908, + "learning_rate": 0.002877, + "loss": 4.4693, + "step": 959 + }, + { + "epoch": 0.0096, + "grad_norm": 1.0277128972108451, + "learning_rate": 0.0028799999999999997, + "loss": 4.4747, + "step": 960 + }, + { + "epoch": 0.00961, + "grad_norm": 0.9925275083373442, + "learning_rate": 0.002883, + "loss": 4.4507, + "step": 961 + }, + { + "epoch": 0.00962, + "grad_norm": 1.0177847800658486, + "learning_rate": 0.002886, + "loss": 4.4861, + "step": 962 + }, + { + "epoch": 0.00963, + "grad_norm": 1.0118802628275685, + "learning_rate": 0.002889, + "loss": 4.4865, + "step": 963 + }, + { + "epoch": 0.00964, + "grad_norm": 1.1856958623453784, + "learning_rate": 0.002892, + "loss": 4.4868, + "step": 964 + }, + { + "epoch": 0.00965, + "grad_norm": 0.8341279969334199, + "learning_rate": 0.002895, + "loss": 4.4524, + "step": 965 + }, + { + "epoch": 0.00966, + "grad_norm": 0.6756655743310646, + "learning_rate": 0.002898, + "loss": 4.4552, + "step": 966 + }, + { + "epoch": 0.00967, + "grad_norm": 0.6634929266596646, + "learning_rate": 0.002901, + "loss": 4.4933, + "step": 967 + }, + { + "epoch": 0.00968, + "grad_norm": 0.6850072615364151, + "learning_rate": 0.002904, + "loss": 4.4564, + "step": 968 + }, + { + "epoch": 0.00969, + "grad_norm": 0.6166935294692494, + "learning_rate": 0.002907, + "loss": 4.4862, + "step": 969 + }, + { + "epoch": 0.0097, + "grad_norm": 0.6616720297110108, + "learning_rate": 0.00291, + "loss": 4.4401, + "step": 970 + }, + { + "epoch": 0.00971, + "grad_norm": 0.5463010292381552, + "learning_rate": 0.002913, + "loss": 4.4415, + "step": 971 + }, + { + "epoch": 0.00972, + "grad_norm": 0.5115285821904162, + "learning_rate": 0.002916, + "loss": 4.4454, + "step": 972 + }, + { + "epoch": 0.00973, + "grad_norm": 0.543781975527911, + "learning_rate": 0.002919, + "loss": 4.4402, + "step": 973 + }, + { + "epoch": 0.00974, + "grad_norm": 0.5167263203162235, + "learning_rate": 0.002922, + "loss": 4.4246, + "step": 974 + }, + { + "epoch": 0.00975, + "grad_norm": 0.4676730441900605, + "learning_rate": 0.002925, + "loss": 4.4434, + "step": 975 + }, + { + "epoch": 0.00976, + "grad_norm": 0.4483140513716565, + "learning_rate": 0.002928, + "loss": 4.4259, + "step": 976 + }, + { + "epoch": 0.00977, + "grad_norm": 0.45300675248114197, + "learning_rate": 0.002931, + "loss": 4.4582, + "step": 977 + }, + { + "epoch": 0.00978, + "grad_norm": 0.524704391185419, + "learning_rate": 0.002934, + "loss": 4.4335, + "step": 978 + }, + { + "epoch": 0.00979, + "grad_norm": 0.6643809565846889, + "learning_rate": 0.002937, + "loss": 4.4266, + "step": 979 + }, + { + "epoch": 0.0098, + "grad_norm": 0.9159629342753403, + "learning_rate": 0.00294, + "loss": 4.433, + "step": 980 + }, + { + "epoch": 0.00981, + "grad_norm": 1.0134318026936866, + "learning_rate": 0.002943, + "loss": 4.4516, + "step": 981 + }, + { + "epoch": 0.00982, + "grad_norm": 0.7532241089289973, + "learning_rate": 0.002946, + "loss": 4.4387, + "step": 982 + }, + { + "epoch": 0.00983, + "grad_norm": 0.7539130672753217, + "learning_rate": 0.0029490000000000002, + "loss": 4.4372, + "step": 983 + }, + { + "epoch": 0.00984, + "grad_norm": 0.7045823796689694, + "learning_rate": 0.002952, + "loss": 4.4213, + "step": 984 + }, + { + "epoch": 0.00985, + "grad_norm": 0.6478785171714704, + "learning_rate": 0.002955, + "loss": 4.444, + "step": 985 + }, + { + "epoch": 0.00986, + "grad_norm": 0.5640139479823427, + "learning_rate": 0.002958, + "loss": 4.3997, + "step": 986 + }, + { + "epoch": 0.00987, + "grad_norm": 0.6145958247621988, + "learning_rate": 0.002961, + "loss": 4.4141, + "step": 987 + }, + { + "epoch": 0.00988, + "grad_norm": 0.7121366087401472, + "learning_rate": 0.002964, + "loss": 4.4512, + "step": 988 + }, + { + "epoch": 0.00989, + "grad_norm": 0.7474063595618389, + "learning_rate": 0.002967, + "loss": 4.3897, + "step": 989 + }, + { + "epoch": 0.0099, + "grad_norm": 0.7608901784540225, + "learning_rate": 0.00297, + "loss": 4.4296, + "step": 990 + }, + { + "epoch": 0.00991, + "grad_norm": 0.8136541225519112, + "learning_rate": 0.002973, + "loss": 4.4314, + "step": 991 + }, + { + "epoch": 0.00992, + "grad_norm": 0.7865701920195308, + "learning_rate": 0.002976, + "loss": 4.4266, + "step": 992 + }, + { + "epoch": 0.00993, + "grad_norm": 0.8315737176917932, + "learning_rate": 0.002979, + "loss": 4.4267, + "step": 993 + }, + { + "epoch": 0.00994, + "grad_norm": 0.7821802322270756, + "learning_rate": 0.002982, + "loss": 4.4281, + "step": 994 + }, + { + "epoch": 0.00995, + "grad_norm": 0.8705272973695986, + "learning_rate": 0.0029850000000000002, + "loss": 4.4331, + "step": 995 + }, + { + "epoch": 0.00996, + "grad_norm": 0.9954881536889274, + "learning_rate": 0.002988, + "loss": 4.4408, + "step": 996 + }, + { + "epoch": 0.00997, + "grad_norm": 0.9720370341600497, + "learning_rate": 0.002991, + "loss": 4.4354, + "step": 997 + }, + { + "epoch": 0.00998, + "grad_norm": 0.7522042168889891, + "learning_rate": 0.002994, + "loss": 4.4331, + "step": 998 + }, + { + "epoch": 0.00999, + "grad_norm": 0.9425882614336212, + "learning_rate": 0.002997, + "loss": 4.4261, + "step": 999 + }, + { + "epoch": 0.01, + "grad_norm": 0.9689827453070083, + "learning_rate": 0.003, + "loss": 4.458, + "step": 1000 + }, + { + "epoch": 0.01001, + "grad_norm": 0.9802085236972514, + "learning_rate": 0.003, + "loss": 4.4569, + "step": 1001 + }, + { + "epoch": 0.01002, + "grad_norm": 0.9742392022619593, + "learning_rate": 0.003, + "loss": 4.4262, + "step": 1002 + }, + { + "epoch": 0.01003, + "grad_norm": 0.9115330327806416, + "learning_rate": 0.003, + "loss": 4.4513, + "step": 1003 + }, + { + "epoch": 0.01004, + "grad_norm": 0.8562148232052564, + "learning_rate": 0.003, + "loss": 4.4397, + "step": 1004 + }, + { + "epoch": 0.01005, + "grad_norm": 0.9105663755998641, + "learning_rate": 0.003, + "loss": 4.4154, + "step": 1005 + }, + { + "epoch": 0.01006, + "grad_norm": 1.018045955439956, + "learning_rate": 0.003, + "loss": 4.433, + "step": 1006 + }, + { + "epoch": 0.01007, + "grad_norm": 0.9357911521230117, + "learning_rate": 0.003, + "loss": 4.4713, + "step": 1007 + }, + { + "epoch": 0.01008, + "grad_norm": 0.9681416399703533, + "learning_rate": 0.003, + "loss": 4.4846, + "step": 1008 + }, + { + "epoch": 0.01009, + "grad_norm": 1.030244827646589, + "learning_rate": 0.003, + "loss": 4.4619, + "step": 1009 + }, + { + "epoch": 0.0101, + "grad_norm": 1.0010362528613534, + "learning_rate": 0.003, + "loss": 4.4688, + "step": 1010 + }, + { + "epoch": 0.01011, + "grad_norm": 1.0284858383133122, + "learning_rate": 0.003, + "loss": 4.4881, + "step": 1011 + }, + { + "epoch": 0.01012, + "grad_norm": 0.9778345429910184, + "learning_rate": 0.003, + "loss": 4.4674, + "step": 1012 + }, + { + "epoch": 0.01013, + "grad_norm": 0.8705611016855861, + "learning_rate": 0.003, + "loss": 4.4919, + "step": 1013 + }, + { + "epoch": 0.01014, + "grad_norm": 0.8083144078788229, + "learning_rate": 0.003, + "loss": 4.4268, + "step": 1014 + }, + { + "epoch": 0.01015, + "grad_norm": 0.7155205086193928, + "learning_rate": 0.003, + "loss": 4.4657, + "step": 1015 + }, + { + "epoch": 0.01016, + "grad_norm": 0.5844948538982189, + "learning_rate": 0.003, + "loss": 4.466, + "step": 1016 + }, + { + "epoch": 0.01017, + "grad_norm": 0.50419422786756, + "learning_rate": 0.003, + "loss": 4.4019, + "step": 1017 + }, + { + "epoch": 0.01018, + "grad_norm": 0.5273528303530755, + "learning_rate": 0.003, + "loss": 4.4545, + "step": 1018 + }, + { + "epoch": 0.01019, + "grad_norm": 0.5329688106940915, + "learning_rate": 0.003, + "loss": 4.4315, + "step": 1019 + }, + { + "epoch": 0.0102, + "grad_norm": 0.6285459843623249, + "learning_rate": 0.003, + "loss": 4.4292, + "step": 1020 + }, + { + "epoch": 0.01021, + "grad_norm": 0.6444525489483212, + "learning_rate": 0.003, + "loss": 4.412, + "step": 1021 + }, + { + "epoch": 0.01022, + "grad_norm": 0.6609826768689684, + "learning_rate": 0.003, + "loss": 4.4194, + "step": 1022 + }, + { + "epoch": 0.01023, + "grad_norm": 0.6479610287689606, + "learning_rate": 0.003, + "loss": 4.4151, + "step": 1023 + }, + { + "epoch": 0.01024, + "grad_norm": 0.7091931447524652, + "learning_rate": 0.003, + "loss": 4.3798, + "step": 1024 + }, + { + "epoch": 0.01025, + "grad_norm": 0.7026391214213478, + "learning_rate": 0.003, + "loss": 4.4116, + "step": 1025 + }, + { + "epoch": 0.01026, + "grad_norm": 0.6731332289892269, + "learning_rate": 0.003, + "loss": 4.3991, + "step": 1026 + }, + { + "epoch": 0.01027, + "grad_norm": 0.6590644472165706, + "learning_rate": 0.003, + "loss": 4.4177, + "step": 1027 + }, + { + "epoch": 0.01028, + "grad_norm": 0.7485101036485022, + "learning_rate": 0.003, + "loss": 4.4014, + "step": 1028 + }, + { + "epoch": 0.01029, + "grad_norm": 0.7198477689690366, + "learning_rate": 0.003, + "loss": 4.3803, + "step": 1029 + }, + { + "epoch": 0.0103, + "grad_norm": 0.5542753635749327, + "learning_rate": 0.003, + "loss": 4.4023, + "step": 1030 + }, + { + "epoch": 0.01031, + "grad_norm": 0.5292390629019561, + "learning_rate": 0.003, + "loss": 4.4144, + "step": 1031 + }, + { + "epoch": 0.01032, + "grad_norm": 0.500926566427153, + "learning_rate": 0.003, + "loss": 4.3837, + "step": 1032 + }, + { + "epoch": 0.01033, + "grad_norm": 0.49056326531128164, + "learning_rate": 0.003, + "loss": 4.4201, + "step": 1033 + }, + { + "epoch": 0.01034, + "grad_norm": 0.5309292795236984, + "learning_rate": 0.003, + "loss": 4.3865, + "step": 1034 + }, + { + "epoch": 0.01035, + "grad_norm": 0.6084296376545847, + "learning_rate": 0.003, + "loss": 4.3892, + "step": 1035 + }, + { + "epoch": 0.01036, + "grad_norm": 0.8615895093148164, + "learning_rate": 0.003, + "loss": 4.3883, + "step": 1036 + }, + { + "epoch": 0.01037, + "grad_norm": 0.8936988458454226, + "learning_rate": 0.003, + "loss": 4.4336, + "step": 1037 + }, + { + "epoch": 0.01038, + "grad_norm": 0.6948017038229403, + "learning_rate": 0.003, + "loss": 4.4116, + "step": 1038 + }, + { + "epoch": 0.01039, + "grad_norm": 0.9114790426144561, + "learning_rate": 0.003, + "loss": 4.4152, + "step": 1039 + }, + { + "epoch": 0.0104, + "grad_norm": 0.8448966818619524, + "learning_rate": 0.003, + "loss": 4.4288, + "step": 1040 + }, + { + "epoch": 0.01041, + "grad_norm": 0.7995140521375168, + "learning_rate": 0.003, + "loss": 4.404, + "step": 1041 + }, + { + "epoch": 0.01042, + "grad_norm": 0.8979223927667839, + "learning_rate": 0.003, + "loss": 4.4357, + "step": 1042 + }, + { + "epoch": 0.01043, + "grad_norm": 0.7233892787514891, + "learning_rate": 0.003, + "loss": 4.3775, + "step": 1043 + }, + { + "epoch": 0.01044, + "grad_norm": 0.782866039824708, + "learning_rate": 0.003, + "loss": 4.4144, + "step": 1044 + }, + { + "epoch": 0.01045, + "grad_norm": 0.841755171402396, + "learning_rate": 0.003, + "loss": 4.4353, + "step": 1045 + }, + { + "epoch": 0.01046, + "grad_norm": 0.780348729633882, + "learning_rate": 0.003, + "loss": 4.4043, + "step": 1046 + }, + { + "epoch": 0.01047, + "grad_norm": 0.8623775519511051, + "learning_rate": 0.003, + "loss": 4.4135, + "step": 1047 + }, + { + "epoch": 0.01048, + "grad_norm": 0.9009461265034386, + "learning_rate": 0.003, + "loss": 4.4049, + "step": 1048 + }, + { + "epoch": 0.01049, + "grad_norm": 0.739794637514069, + "learning_rate": 0.003, + "loss": 4.3786, + "step": 1049 + }, + { + "epoch": 0.0105, + "grad_norm": 0.6259101207085414, + "learning_rate": 0.003, + "loss": 4.4154, + "step": 1050 + }, + { + "epoch": 0.01051, + "grad_norm": 0.6694575857098493, + "learning_rate": 0.003, + "loss": 4.393, + "step": 1051 + }, + { + "epoch": 0.01052, + "grad_norm": 0.7141380180592181, + "learning_rate": 0.003, + "loss": 4.4052, + "step": 1052 + }, + { + "epoch": 0.01053, + "grad_norm": 0.7943106393078887, + "learning_rate": 0.003, + "loss": 4.4143, + "step": 1053 + }, + { + "epoch": 0.01054, + "grad_norm": 0.7478655479661611, + "learning_rate": 0.003, + "loss": 4.4077, + "step": 1054 + }, + { + "epoch": 0.01055, + "grad_norm": 0.707966249364238, + "learning_rate": 0.003, + "loss": 4.3861, + "step": 1055 + }, + { + "epoch": 0.01056, + "grad_norm": 0.6358438963919717, + "learning_rate": 0.003, + "loss": 4.3823, + "step": 1056 + }, + { + "epoch": 0.01057, + "grad_norm": 0.5691063929797074, + "learning_rate": 0.003, + "loss": 4.3849, + "step": 1057 + }, + { + "epoch": 0.01058, + "grad_norm": 0.5109096197008457, + "learning_rate": 0.003, + "loss": 4.3737, + "step": 1058 + }, + { + "epoch": 0.01059, + "grad_norm": 0.419033279294021, + "learning_rate": 0.003, + "loss": 4.3749, + "step": 1059 + }, + { + "epoch": 0.0106, + "grad_norm": 0.4506837031966603, + "learning_rate": 0.003, + "loss": 4.3654, + "step": 1060 + }, + { + "epoch": 0.01061, + "grad_norm": 0.42662194113596513, + "learning_rate": 0.003, + "loss": 4.361, + "step": 1061 + }, + { + "epoch": 0.01062, + "grad_norm": 0.5018523046776621, + "learning_rate": 0.003, + "loss": 4.3568, + "step": 1062 + }, + { + "epoch": 0.01063, + "grad_norm": 0.6580287488917459, + "learning_rate": 0.003, + "loss": 4.3552, + "step": 1063 + }, + { + "epoch": 0.01064, + "grad_norm": 0.831793270479749, + "learning_rate": 0.003, + "loss": 4.3844, + "step": 1064 + }, + { + "epoch": 0.01065, + "grad_norm": 1.0701105543701324, + "learning_rate": 0.003, + "loss": 4.3903, + "step": 1065 + }, + { + "epoch": 0.01066, + "grad_norm": 0.943731953453328, + "learning_rate": 0.003, + "loss": 4.4118, + "step": 1066 + }, + { + "epoch": 0.01067, + "grad_norm": 0.9236025360095688, + "learning_rate": 0.003, + "loss": 4.4136, + "step": 1067 + }, + { + "epoch": 0.01068, + "grad_norm": 1.1282330525862354, + "learning_rate": 0.003, + "loss": 4.4118, + "step": 1068 + }, + { + "epoch": 0.01069, + "grad_norm": 0.840410887973158, + "learning_rate": 0.003, + "loss": 4.4054, + "step": 1069 + }, + { + "epoch": 0.0107, + "grad_norm": 0.733631994312363, + "learning_rate": 0.003, + "loss": 4.4566, + "step": 1070 + }, + { + "epoch": 0.01071, + "grad_norm": 0.7799052277350242, + "learning_rate": 0.003, + "loss": 4.3774, + "step": 1071 + }, + { + "epoch": 0.01072, + "grad_norm": 0.7960071687741883, + "learning_rate": 0.003, + "loss": 4.4136, + "step": 1072 + }, + { + "epoch": 0.01073, + "grad_norm": 0.6975430946737453, + "learning_rate": 0.003, + "loss": 4.4079, + "step": 1073 + }, + { + "epoch": 0.01074, + "grad_norm": 0.6813769264433432, + "learning_rate": 0.003, + "loss": 4.37, + "step": 1074 + }, + { + "epoch": 0.01075, + "grad_norm": 0.5971783233576602, + "learning_rate": 0.003, + "loss": 4.4063, + "step": 1075 + }, + { + "epoch": 0.01076, + "grad_norm": 0.5510713740534053, + "learning_rate": 0.003, + "loss": 4.3967, + "step": 1076 + }, + { + "epoch": 0.01077, + "grad_norm": 0.5595097233894012, + "learning_rate": 0.003, + "loss": 4.3917, + "step": 1077 + }, + { + "epoch": 0.01078, + "grad_norm": 0.6230934184673731, + "learning_rate": 0.003, + "loss": 4.3491, + "step": 1078 + }, + { + "epoch": 0.01079, + "grad_norm": 0.7178707958532615, + "learning_rate": 0.003, + "loss": 4.3825, + "step": 1079 + }, + { + "epoch": 0.0108, + "grad_norm": 0.7484447643520958, + "learning_rate": 0.003, + "loss": 4.3665, + "step": 1080 + }, + { + "epoch": 0.01081, + "grad_norm": 0.780034137982597, + "learning_rate": 0.003, + "loss": 4.3727, + "step": 1081 + }, + { + "epoch": 0.01082, + "grad_norm": 0.7403701975238451, + "learning_rate": 0.003, + "loss": 4.3894, + "step": 1082 + }, + { + "epoch": 0.01083, + "grad_norm": 0.6305537452051372, + "learning_rate": 0.003, + "loss": 4.3733, + "step": 1083 + }, + { + "epoch": 0.01084, + "grad_norm": 0.7435200213630843, + "learning_rate": 0.003, + "loss": 4.3825, + "step": 1084 + }, + { + "epoch": 0.01085, + "grad_norm": 0.7405129376016774, + "learning_rate": 0.003, + "loss": 4.3735, + "step": 1085 + }, + { + "epoch": 0.01086, + "grad_norm": 0.6037526750507048, + "learning_rate": 0.003, + "loss": 4.3449, + "step": 1086 + }, + { + "epoch": 0.01087, + "grad_norm": 0.5349574840042758, + "learning_rate": 0.003, + "loss": 4.3549, + "step": 1087 + }, + { + "epoch": 0.01088, + "grad_norm": 0.5555341529780033, + "learning_rate": 0.003, + "loss": 4.3917, + "step": 1088 + }, + { + "epoch": 0.01089, + "grad_norm": 0.6254527345250118, + "learning_rate": 0.003, + "loss": 4.3599, + "step": 1089 + }, + { + "epoch": 0.0109, + "grad_norm": 0.6828156926551208, + "learning_rate": 0.003, + "loss": 4.3884, + "step": 1090 + }, + { + "epoch": 0.01091, + "grad_norm": 0.5753021180468998, + "learning_rate": 0.003, + "loss": 4.3549, + "step": 1091 + }, + { + "epoch": 0.01092, + "grad_norm": 0.4077341561870379, + "learning_rate": 0.003, + "loss": 4.3436, + "step": 1092 + }, + { + "epoch": 0.01093, + "grad_norm": 0.47615819716416974, + "learning_rate": 0.003, + "loss": 4.3664, + "step": 1093 + }, + { + "epoch": 0.01094, + "grad_norm": 0.5845002574061751, + "learning_rate": 0.003, + "loss": 4.3553, + "step": 1094 + }, + { + "epoch": 0.01095, + "grad_norm": 0.8583788887394415, + "learning_rate": 0.003, + "loss": 4.3944, + "step": 1095 + }, + { + "epoch": 0.01096, + "grad_norm": 1.0479503783625854, + "learning_rate": 0.003, + "loss": 4.3824, + "step": 1096 + }, + { + "epoch": 0.01097, + "grad_norm": 0.9584408455564232, + "learning_rate": 0.003, + "loss": 4.3695, + "step": 1097 + }, + { + "epoch": 0.01098, + "grad_norm": 1.2545731298718157, + "learning_rate": 0.003, + "loss": 4.4068, + "step": 1098 + }, + { + "epoch": 0.01099, + "grad_norm": 0.8305012272413557, + "learning_rate": 0.003, + "loss": 4.3706, + "step": 1099 + }, + { + "epoch": 0.011, + "grad_norm": 0.7311868902259275, + "learning_rate": 0.003, + "loss": 4.3677, + "step": 1100 + }, + { + "epoch": 0.01101, + "grad_norm": 0.6742597336847759, + "learning_rate": 0.003, + "loss": 4.3752, + "step": 1101 + }, + { + "epoch": 0.01102, + "grad_norm": 0.6773795133108548, + "learning_rate": 0.003, + "loss": 4.3802, + "step": 1102 + }, + { + "epoch": 0.01103, + "grad_norm": 0.6414092553075785, + "learning_rate": 0.003, + "loss": 4.3701, + "step": 1103 + }, + { + "epoch": 0.01104, + "grad_norm": 0.6378763268399013, + "learning_rate": 0.003, + "loss": 4.3589, + "step": 1104 + }, + { + "epoch": 0.01105, + "grad_norm": 0.6837675260878959, + "learning_rate": 0.003, + "loss": 4.3856, + "step": 1105 + }, + { + "epoch": 0.01106, + "grad_norm": 0.667950958790211, + "learning_rate": 0.003, + "loss": 4.3574, + "step": 1106 + }, + { + "epoch": 0.01107, + "grad_norm": 0.6082482122734244, + "learning_rate": 0.003, + "loss": 4.3623, + "step": 1107 + }, + { + "epoch": 0.01108, + "grad_norm": 0.5983800302348057, + "learning_rate": 0.003, + "loss": 4.3313, + "step": 1108 + }, + { + "epoch": 0.01109, + "grad_norm": 0.6046679697278923, + "learning_rate": 0.003, + "loss": 4.3476, + "step": 1109 + }, + { + "epoch": 0.0111, + "grad_norm": 0.6353880163110637, + "learning_rate": 0.003, + "loss": 4.3675, + "step": 1110 + }, + { + "epoch": 0.01111, + "grad_norm": 0.7049628214149826, + "learning_rate": 0.003, + "loss": 4.3473, + "step": 1111 + }, + { + "epoch": 0.01112, + "grad_norm": 0.758960503794264, + "learning_rate": 0.003, + "loss": 4.369, + "step": 1112 + }, + { + "epoch": 0.01113, + "grad_norm": 0.6091686772357772, + "learning_rate": 0.003, + "loss": 4.3335, + "step": 1113 + }, + { + "epoch": 0.01114, + "grad_norm": 0.628169460357088, + "learning_rate": 0.003, + "loss": 4.3462, + "step": 1114 + }, + { + "epoch": 0.01115, + "grad_norm": 0.7453432538813878, + "learning_rate": 0.003, + "loss": 4.3779, + "step": 1115 + }, + { + "epoch": 0.01116, + "grad_norm": 0.9959163456029771, + "learning_rate": 0.003, + "loss": 4.3771, + "step": 1116 + }, + { + "epoch": 0.01117, + "grad_norm": 0.9566018042437344, + "learning_rate": 0.003, + "loss": 4.3915, + "step": 1117 + }, + { + "epoch": 0.01118, + "grad_norm": 0.6794219206934775, + "learning_rate": 0.003, + "loss": 4.372, + "step": 1118 + }, + { + "epoch": 0.01119, + "grad_norm": 0.6027310157842346, + "learning_rate": 0.003, + "loss": 4.375, + "step": 1119 + }, + { + "epoch": 0.0112, + "grad_norm": 0.6064414406921254, + "learning_rate": 0.003, + "loss": 4.3588, + "step": 1120 + }, + { + "epoch": 0.01121, + "grad_norm": 0.643465354938861, + "learning_rate": 0.003, + "loss": 4.3691, + "step": 1121 + }, + { + "epoch": 0.01122, + "grad_norm": 0.8439491151148678, + "learning_rate": 0.003, + "loss": 4.3539, + "step": 1122 + }, + { + "epoch": 0.01123, + "grad_norm": 0.857123487934385, + "learning_rate": 0.003, + "loss": 4.345, + "step": 1123 + }, + { + "epoch": 0.01124, + "grad_norm": 0.7412296015988188, + "learning_rate": 0.003, + "loss": 4.3865, + "step": 1124 + }, + { + "epoch": 0.01125, + "grad_norm": 0.8143761816040683, + "learning_rate": 0.003, + "loss": 4.3583, + "step": 1125 + }, + { + "epoch": 0.01126, + "grad_norm": 0.741767094028342, + "learning_rate": 0.003, + "loss": 4.3618, + "step": 1126 + }, + { + "epoch": 0.01127, + "grad_norm": 0.737420855232079, + "learning_rate": 0.003, + "loss": 4.3948, + "step": 1127 + }, + { + "epoch": 0.01128, + "grad_norm": 0.7656500975813709, + "learning_rate": 0.003, + "loss": 4.372, + "step": 1128 + }, + { + "epoch": 0.01129, + "grad_norm": 0.8248659664813461, + "learning_rate": 0.003, + "loss": 4.3685, + "step": 1129 + }, + { + "epoch": 0.0113, + "grad_norm": 0.7756542452356111, + "learning_rate": 0.003, + "loss": 4.3447, + "step": 1130 + }, + { + "epoch": 0.01131, + "grad_norm": 0.7965849423705411, + "learning_rate": 0.003, + "loss": 4.3615, + "step": 1131 + }, + { + "epoch": 0.01132, + "grad_norm": 0.718692737781693, + "learning_rate": 0.003, + "loss": 4.3824, + "step": 1132 + }, + { + "epoch": 0.01133, + "grad_norm": 0.6500905111846983, + "learning_rate": 0.003, + "loss": 4.34, + "step": 1133 + }, + { + "epoch": 0.01134, + "grad_norm": 0.6052834019155894, + "learning_rate": 0.003, + "loss": 4.3525, + "step": 1134 + }, + { + "epoch": 0.01135, + "grad_norm": 0.57728723025885, + "learning_rate": 0.003, + "loss": 4.3767, + "step": 1135 + }, + { + "epoch": 0.01136, + "grad_norm": 0.6622614298653483, + "learning_rate": 0.003, + "loss": 4.3514, + "step": 1136 + }, + { + "epoch": 0.01137, + "grad_norm": 0.6945355696134266, + "learning_rate": 0.003, + "loss": 4.3784, + "step": 1137 + }, + { + "epoch": 0.01138, + "grad_norm": 0.7423434054596434, + "learning_rate": 0.003, + "loss": 4.3361, + "step": 1138 + }, + { + "epoch": 0.01139, + "grad_norm": 0.7678127530736037, + "learning_rate": 0.003, + "loss": 4.3463, + "step": 1139 + }, + { + "epoch": 0.0114, + "grad_norm": 0.7470822118949633, + "learning_rate": 0.003, + "loss": 4.3322, + "step": 1140 + }, + { + "epoch": 0.01141, + "grad_norm": 0.903198095826591, + "learning_rate": 0.003, + "loss": 4.3608, + "step": 1141 + }, + { + "epoch": 0.01142, + "grad_norm": 1.0797547624524322, + "learning_rate": 0.003, + "loss": 4.3612, + "step": 1142 + }, + { + "epoch": 0.01143, + "grad_norm": 0.7877548466868329, + "learning_rate": 0.003, + "loss": 4.3542, + "step": 1143 + }, + { + "epoch": 0.01144, + "grad_norm": 0.8147882944795127, + "learning_rate": 0.003, + "loss": 4.3253, + "step": 1144 + }, + { + "epoch": 0.01145, + "grad_norm": 0.7148112106554612, + "learning_rate": 0.003, + "loss": 4.3655, + "step": 1145 + }, + { + "epoch": 0.01146, + "grad_norm": 0.6580426381015894, + "learning_rate": 0.003, + "loss": 4.3445, + "step": 1146 + }, + { + "epoch": 0.01147, + "grad_norm": 0.7448256416265617, + "learning_rate": 0.003, + "loss": 4.3367, + "step": 1147 + }, + { + "epoch": 0.01148, + "grad_norm": 0.6968832253815664, + "learning_rate": 0.003, + "loss": 4.3386, + "step": 1148 + }, + { + "epoch": 0.01149, + "grad_norm": 0.627190656091196, + "learning_rate": 0.003, + "loss": 4.3527, + "step": 1149 + }, + { + "epoch": 0.0115, + "grad_norm": 0.6739971655272956, + "learning_rate": 0.003, + "loss": 4.3445, + "step": 1150 + }, + { + "epoch": 0.01151, + "grad_norm": 0.7518825421443525, + "learning_rate": 0.003, + "loss": 4.3445, + "step": 1151 + }, + { + "epoch": 0.01152, + "grad_norm": 0.7628403894732813, + "learning_rate": 0.003, + "loss": 4.3529, + "step": 1152 + }, + { + "epoch": 0.01153, + "grad_norm": 0.726109999889872, + "learning_rate": 0.003, + "loss": 4.3668, + "step": 1153 + }, + { + "epoch": 0.01154, + "grad_norm": 0.7747132516080261, + "learning_rate": 0.003, + "loss": 4.3438, + "step": 1154 + }, + { + "epoch": 0.01155, + "grad_norm": 0.8180276502719757, + "learning_rate": 0.003, + "loss": 4.3369, + "step": 1155 + }, + { + "epoch": 0.01156, + "grad_norm": 0.7757994684934312, + "learning_rate": 0.003, + "loss": 4.3405, + "step": 1156 + }, + { + "epoch": 0.01157, + "grad_norm": 0.7233155240278132, + "learning_rate": 0.003, + "loss": 4.3659, + "step": 1157 + }, + { + "epoch": 0.01158, + "grad_norm": 0.6552894912204377, + "learning_rate": 0.003, + "loss": 4.332, + "step": 1158 + }, + { + "epoch": 0.01159, + "grad_norm": 0.7654422010962233, + "learning_rate": 0.003, + "loss": 4.3316, + "step": 1159 + }, + { + "epoch": 0.0116, + "grad_norm": 0.7129248873282286, + "learning_rate": 0.003, + "loss": 4.3462, + "step": 1160 + }, + { + "epoch": 0.01161, + "grad_norm": 0.728340877307682, + "learning_rate": 0.003, + "loss": 4.3213, + "step": 1161 + }, + { + "epoch": 0.01162, + "grad_norm": 0.6672912532139038, + "learning_rate": 0.003, + "loss": 4.3344, + "step": 1162 + }, + { + "epoch": 0.01163, + "grad_norm": 0.6087688372398936, + "learning_rate": 0.003, + "loss": 4.3512, + "step": 1163 + }, + { + "epoch": 0.01164, + "grad_norm": 0.5012921787632111, + "learning_rate": 0.003, + "loss": 4.3329, + "step": 1164 + }, + { + "epoch": 0.01165, + "grad_norm": 0.4972192156782153, + "learning_rate": 0.003, + "loss": 4.3277, + "step": 1165 + }, + { + "epoch": 0.01166, + "grad_norm": 0.48804976467805194, + "learning_rate": 0.003, + "loss": 4.3262, + "step": 1166 + }, + { + "epoch": 0.01167, + "grad_norm": 0.5481538887564317, + "learning_rate": 0.003, + "loss": 4.2952, + "step": 1167 + }, + { + "epoch": 0.01168, + "grad_norm": 0.6090792597366076, + "learning_rate": 0.003, + "loss": 4.3026, + "step": 1168 + }, + { + "epoch": 0.01169, + "grad_norm": 0.653085333922602, + "learning_rate": 0.003, + "loss": 4.347, + "step": 1169 + }, + { + "epoch": 0.0117, + "grad_norm": 0.62436544437129, + "learning_rate": 0.003, + "loss": 4.3309, + "step": 1170 + }, + { + "epoch": 0.01171, + "grad_norm": 0.654802486268078, + "learning_rate": 0.003, + "loss": 4.3143, + "step": 1171 + }, + { + "epoch": 0.01172, + "grad_norm": 0.6246624927886145, + "learning_rate": 0.003, + "loss": 4.3038, + "step": 1172 + }, + { + "epoch": 0.01173, + "grad_norm": 0.6813303457409844, + "learning_rate": 0.003, + "loss": 4.3382, + "step": 1173 + }, + { + "epoch": 0.01174, + "grad_norm": 0.7721136771646261, + "learning_rate": 0.003, + "loss": 4.3203, + "step": 1174 + }, + { + "epoch": 0.01175, + "grad_norm": 0.823782632958531, + "learning_rate": 0.003, + "loss": 4.3152, + "step": 1175 + }, + { + "epoch": 0.01176, + "grad_norm": 0.9763749462159719, + "learning_rate": 0.003, + "loss": 4.3509, + "step": 1176 + }, + { + "epoch": 0.01177, + "grad_norm": 1.0894015492716702, + "learning_rate": 0.003, + "loss": 4.3444, + "step": 1177 + }, + { + "epoch": 0.01178, + "grad_norm": 0.8156537670820067, + "learning_rate": 0.003, + "loss": 4.3473, + "step": 1178 + }, + { + "epoch": 0.01179, + "grad_norm": 0.8529362760715462, + "learning_rate": 0.003, + "loss": 4.3276, + "step": 1179 + }, + { + "epoch": 0.0118, + "grad_norm": 0.857392382910556, + "learning_rate": 0.003, + "loss": 4.395, + "step": 1180 + }, + { + "epoch": 0.01181, + "grad_norm": 0.8128436446210868, + "learning_rate": 0.003, + "loss": 4.3809, + "step": 1181 + }, + { + "epoch": 0.01182, + "grad_norm": 0.911584010746237, + "learning_rate": 0.003, + "loss": 4.3433, + "step": 1182 + }, + { + "epoch": 0.01183, + "grad_norm": 0.8460613119287651, + "learning_rate": 0.003, + "loss": 4.337, + "step": 1183 + }, + { + "epoch": 0.01184, + "grad_norm": 0.7730717809982315, + "learning_rate": 0.003, + "loss": 4.3753, + "step": 1184 + }, + { + "epoch": 0.01185, + "grad_norm": 0.7718355908910234, + "learning_rate": 0.003, + "loss": 4.3449, + "step": 1185 + }, + { + "epoch": 0.01186, + "grad_norm": 0.749856576054878, + "learning_rate": 0.003, + "loss": 4.3669, + "step": 1186 + }, + { + "epoch": 0.01187, + "grad_norm": 0.6703031442863907, + "learning_rate": 0.003, + "loss": 4.33, + "step": 1187 + }, + { + "epoch": 0.01188, + "grad_norm": 0.6597409636847196, + "learning_rate": 0.003, + "loss": 4.3387, + "step": 1188 + }, + { + "epoch": 0.01189, + "grad_norm": 0.7528540145175824, + "learning_rate": 0.003, + "loss": 4.3431, + "step": 1189 + }, + { + "epoch": 0.0119, + "grad_norm": 0.7990197433249405, + "learning_rate": 0.003, + "loss": 4.3273, + "step": 1190 + }, + { + "epoch": 0.01191, + "grad_norm": 0.8012185403960035, + "learning_rate": 0.003, + "loss": 4.3081, + "step": 1191 + }, + { + "epoch": 0.01192, + "grad_norm": 0.7571323847445259, + "learning_rate": 0.003, + "loss": 4.3381, + "step": 1192 + }, + { + "epoch": 0.01193, + "grad_norm": 0.6624809456477357, + "learning_rate": 0.003, + "loss": 4.3371, + "step": 1193 + }, + { + "epoch": 0.01194, + "grad_norm": 0.7303795208901569, + "learning_rate": 0.003, + "loss": 4.3395, + "step": 1194 + }, + { + "epoch": 0.01195, + "grad_norm": 0.7841704264970735, + "learning_rate": 0.003, + "loss": 4.3211, + "step": 1195 + }, + { + "epoch": 0.01196, + "grad_norm": 0.8218420512118195, + "learning_rate": 0.003, + "loss": 4.3286, + "step": 1196 + }, + { + "epoch": 0.01197, + "grad_norm": 0.6346054129298183, + "learning_rate": 0.003, + "loss": 4.2874, + "step": 1197 + }, + { + "epoch": 0.01198, + "grad_norm": 0.648198494061273, + "learning_rate": 0.003, + "loss": 4.3575, + "step": 1198 + }, + { + "epoch": 0.01199, + "grad_norm": 0.719551428234481, + "learning_rate": 0.003, + "loss": 4.3378, + "step": 1199 + }, + { + "epoch": 0.012, + "grad_norm": 0.7355036739818444, + "learning_rate": 0.003, + "loss": 4.3295, + "step": 1200 + }, + { + "epoch": 0.01201, + "grad_norm": 0.9984238489578499, + "learning_rate": 0.003, + "loss": 4.3657, + "step": 1201 + }, + { + "epoch": 0.01202, + "grad_norm": 1.2422658156142912, + "learning_rate": 0.003, + "loss": 4.3719, + "step": 1202 + }, + { + "epoch": 0.01203, + "grad_norm": 0.7461879644958674, + "learning_rate": 0.003, + "loss": 4.3511, + "step": 1203 + }, + { + "epoch": 0.01204, + "grad_norm": 0.8846740490522026, + "learning_rate": 0.003, + "loss": 4.3624, + "step": 1204 + }, + { + "epoch": 0.01205, + "grad_norm": 0.7423867106895455, + "learning_rate": 0.003, + "loss": 4.3403, + "step": 1205 + }, + { + "epoch": 0.01206, + "grad_norm": 0.7502877844645354, + "learning_rate": 0.003, + "loss": 4.3858, + "step": 1206 + }, + { + "epoch": 0.01207, + "grad_norm": 0.8010490453321524, + "learning_rate": 0.003, + "loss": 4.3053, + "step": 1207 + }, + { + "epoch": 0.01208, + "grad_norm": 0.7863992142209323, + "learning_rate": 0.003, + "loss": 4.3665, + "step": 1208 + }, + { + "epoch": 0.01209, + "grad_norm": 0.6936314129825434, + "learning_rate": 0.003, + "loss": 4.3419, + "step": 1209 + }, + { + "epoch": 0.0121, + "grad_norm": 0.5834058626530795, + "learning_rate": 0.003, + "loss": 4.3425, + "step": 1210 + }, + { + "epoch": 0.01211, + "grad_norm": 0.5442454390371272, + "learning_rate": 0.003, + "loss": 4.3248, + "step": 1211 + }, + { + "epoch": 0.01212, + "grad_norm": 0.5640531404426855, + "learning_rate": 0.003, + "loss": 4.3485, + "step": 1212 + }, + { + "epoch": 0.01213, + "grad_norm": 0.4686052961934471, + "learning_rate": 0.003, + "loss": 4.3204, + "step": 1213 + }, + { + "epoch": 0.01214, + "grad_norm": 0.4623933928179833, + "learning_rate": 0.003, + "loss": 4.3023, + "step": 1214 + }, + { + "epoch": 0.01215, + "grad_norm": 0.390314877629531, + "learning_rate": 0.003, + "loss": 4.3075, + "step": 1215 + }, + { + "epoch": 0.01216, + "grad_norm": 0.3573034139707242, + "learning_rate": 0.003, + "loss": 4.2995, + "step": 1216 + }, + { + "epoch": 0.01217, + "grad_norm": 0.38593293619021, + "learning_rate": 0.003, + "loss": 4.3063, + "step": 1217 + }, + { + "epoch": 0.01218, + "grad_norm": 0.40705674810377235, + "learning_rate": 0.003, + "loss": 4.3006, + "step": 1218 + }, + { + "epoch": 0.01219, + "grad_norm": 0.45251062517773044, + "learning_rate": 0.003, + "loss": 4.3, + "step": 1219 + }, + { + "epoch": 0.0122, + "grad_norm": 0.5627372695280709, + "learning_rate": 0.003, + "loss": 4.2943, + "step": 1220 + }, + { + "epoch": 0.01221, + "grad_norm": 0.6302297504421508, + "learning_rate": 0.003, + "loss": 4.3122, + "step": 1221 + }, + { + "epoch": 0.01222, + "grad_norm": 0.695146633094906, + "learning_rate": 0.003, + "loss": 4.2975, + "step": 1222 + }, + { + "epoch": 0.01223, + "grad_norm": 0.6745684022343855, + "learning_rate": 0.003, + "loss": 4.2965, + "step": 1223 + }, + { + "epoch": 0.01224, + "grad_norm": 0.6275537412165774, + "learning_rate": 0.003, + "loss": 4.2794, + "step": 1224 + }, + { + "epoch": 0.01225, + "grad_norm": 0.732873035624088, + "learning_rate": 0.003, + "loss": 4.3084, + "step": 1225 + }, + { + "epoch": 0.01226, + "grad_norm": 0.8536265776058696, + "learning_rate": 0.003, + "loss": 4.3061, + "step": 1226 + }, + { + "epoch": 0.01227, + "grad_norm": 0.7710687106511338, + "learning_rate": 0.003, + "loss": 4.3371, + "step": 1227 + }, + { + "epoch": 0.01228, + "grad_norm": 0.6158196915186589, + "learning_rate": 0.003, + "loss": 4.3122, + "step": 1228 + }, + { + "epoch": 0.01229, + "grad_norm": 0.5647355795097493, + "learning_rate": 0.003, + "loss": 4.2952, + "step": 1229 + }, + { + "epoch": 0.0123, + "grad_norm": 0.5989687141441427, + "learning_rate": 0.003, + "loss": 4.3304, + "step": 1230 + }, + { + "epoch": 0.01231, + "grad_norm": 0.6070784312189734, + "learning_rate": 0.003, + "loss": 4.2993, + "step": 1231 + }, + { + "epoch": 0.01232, + "grad_norm": 0.6098189672837228, + "learning_rate": 0.003, + "loss": 4.276, + "step": 1232 + }, + { + "epoch": 0.01233, + "grad_norm": 0.6187452049200426, + "learning_rate": 0.003, + "loss": 4.2843, + "step": 1233 + }, + { + "epoch": 0.01234, + "grad_norm": 0.6847326322735632, + "learning_rate": 0.003, + "loss": 4.3016, + "step": 1234 + }, + { + "epoch": 0.01235, + "grad_norm": 0.7347950056629514, + "learning_rate": 0.003, + "loss": 4.2975, + "step": 1235 + }, + { + "epoch": 0.01236, + "grad_norm": 0.8404243381213906, + "learning_rate": 0.003, + "loss": 4.3037, + "step": 1236 + }, + { + "epoch": 0.01237, + "grad_norm": 1.1666025530795177, + "learning_rate": 0.003, + "loss": 4.3338, + "step": 1237 + }, + { + "epoch": 0.01238, + "grad_norm": 0.7706517401085584, + "learning_rate": 0.003, + "loss": 4.3184, + "step": 1238 + }, + { + "epoch": 0.01239, + "grad_norm": 0.8892529662707559, + "learning_rate": 0.003, + "loss": 4.3055, + "step": 1239 + }, + { + "epoch": 0.0124, + "grad_norm": 0.9665932126023419, + "learning_rate": 0.003, + "loss": 4.3004, + "step": 1240 + }, + { + "epoch": 0.01241, + "grad_norm": 1.0935280340663984, + "learning_rate": 0.003, + "loss": 4.3233, + "step": 1241 + }, + { + "epoch": 0.01242, + "grad_norm": 1.117726648141726, + "learning_rate": 0.003, + "loss": 4.3312, + "step": 1242 + }, + { + "epoch": 0.01243, + "grad_norm": 1.0553809973617114, + "learning_rate": 0.003, + "loss": 4.3181, + "step": 1243 + }, + { + "epoch": 0.01244, + "grad_norm": 0.9192132297518948, + "learning_rate": 0.003, + "loss": 4.3267, + "step": 1244 + }, + { + "epoch": 0.01245, + "grad_norm": 0.9382654947912115, + "learning_rate": 0.003, + "loss": 4.3187, + "step": 1245 + }, + { + "epoch": 0.01246, + "grad_norm": 1.0892504099540619, + "learning_rate": 0.003, + "loss": 4.3062, + "step": 1246 + }, + { + "epoch": 0.01247, + "grad_norm": 0.8993491326759976, + "learning_rate": 0.003, + "loss": 4.3511, + "step": 1247 + }, + { + "epoch": 0.01248, + "grad_norm": 0.908268119830267, + "learning_rate": 0.003, + "loss": 4.3379, + "step": 1248 + }, + { + "epoch": 0.01249, + "grad_norm": 0.8800131758337482, + "learning_rate": 0.003, + "loss": 4.3278, + "step": 1249 + }, + { + "epoch": 0.0125, + "grad_norm": 0.8337196500160362, + "learning_rate": 0.003, + "loss": 4.3052, + "step": 1250 + }, + { + "epoch": 0.01251, + "grad_norm": 0.9249131658343839, + "learning_rate": 0.003, + "loss": 4.3424, + "step": 1251 + }, + { + "epoch": 0.01252, + "grad_norm": 0.8524650338611843, + "learning_rate": 0.003, + "loss": 4.3172, + "step": 1252 + }, + { + "epoch": 0.01253, + "grad_norm": 0.7163285364378025, + "learning_rate": 0.003, + "loss": 4.2931, + "step": 1253 + }, + { + "epoch": 0.01254, + "grad_norm": 0.6991567907525832, + "learning_rate": 0.003, + "loss": 4.3341, + "step": 1254 + }, + { + "epoch": 0.01255, + "grad_norm": 0.6949992430981787, + "learning_rate": 0.003, + "loss": 4.2856, + "step": 1255 + }, + { + "epoch": 0.01256, + "grad_norm": 0.7100922258125638, + "learning_rate": 0.003, + "loss": 4.3118, + "step": 1256 + }, + { + "epoch": 0.01257, + "grad_norm": 0.6612217096643498, + "learning_rate": 0.003, + "loss": 4.2941, + "step": 1257 + }, + { + "epoch": 0.01258, + "grad_norm": 0.7747579760551054, + "learning_rate": 0.003, + "loss": 4.3107, + "step": 1258 + }, + { + "epoch": 0.01259, + "grad_norm": 1.124107981669201, + "learning_rate": 0.003, + "loss": 4.3222, + "step": 1259 + }, + { + "epoch": 0.0126, + "grad_norm": 0.9528743820284102, + "learning_rate": 0.003, + "loss": 4.3627, + "step": 1260 + }, + { + "epoch": 0.01261, + "grad_norm": 0.842570464638209, + "learning_rate": 0.003, + "loss": 4.3486, + "step": 1261 + }, + { + "epoch": 0.01262, + "grad_norm": 0.8393239407149024, + "learning_rate": 0.003, + "loss": 4.3397, + "step": 1262 + }, + { + "epoch": 0.01263, + "grad_norm": 0.8299261180456607, + "learning_rate": 0.003, + "loss": 4.3362, + "step": 1263 + }, + { + "epoch": 0.01264, + "grad_norm": 0.8141230470068299, + "learning_rate": 0.003, + "loss": 4.3187, + "step": 1264 + }, + { + "epoch": 0.01265, + "grad_norm": 1.0141550421015684, + "learning_rate": 0.003, + "loss": 4.3395, + "step": 1265 + }, + { + "epoch": 0.01266, + "grad_norm": 0.7612600647582426, + "learning_rate": 0.003, + "loss": 4.3297, + "step": 1266 + }, + { + "epoch": 0.01267, + "grad_norm": 0.6187340841665547, + "learning_rate": 0.003, + "loss": 4.3447, + "step": 1267 + }, + { + "epoch": 0.01268, + "grad_norm": 0.5822458314315527, + "learning_rate": 0.003, + "loss": 4.3275, + "step": 1268 + }, + { + "epoch": 0.01269, + "grad_norm": 0.4492593229301203, + "learning_rate": 0.003, + "loss": 4.2855, + "step": 1269 + }, + { + "epoch": 0.0127, + "grad_norm": 0.5016783253235925, + "learning_rate": 0.003, + "loss": 4.3213, + "step": 1270 + }, + { + "epoch": 0.01271, + "grad_norm": 0.44977006545763915, + "learning_rate": 0.003, + "loss": 4.297, + "step": 1271 + }, + { + "epoch": 0.01272, + "grad_norm": 0.43028280089396154, + "learning_rate": 0.003, + "loss": 4.291, + "step": 1272 + }, + { + "epoch": 0.01273, + "grad_norm": 0.4520162301705406, + "learning_rate": 0.003, + "loss": 4.3158, + "step": 1273 + }, + { + "epoch": 0.01274, + "grad_norm": 0.451409910695501, + "learning_rate": 0.003, + "loss": 4.3063, + "step": 1274 + }, + { + "epoch": 0.01275, + "grad_norm": 0.4633201534438662, + "learning_rate": 0.003, + "loss": 4.2756, + "step": 1275 + }, + { + "epoch": 0.01276, + "grad_norm": 0.4818283038515168, + "learning_rate": 0.003, + "loss": 4.2804, + "step": 1276 + }, + { + "epoch": 0.01277, + "grad_norm": 0.5612989347677225, + "learning_rate": 0.003, + "loss": 4.2895, + "step": 1277 + }, + { + "epoch": 0.01278, + "grad_norm": 0.6134565833773333, + "learning_rate": 0.003, + "loss": 4.2966, + "step": 1278 + }, + { + "epoch": 0.01279, + "grad_norm": 0.548830310802857, + "learning_rate": 0.003, + "loss": 4.2778, + "step": 1279 + }, + { + "epoch": 0.0128, + "grad_norm": 0.5192954436192417, + "learning_rate": 0.003, + "loss": 4.2736, + "step": 1280 + }, + { + "epoch": 0.01281, + "grad_norm": 0.7079203023923951, + "learning_rate": 0.003, + "loss": 4.2645, + "step": 1281 + }, + { + "epoch": 0.01282, + "grad_norm": 0.8640412732426497, + "learning_rate": 0.003, + "loss": 4.3216, + "step": 1282 + }, + { + "epoch": 0.01283, + "grad_norm": 0.7685939231737142, + "learning_rate": 0.003, + "loss": 4.2859, + "step": 1283 + }, + { + "epoch": 0.01284, + "grad_norm": 0.6000861053702143, + "learning_rate": 0.003, + "loss": 4.2792, + "step": 1284 + }, + { + "epoch": 0.01285, + "grad_norm": 0.7254306454469097, + "learning_rate": 0.003, + "loss": 4.2881, + "step": 1285 + }, + { + "epoch": 0.01286, + "grad_norm": 0.6733987067909273, + "learning_rate": 0.003, + "loss": 4.3069, + "step": 1286 + }, + { + "epoch": 0.01287, + "grad_norm": 0.6347430576087306, + "learning_rate": 0.003, + "loss": 4.2706, + "step": 1287 + }, + { + "epoch": 0.01288, + "grad_norm": 0.6742039696219508, + "learning_rate": 0.003, + "loss": 4.2877, + "step": 1288 + }, + { + "epoch": 0.01289, + "grad_norm": 0.7964754563506287, + "learning_rate": 0.003, + "loss": 4.2851, + "step": 1289 + }, + { + "epoch": 0.0129, + "grad_norm": 0.7466700913788555, + "learning_rate": 0.003, + "loss": 4.3108, + "step": 1290 + }, + { + "epoch": 0.01291, + "grad_norm": 0.7257835946382555, + "learning_rate": 0.003, + "loss": 4.2878, + "step": 1291 + }, + { + "epoch": 0.01292, + "grad_norm": 0.7356576793031216, + "learning_rate": 0.003, + "loss": 4.3222, + "step": 1292 + }, + { + "epoch": 0.01293, + "grad_norm": 0.7652258994712906, + "learning_rate": 0.003, + "loss": 4.312, + "step": 1293 + }, + { + "epoch": 0.01294, + "grad_norm": 0.8074663392977137, + "learning_rate": 0.003, + "loss": 4.2851, + "step": 1294 + }, + { + "epoch": 0.01295, + "grad_norm": 0.8419246282427761, + "learning_rate": 0.003, + "loss": 4.3171, + "step": 1295 + }, + { + "epoch": 0.01296, + "grad_norm": 0.9067886743130651, + "learning_rate": 0.003, + "loss": 4.2936, + "step": 1296 + }, + { + "epoch": 0.01297, + "grad_norm": 0.9867944428505203, + "learning_rate": 0.003, + "loss": 4.3168, + "step": 1297 + }, + { + "epoch": 0.01298, + "grad_norm": 0.9008581643124973, + "learning_rate": 0.003, + "loss": 4.2927, + "step": 1298 + }, + { + "epoch": 0.01299, + "grad_norm": 0.9659427448976995, + "learning_rate": 0.003, + "loss": 4.3085, + "step": 1299 + }, + { + "epoch": 0.013, + "grad_norm": 1.004138764919463, + "learning_rate": 0.003, + "loss": 4.3011, + "step": 1300 + }, + { + "epoch": 0.01301, + "grad_norm": 0.9518499756671687, + "learning_rate": 0.003, + "loss": 4.3191, + "step": 1301 + }, + { + "epoch": 0.01302, + "grad_norm": 0.8711824338075493, + "learning_rate": 0.003, + "loss": 4.361, + "step": 1302 + }, + { + "epoch": 0.01303, + "grad_norm": 0.8676764239234932, + "learning_rate": 0.003, + "loss": 4.3063, + "step": 1303 + }, + { + "epoch": 0.01304, + "grad_norm": 0.7125360490503003, + "learning_rate": 0.003, + "loss": 4.3284, + "step": 1304 + }, + { + "epoch": 0.01305, + "grad_norm": 0.6046564741110017, + "learning_rate": 0.003, + "loss": 4.3089, + "step": 1305 + }, + { + "epoch": 0.01306, + "grad_norm": 0.5647927538938984, + "learning_rate": 0.003, + "loss": 4.2992, + "step": 1306 + }, + { + "epoch": 0.01307, + "grad_norm": 0.5913576309809552, + "learning_rate": 0.003, + "loss": 4.3149, + "step": 1307 + }, + { + "epoch": 0.01308, + "grad_norm": 0.5924251650199668, + "learning_rate": 0.003, + "loss": 4.2932, + "step": 1308 + }, + { + "epoch": 0.01309, + "grad_norm": 0.507504674715556, + "learning_rate": 0.003, + "loss": 4.326, + "step": 1309 + }, + { + "epoch": 0.0131, + "grad_norm": 0.6387323300673942, + "learning_rate": 0.003, + "loss": 4.2756, + "step": 1310 + }, + { + "epoch": 0.01311, + "grad_norm": 0.7281996988052075, + "learning_rate": 0.003, + "loss": 4.3202, + "step": 1311 + }, + { + "epoch": 0.01312, + "grad_norm": 0.8834351506579509, + "learning_rate": 0.003, + "loss": 4.2908, + "step": 1312 + }, + { + "epoch": 0.01313, + "grad_norm": 0.840185011015699, + "learning_rate": 0.003, + "loss": 4.2898, + "step": 1313 + }, + { + "epoch": 0.01314, + "grad_norm": 0.6830490768340751, + "learning_rate": 0.003, + "loss": 4.2745, + "step": 1314 + }, + { + "epoch": 0.01315, + "grad_norm": 0.6358981723627178, + "learning_rate": 0.003, + "loss": 4.2992, + "step": 1315 + }, + { + "epoch": 0.01316, + "grad_norm": 0.6007791887068558, + "learning_rate": 0.003, + "loss": 4.2829, + "step": 1316 + }, + { + "epoch": 0.01317, + "grad_norm": 0.5048125779943076, + "learning_rate": 0.003, + "loss": 4.3051, + "step": 1317 + }, + { + "epoch": 0.01318, + "grad_norm": 0.49450675942854344, + "learning_rate": 0.003, + "loss": 4.2837, + "step": 1318 + }, + { + "epoch": 0.01319, + "grad_norm": 0.5258905508186564, + "learning_rate": 0.003, + "loss": 4.292, + "step": 1319 + }, + { + "epoch": 0.0132, + "grad_norm": 0.5585710036479011, + "learning_rate": 0.003, + "loss": 4.2888, + "step": 1320 + }, + { + "epoch": 0.01321, + "grad_norm": 0.5365220915283813, + "learning_rate": 0.003, + "loss": 4.2608, + "step": 1321 + }, + { + "epoch": 0.01322, + "grad_norm": 0.5557887511106342, + "learning_rate": 0.003, + "loss": 4.292, + "step": 1322 + }, + { + "epoch": 0.01323, + "grad_norm": 0.5531377716759448, + "learning_rate": 0.003, + "loss": 4.2837, + "step": 1323 + }, + { + "epoch": 0.01324, + "grad_norm": 0.48156733948461256, + "learning_rate": 0.003, + "loss": 4.2672, + "step": 1324 + }, + { + "epoch": 0.01325, + "grad_norm": 0.4219074369900267, + "learning_rate": 0.003, + "loss": 4.2693, + "step": 1325 + }, + { + "epoch": 0.01326, + "grad_norm": 0.43890778305013683, + "learning_rate": 0.003, + "loss": 4.2559, + "step": 1326 + }, + { + "epoch": 0.01327, + "grad_norm": 0.4164892745664214, + "learning_rate": 0.003, + "loss": 4.2724, + "step": 1327 + }, + { + "epoch": 0.01328, + "grad_norm": 0.49229273243603294, + "learning_rate": 0.003, + "loss": 4.2646, + "step": 1328 + }, + { + "epoch": 0.01329, + "grad_norm": 0.6121730970661665, + "learning_rate": 0.003, + "loss": 4.2329, + "step": 1329 + }, + { + "epoch": 0.0133, + "grad_norm": 0.8848354653817555, + "learning_rate": 0.003, + "loss": 4.282, + "step": 1330 + }, + { + "epoch": 0.01331, + "grad_norm": 1.0334105334221095, + "learning_rate": 0.003, + "loss": 4.2945, + "step": 1331 + }, + { + "epoch": 0.01332, + "grad_norm": 0.8234894915328277, + "learning_rate": 0.003, + "loss": 4.2994, + "step": 1332 + }, + { + "epoch": 0.01333, + "grad_norm": 0.7856986790322864, + "learning_rate": 0.003, + "loss": 4.2716, + "step": 1333 + }, + { + "epoch": 0.01334, + "grad_norm": 0.7455452553988269, + "learning_rate": 0.003, + "loss": 4.2662, + "step": 1334 + }, + { + "epoch": 0.01335, + "grad_norm": 0.623751194689146, + "learning_rate": 0.003, + "loss": 4.2787, + "step": 1335 + }, + { + "epoch": 0.01336, + "grad_norm": 0.8238963440270202, + "learning_rate": 0.003, + "loss": 4.2839, + "step": 1336 + }, + { + "epoch": 0.01337, + "grad_norm": 0.8159528784128957, + "learning_rate": 0.003, + "loss": 4.3035, + "step": 1337 + }, + { + "epoch": 0.01338, + "grad_norm": 0.691075290250966, + "learning_rate": 0.003, + "loss": 4.2694, + "step": 1338 + }, + { + "epoch": 0.01339, + "grad_norm": 0.6872888519447673, + "learning_rate": 0.003, + "loss": 4.2893, + "step": 1339 + }, + { + "epoch": 0.0134, + "grad_norm": 0.6938872016813308, + "learning_rate": 0.003, + "loss": 4.2701, + "step": 1340 + }, + { + "epoch": 0.01341, + "grad_norm": 0.7051901323106005, + "learning_rate": 0.003, + "loss": 4.2883, + "step": 1341 + }, + { + "epoch": 0.01342, + "grad_norm": 0.7365511131614688, + "learning_rate": 0.003, + "loss": 4.257, + "step": 1342 + }, + { + "epoch": 0.01343, + "grad_norm": 0.7564141855011662, + "learning_rate": 0.003, + "loss": 4.2701, + "step": 1343 + }, + { + "epoch": 0.01344, + "grad_norm": 0.695387503279461, + "learning_rate": 0.003, + "loss": 4.2649, + "step": 1344 + }, + { + "epoch": 0.01345, + "grad_norm": 0.6914588150610467, + "learning_rate": 0.003, + "loss": 4.3033, + "step": 1345 + }, + { + "epoch": 0.01346, + "grad_norm": 0.6768362060970508, + "learning_rate": 0.003, + "loss": 4.277, + "step": 1346 + }, + { + "epoch": 0.01347, + "grad_norm": 0.5492073753536512, + "learning_rate": 0.003, + "loss": 4.2633, + "step": 1347 + }, + { + "epoch": 0.01348, + "grad_norm": 0.5626283903224933, + "learning_rate": 0.003, + "loss": 4.2635, + "step": 1348 + }, + { + "epoch": 0.01349, + "grad_norm": 0.6438865489267382, + "learning_rate": 0.003, + "loss": 4.2856, + "step": 1349 + }, + { + "epoch": 0.0135, + "grad_norm": 0.7199005579630849, + "learning_rate": 0.003, + "loss": 4.2796, + "step": 1350 + }, + { + "epoch": 0.01351, + "grad_norm": 0.5428932527008233, + "learning_rate": 0.003, + "loss": 4.3038, + "step": 1351 + }, + { + "epoch": 0.01352, + "grad_norm": 0.5284901745728212, + "learning_rate": 0.003, + "loss": 4.2779, + "step": 1352 + }, + { + "epoch": 0.01353, + "grad_norm": 0.5600878945602594, + "learning_rate": 0.003, + "loss": 4.2635, + "step": 1353 + }, + { + "epoch": 0.01354, + "grad_norm": 0.6391444612725596, + "learning_rate": 0.003, + "loss": 4.2525, + "step": 1354 + }, + { + "epoch": 0.01355, + "grad_norm": 0.5641606375474643, + "learning_rate": 0.003, + "loss": 4.2634, + "step": 1355 + }, + { + "epoch": 0.01356, + "grad_norm": 0.5080744885475224, + "learning_rate": 0.003, + "loss": 4.2545, + "step": 1356 + }, + { + "epoch": 0.01357, + "grad_norm": 0.5093639445431166, + "learning_rate": 0.003, + "loss": 4.2441, + "step": 1357 + }, + { + "epoch": 0.01358, + "grad_norm": 0.5049437394862782, + "learning_rate": 0.003, + "loss": 4.2383, + "step": 1358 + }, + { + "epoch": 0.01359, + "grad_norm": 0.534762392832106, + "learning_rate": 0.003, + "loss": 4.2789, + "step": 1359 + }, + { + "epoch": 0.0136, + "grad_norm": 0.5654562399234901, + "learning_rate": 0.003, + "loss": 4.2447, + "step": 1360 + }, + { + "epoch": 0.01361, + "grad_norm": 0.7911685516807686, + "learning_rate": 0.003, + "loss": 4.2666, + "step": 1361 + }, + { + "epoch": 0.01362, + "grad_norm": 1.133172158468294, + "learning_rate": 0.003, + "loss": 4.3073, + "step": 1362 + }, + { + "epoch": 0.01363, + "grad_norm": 0.9631930238400119, + "learning_rate": 0.003, + "loss": 4.2987, + "step": 1363 + }, + { + "epoch": 0.01364, + "grad_norm": 0.7443319461425727, + "learning_rate": 0.003, + "loss": 4.2839, + "step": 1364 + }, + { + "epoch": 0.01365, + "grad_norm": 0.7368707977522249, + "learning_rate": 0.003, + "loss": 4.3041, + "step": 1365 + }, + { + "epoch": 0.01366, + "grad_norm": 0.6506933778775009, + "learning_rate": 0.003, + "loss": 4.255, + "step": 1366 + }, + { + "epoch": 0.01367, + "grad_norm": 0.7425630695378723, + "learning_rate": 0.003, + "loss": 4.3072, + "step": 1367 + }, + { + "epoch": 0.01368, + "grad_norm": 0.7495522896004446, + "learning_rate": 0.003, + "loss": 4.2833, + "step": 1368 + }, + { + "epoch": 0.01369, + "grad_norm": 0.7542187487500857, + "learning_rate": 0.003, + "loss": 4.2721, + "step": 1369 + }, + { + "epoch": 0.0137, + "grad_norm": 0.9330809300322568, + "learning_rate": 0.003, + "loss": 4.3029, + "step": 1370 + }, + { + "epoch": 0.01371, + "grad_norm": 1.2015218016331504, + "learning_rate": 0.003, + "loss": 4.3156, + "step": 1371 + }, + { + "epoch": 0.01372, + "grad_norm": 1.0453049109497266, + "learning_rate": 0.003, + "loss": 4.2577, + "step": 1372 + }, + { + "epoch": 0.01373, + "grad_norm": 1.086666562724284, + "learning_rate": 0.003, + "loss": 4.308, + "step": 1373 + }, + { + "epoch": 0.01374, + "grad_norm": 1.0320592526365153, + "learning_rate": 0.003, + "loss": 4.3227, + "step": 1374 + }, + { + "epoch": 0.01375, + "grad_norm": 1.1967086046120219, + "learning_rate": 0.003, + "loss": 4.2884, + "step": 1375 + }, + { + "epoch": 0.01376, + "grad_norm": 0.8548157633990001, + "learning_rate": 0.003, + "loss": 4.307, + "step": 1376 + }, + { + "epoch": 0.01377, + "grad_norm": 0.9102072373918407, + "learning_rate": 0.003, + "loss": 4.2985, + "step": 1377 + }, + { + "epoch": 0.01378, + "grad_norm": 0.9843975361934545, + "learning_rate": 0.003, + "loss": 4.3358, + "step": 1378 + }, + { + "epoch": 0.01379, + "grad_norm": 1.0283723711953892, + "learning_rate": 0.003, + "loss": 4.3102, + "step": 1379 + }, + { + "epoch": 0.0138, + "grad_norm": 0.9006954176701093, + "learning_rate": 0.003, + "loss": 4.3045, + "step": 1380 + }, + { + "epoch": 0.01381, + "grad_norm": 1.123717795950056, + "learning_rate": 0.003, + "loss": 4.3217, + "step": 1381 + }, + { + "epoch": 0.01382, + "grad_norm": 1.0986902300743402, + "learning_rate": 0.003, + "loss": 4.3218, + "step": 1382 + }, + { + "epoch": 0.01383, + "grad_norm": 0.9410643097534779, + "learning_rate": 0.003, + "loss": 4.3113, + "step": 1383 + }, + { + "epoch": 0.01384, + "grad_norm": 0.8436673398281763, + "learning_rate": 0.003, + "loss": 4.3327, + "step": 1384 + }, + { + "epoch": 0.01385, + "grad_norm": 0.8422095588989059, + "learning_rate": 0.003, + "loss": 4.3233, + "step": 1385 + }, + { + "epoch": 0.01386, + "grad_norm": 0.9633211233231406, + "learning_rate": 0.003, + "loss": 4.3006, + "step": 1386 + }, + { + "epoch": 0.01387, + "grad_norm": 1.2716300561274656, + "learning_rate": 0.003, + "loss": 4.3052, + "step": 1387 + }, + { + "epoch": 0.01388, + "grad_norm": 0.786746450686237, + "learning_rate": 0.003, + "loss": 4.33, + "step": 1388 + }, + { + "epoch": 0.01389, + "grad_norm": 0.763957561776965, + "learning_rate": 0.003, + "loss": 4.3069, + "step": 1389 + }, + { + "epoch": 0.0139, + "grad_norm": 0.7874363277644975, + "learning_rate": 0.003, + "loss": 4.3041, + "step": 1390 + }, + { + "epoch": 0.01391, + "grad_norm": 0.7493071050038139, + "learning_rate": 0.003, + "loss": 4.3036, + "step": 1391 + }, + { + "epoch": 0.01392, + "grad_norm": 0.5930612778435612, + "learning_rate": 0.003, + "loss": 4.3076, + "step": 1392 + }, + { + "epoch": 0.01393, + "grad_norm": 0.5012167316836135, + "learning_rate": 0.003, + "loss": 4.2738, + "step": 1393 + }, + { + "epoch": 0.01394, + "grad_norm": 0.4948661937091383, + "learning_rate": 0.003, + "loss": 4.2568, + "step": 1394 + }, + { + "epoch": 0.01395, + "grad_norm": 0.4622888611133768, + "learning_rate": 0.003, + "loss": 4.2868, + "step": 1395 + }, + { + "epoch": 0.01396, + "grad_norm": 0.4205886993889684, + "learning_rate": 0.003, + "loss": 4.275, + "step": 1396 + }, + { + "epoch": 0.01397, + "grad_norm": 0.47767051412590744, + "learning_rate": 0.003, + "loss": 4.2862, + "step": 1397 + }, + { + "epoch": 0.01398, + "grad_norm": 0.417341308437963, + "learning_rate": 0.003, + "loss": 4.2647, + "step": 1398 + }, + { + "epoch": 0.01399, + "grad_norm": 0.38332601666132154, + "learning_rate": 0.003, + "loss": 4.2835, + "step": 1399 + }, + { + "epoch": 0.014, + "grad_norm": 0.49255586643940574, + "learning_rate": 0.003, + "loss": 4.2444, + "step": 1400 + }, + { + "epoch": 0.01401, + "grad_norm": 0.8450629843994065, + "learning_rate": 0.003, + "loss": 4.2739, + "step": 1401 + }, + { + "epoch": 0.01402, + "grad_norm": 1.0913061976723808, + "learning_rate": 0.003, + "loss": 4.3172, + "step": 1402 + }, + { + "epoch": 0.01403, + "grad_norm": 0.6222948640126749, + "learning_rate": 0.003, + "loss": 4.2634, + "step": 1403 + }, + { + "epoch": 0.01404, + "grad_norm": 0.6692390535634434, + "learning_rate": 0.003, + "loss": 4.291, + "step": 1404 + }, + { + "epoch": 0.01405, + "grad_norm": 0.531565118139576, + "learning_rate": 0.003, + "loss": 4.2997, + "step": 1405 + }, + { + "epoch": 0.01406, + "grad_norm": 0.4824477961295427, + "learning_rate": 0.003, + "loss": 4.2779, + "step": 1406 + }, + { + "epoch": 0.01407, + "grad_norm": 0.5759838031356613, + "learning_rate": 0.003, + "loss": 4.2942, + "step": 1407 + }, + { + "epoch": 0.01408, + "grad_norm": 0.4476558981995588, + "learning_rate": 0.003, + "loss": 4.245, + "step": 1408 + }, + { + "epoch": 0.01409, + "grad_norm": 0.43291671999910936, + "learning_rate": 0.003, + "loss": 4.2753, + "step": 1409 + }, + { + "epoch": 0.0141, + "grad_norm": 0.48354199581441126, + "learning_rate": 0.003, + "loss": 4.2565, + "step": 1410 + }, + { + "epoch": 0.01411, + "grad_norm": 0.4777579756136276, + "learning_rate": 0.003, + "loss": 4.2302, + "step": 1411 + }, + { + "epoch": 0.01412, + "grad_norm": 0.5224654831649898, + "learning_rate": 0.003, + "loss": 4.2693, + "step": 1412 + }, + { + "epoch": 0.01413, + "grad_norm": 0.5669597734894755, + "learning_rate": 0.003, + "loss": 4.2639, + "step": 1413 + }, + { + "epoch": 0.01414, + "grad_norm": 0.5645597477136786, + "learning_rate": 0.003, + "loss": 4.2453, + "step": 1414 + }, + { + "epoch": 0.01415, + "grad_norm": 0.5887759339552011, + "learning_rate": 0.003, + "loss": 4.238, + "step": 1415 + }, + { + "epoch": 0.01416, + "grad_norm": 0.6104838140849604, + "learning_rate": 0.003, + "loss": 4.2641, + "step": 1416 + }, + { + "epoch": 0.01417, + "grad_norm": 0.6908167320390044, + "learning_rate": 0.003, + "loss": 4.2559, + "step": 1417 + }, + { + "epoch": 0.01418, + "grad_norm": 0.8241127076876378, + "learning_rate": 0.003, + "loss": 4.28, + "step": 1418 + }, + { + "epoch": 0.01419, + "grad_norm": 0.8554302831681504, + "learning_rate": 0.003, + "loss": 4.2733, + "step": 1419 + }, + { + "epoch": 0.0142, + "grad_norm": 0.8153359619668534, + "learning_rate": 0.003, + "loss": 4.28, + "step": 1420 + }, + { + "epoch": 0.01421, + "grad_norm": 0.808004884443839, + "learning_rate": 0.003, + "loss": 4.2527, + "step": 1421 + }, + { + "epoch": 0.01422, + "grad_norm": 0.6302573298023686, + "learning_rate": 0.003, + "loss": 4.2516, + "step": 1422 + }, + { + "epoch": 0.01423, + "grad_norm": 0.6834582033971294, + "learning_rate": 0.003, + "loss": 4.2585, + "step": 1423 + }, + { + "epoch": 0.01424, + "grad_norm": 0.7665806417295189, + "learning_rate": 0.003, + "loss": 4.2761, + "step": 1424 + }, + { + "epoch": 0.01425, + "grad_norm": 0.8479056405573007, + "learning_rate": 0.003, + "loss": 4.2857, + "step": 1425 + }, + { + "epoch": 0.01426, + "grad_norm": 0.8307259541179206, + "learning_rate": 0.003, + "loss": 4.2636, + "step": 1426 + }, + { + "epoch": 0.01427, + "grad_norm": 0.7328418116871758, + "learning_rate": 0.003, + "loss": 4.257, + "step": 1427 + }, + { + "epoch": 0.01428, + "grad_norm": 0.834478644245489, + "learning_rate": 0.003, + "loss": 4.2596, + "step": 1428 + }, + { + "epoch": 0.01429, + "grad_norm": 0.8679311866430979, + "learning_rate": 0.003, + "loss": 4.2742, + "step": 1429 + }, + { + "epoch": 0.0143, + "grad_norm": 0.9567780417694972, + "learning_rate": 0.003, + "loss": 4.2902, + "step": 1430 + }, + { + "epoch": 0.01431, + "grad_norm": 0.99274156172833, + "learning_rate": 0.003, + "loss": 4.2901, + "step": 1431 + }, + { + "epoch": 0.01432, + "grad_norm": 0.9575505507003149, + "learning_rate": 0.003, + "loss": 4.2782, + "step": 1432 + }, + { + "epoch": 0.01433, + "grad_norm": 1.0120649680762432, + "learning_rate": 0.003, + "loss": 4.2802, + "step": 1433 + }, + { + "epoch": 0.01434, + "grad_norm": 0.9698542777423387, + "learning_rate": 0.003, + "loss": 4.2694, + "step": 1434 + }, + { + "epoch": 0.01435, + "grad_norm": 0.8592446572371273, + "learning_rate": 0.003, + "loss": 4.2745, + "step": 1435 + }, + { + "epoch": 0.01436, + "grad_norm": 0.7299491282894184, + "learning_rate": 0.003, + "loss": 4.2863, + "step": 1436 + }, + { + "epoch": 0.01437, + "grad_norm": 0.6523703034324407, + "learning_rate": 0.003, + "loss": 4.296, + "step": 1437 + }, + { + "epoch": 0.01438, + "grad_norm": 0.6201165515106615, + "learning_rate": 0.003, + "loss": 4.2864, + "step": 1438 + }, + { + "epoch": 0.01439, + "grad_norm": 0.6536973612494624, + "learning_rate": 0.003, + "loss": 4.25, + "step": 1439 + }, + { + "epoch": 0.0144, + "grad_norm": 0.729776581664813, + "learning_rate": 0.003, + "loss": 4.2648, + "step": 1440 + }, + { + "epoch": 0.01441, + "grad_norm": 0.7206829643898741, + "learning_rate": 0.003, + "loss": 4.2702, + "step": 1441 + }, + { + "epoch": 0.01442, + "grad_norm": 0.7419558696417659, + "learning_rate": 0.003, + "loss": 4.2561, + "step": 1442 + }, + { + "epoch": 0.01443, + "grad_norm": 0.8093095130841327, + "learning_rate": 0.003, + "loss": 4.3153, + "step": 1443 + }, + { + "epoch": 0.01444, + "grad_norm": 0.7933277975079275, + "learning_rate": 0.003, + "loss": 4.2825, + "step": 1444 + }, + { + "epoch": 0.01445, + "grad_norm": 0.7264821716906565, + "learning_rate": 0.003, + "loss": 4.2399, + "step": 1445 + }, + { + "epoch": 0.01446, + "grad_norm": 0.6240641235999153, + "learning_rate": 0.003, + "loss": 4.2506, + "step": 1446 + }, + { + "epoch": 0.01447, + "grad_norm": 0.5288921284639453, + "learning_rate": 0.003, + "loss": 4.256, + "step": 1447 + }, + { + "epoch": 0.01448, + "grad_norm": 0.534171003657093, + "learning_rate": 0.003, + "loss": 4.2664, + "step": 1448 + }, + { + "epoch": 0.01449, + "grad_norm": 0.504252276028137, + "learning_rate": 0.003, + "loss": 4.2565, + "step": 1449 + }, + { + "epoch": 0.0145, + "grad_norm": 0.44556324681710546, + "learning_rate": 0.003, + "loss": 4.27, + "step": 1450 + }, + { + "epoch": 0.01451, + "grad_norm": 0.4269835752600227, + "learning_rate": 0.003, + "loss": 4.2459, + "step": 1451 + }, + { + "epoch": 0.01452, + "grad_norm": 0.43297277393213274, + "learning_rate": 0.003, + "loss": 4.2457, + "step": 1452 + }, + { + "epoch": 0.01453, + "grad_norm": 0.48066055951097997, + "learning_rate": 0.003, + "loss": 4.2578, + "step": 1453 + }, + { + "epoch": 0.01454, + "grad_norm": 0.7307010489462921, + "learning_rate": 0.003, + "loss": 4.2308, + "step": 1454 + }, + { + "epoch": 0.01455, + "grad_norm": 1.061032754030789, + "learning_rate": 0.003, + "loss": 4.2921, + "step": 1455 + }, + { + "epoch": 0.01456, + "grad_norm": 0.8307341818819957, + "learning_rate": 0.003, + "loss": 4.2586, + "step": 1456 + }, + { + "epoch": 0.01457, + "grad_norm": 0.6035503218291812, + "learning_rate": 0.003, + "loss": 4.2382, + "step": 1457 + }, + { + "epoch": 0.01458, + "grad_norm": 0.6933692267109066, + "learning_rate": 0.003, + "loss": 4.242, + "step": 1458 + }, + { + "epoch": 0.01459, + "grad_norm": 0.6220443333798514, + "learning_rate": 0.003, + "loss": 4.2607, + "step": 1459 + }, + { + "epoch": 0.0146, + "grad_norm": 0.8143223428105976, + "learning_rate": 0.003, + "loss": 4.2715, + "step": 1460 + }, + { + "epoch": 0.01461, + "grad_norm": 0.7987144915177533, + "learning_rate": 0.003, + "loss": 4.2513, + "step": 1461 + }, + { + "epoch": 0.01462, + "grad_norm": 0.6408744129720689, + "learning_rate": 0.003, + "loss": 4.2512, + "step": 1462 + }, + { + "epoch": 0.01463, + "grad_norm": 0.7757864438293093, + "learning_rate": 0.003, + "loss": 4.2548, + "step": 1463 + }, + { + "epoch": 0.01464, + "grad_norm": 0.7699033474111219, + "learning_rate": 0.003, + "loss": 4.2388, + "step": 1464 + }, + { + "epoch": 0.01465, + "grad_norm": 0.5797229129062035, + "learning_rate": 0.003, + "loss": 4.2736, + "step": 1465 + }, + { + "epoch": 0.01466, + "grad_norm": 0.611021916593894, + "learning_rate": 0.003, + "loss": 4.238, + "step": 1466 + }, + { + "epoch": 0.01467, + "grad_norm": 0.7716393804810123, + "learning_rate": 0.003, + "loss": 4.2809, + "step": 1467 + }, + { + "epoch": 0.01468, + "grad_norm": 0.684598948135191, + "learning_rate": 0.003, + "loss": 4.2576, + "step": 1468 + }, + { + "epoch": 0.01469, + "grad_norm": 0.5983303592308452, + "learning_rate": 0.003, + "loss": 4.2537, + "step": 1469 + }, + { + "epoch": 0.0147, + "grad_norm": 0.5951720244159906, + "learning_rate": 0.003, + "loss": 4.246, + "step": 1470 + }, + { + "epoch": 0.01471, + "grad_norm": 0.595230494931454, + "learning_rate": 0.003, + "loss": 4.2579, + "step": 1471 + }, + { + "epoch": 0.01472, + "grad_norm": 0.5179540783713997, + "learning_rate": 0.003, + "loss": 4.2385, + "step": 1472 + }, + { + "epoch": 0.01473, + "grad_norm": 0.513255678190744, + "learning_rate": 0.003, + "loss": 4.2524, + "step": 1473 + }, + { + "epoch": 0.01474, + "grad_norm": 0.5312567488001211, + "learning_rate": 0.003, + "loss": 4.2368, + "step": 1474 + }, + { + "epoch": 0.01475, + "grad_norm": 0.5231277112786824, + "learning_rate": 0.003, + "loss": 4.2547, + "step": 1475 + }, + { + "epoch": 0.01476, + "grad_norm": 0.524915592276574, + "learning_rate": 0.003, + "loss": 4.2412, + "step": 1476 + }, + { + "epoch": 0.01477, + "grad_norm": 0.4868094804842935, + "learning_rate": 0.003, + "loss": 4.2453, + "step": 1477 + }, + { + "epoch": 0.01478, + "grad_norm": 0.46072622581307954, + "learning_rate": 0.003, + "loss": 4.2277, + "step": 1478 + }, + { + "epoch": 0.01479, + "grad_norm": 0.5672195126213446, + "learning_rate": 0.003, + "loss": 4.2625, + "step": 1479 + }, + { + "epoch": 0.0148, + "grad_norm": 0.7550630041249586, + "learning_rate": 0.003, + "loss": 4.2383, + "step": 1480 + }, + { + "epoch": 0.01481, + "grad_norm": 0.9561266542090932, + "learning_rate": 0.003, + "loss": 4.2523, + "step": 1481 + }, + { + "epoch": 0.01482, + "grad_norm": 1.4635954400671285, + "learning_rate": 0.003, + "loss": 4.2602, + "step": 1482 + }, + { + "epoch": 0.01483, + "grad_norm": 0.7350989176290839, + "learning_rate": 0.003, + "loss": 4.2401, + "step": 1483 + }, + { + "epoch": 0.01484, + "grad_norm": 0.687260607047077, + "learning_rate": 0.003, + "loss": 4.2906, + "step": 1484 + }, + { + "epoch": 0.01485, + "grad_norm": 0.6759306581993629, + "learning_rate": 0.003, + "loss": 4.2492, + "step": 1485 + }, + { + "epoch": 0.01486, + "grad_norm": 0.5738231891215967, + "learning_rate": 0.003, + "loss": 4.2438, + "step": 1486 + }, + { + "epoch": 0.01487, + "grad_norm": 0.67662740183017, + "learning_rate": 0.003, + "loss": 4.2486, + "step": 1487 + }, + { + "epoch": 0.01488, + "grad_norm": 0.6465587043004464, + "learning_rate": 0.003, + "loss": 4.2584, + "step": 1488 + }, + { + "epoch": 0.01489, + "grad_norm": 0.6278262825245318, + "learning_rate": 0.003, + "loss": 4.251, + "step": 1489 + }, + { + "epoch": 0.0149, + "grad_norm": 0.7125204242602325, + "learning_rate": 0.003, + "loss": 4.2472, + "step": 1490 + }, + { + "epoch": 0.01491, + "grad_norm": 0.739446677383364, + "learning_rate": 0.003, + "loss": 4.263, + "step": 1491 + }, + { + "epoch": 0.01492, + "grad_norm": 0.7351492446611294, + "learning_rate": 0.003, + "loss": 4.2298, + "step": 1492 + }, + { + "epoch": 0.01493, + "grad_norm": 0.7206112830630914, + "learning_rate": 0.003, + "loss": 4.2585, + "step": 1493 + }, + { + "epoch": 0.01494, + "grad_norm": 0.7389347083990496, + "learning_rate": 0.003, + "loss": 4.239, + "step": 1494 + }, + { + "epoch": 0.01495, + "grad_norm": 0.6924446817172788, + "learning_rate": 0.003, + "loss": 4.2476, + "step": 1495 + }, + { + "epoch": 0.01496, + "grad_norm": 0.6954181917083473, + "learning_rate": 0.003, + "loss": 4.2266, + "step": 1496 + }, + { + "epoch": 0.01497, + "grad_norm": 0.6994096236505176, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 1497 + }, + { + "epoch": 0.01498, + "grad_norm": 0.737027067879948, + "learning_rate": 0.003, + "loss": 4.2577, + "step": 1498 + }, + { + "epoch": 0.01499, + "grad_norm": 0.8018965005148498, + "learning_rate": 0.003, + "loss": 4.2828, + "step": 1499 + }, + { + "epoch": 0.015, + "grad_norm": 0.673322646768884, + "learning_rate": 0.003, + "loss": 4.2302, + "step": 1500 + }, + { + "epoch": 0.01501, + "grad_norm": 0.6899086847228731, + "learning_rate": 0.003, + "loss": 4.23, + "step": 1501 + }, + { + "epoch": 0.01502, + "grad_norm": 0.6591283488970741, + "learning_rate": 0.003, + "loss": 4.2708, + "step": 1502 + }, + { + "epoch": 0.01503, + "grad_norm": 0.6852371530334846, + "learning_rate": 0.003, + "loss": 4.2501, + "step": 1503 + }, + { + "epoch": 0.01504, + "grad_norm": 0.6823538206406574, + "learning_rate": 0.003, + "loss": 4.2505, + "step": 1504 + }, + { + "epoch": 0.01505, + "grad_norm": 0.7899988696399359, + "learning_rate": 0.003, + "loss": 4.2395, + "step": 1505 + }, + { + "epoch": 0.01506, + "grad_norm": 0.8794700649213404, + "learning_rate": 0.003, + "loss": 4.2637, + "step": 1506 + }, + { + "epoch": 0.01507, + "grad_norm": 0.8573682870616134, + "learning_rate": 0.003, + "loss": 4.2788, + "step": 1507 + }, + { + "epoch": 0.01508, + "grad_norm": 0.9379452075670608, + "learning_rate": 0.003, + "loss": 4.257, + "step": 1508 + }, + { + "epoch": 0.01509, + "grad_norm": 1.155120366213554, + "learning_rate": 0.003, + "loss": 4.2868, + "step": 1509 + }, + { + "epoch": 0.0151, + "grad_norm": 0.8564988533087747, + "learning_rate": 0.003, + "loss": 4.2523, + "step": 1510 + }, + { + "epoch": 0.01511, + "grad_norm": 0.7943851233402885, + "learning_rate": 0.003, + "loss": 4.242, + "step": 1511 + }, + { + "epoch": 0.01512, + "grad_norm": 0.7290925664752087, + "learning_rate": 0.003, + "loss": 4.2483, + "step": 1512 + }, + { + "epoch": 0.01513, + "grad_norm": 0.8302189630029514, + "learning_rate": 0.003, + "loss": 4.2538, + "step": 1513 + }, + { + "epoch": 0.01514, + "grad_norm": 0.9180439778149861, + "learning_rate": 0.003, + "loss": 4.2658, + "step": 1514 + }, + { + "epoch": 0.01515, + "grad_norm": 0.8252164561072897, + "learning_rate": 0.003, + "loss": 4.2572, + "step": 1515 + }, + { + "epoch": 0.01516, + "grad_norm": 0.7235677802672953, + "learning_rate": 0.003, + "loss": 4.2628, + "step": 1516 + }, + { + "epoch": 0.01517, + "grad_norm": 0.669771960126658, + "learning_rate": 0.003, + "loss": 4.2114, + "step": 1517 + }, + { + "epoch": 0.01518, + "grad_norm": 0.6555410571298043, + "learning_rate": 0.003, + "loss": 4.248, + "step": 1518 + }, + { + "epoch": 0.01519, + "grad_norm": 0.7596573642242733, + "learning_rate": 0.003, + "loss": 4.2822, + "step": 1519 + }, + { + "epoch": 0.0152, + "grad_norm": 0.8963421668718663, + "learning_rate": 0.003, + "loss": 4.2455, + "step": 1520 + }, + { + "epoch": 0.01521, + "grad_norm": 1.008660081317184, + "learning_rate": 0.003, + "loss": 4.2626, + "step": 1521 + }, + { + "epoch": 0.01522, + "grad_norm": 1.0941342292979577, + "learning_rate": 0.003, + "loss": 4.2769, + "step": 1522 + }, + { + "epoch": 0.01523, + "grad_norm": 0.8501402046319685, + "learning_rate": 0.003, + "loss": 4.2571, + "step": 1523 + }, + { + "epoch": 0.01524, + "grad_norm": 0.9113983493960105, + "learning_rate": 0.003, + "loss": 4.2967, + "step": 1524 + }, + { + "epoch": 0.01525, + "grad_norm": 0.7627206625135139, + "learning_rate": 0.003, + "loss": 4.2846, + "step": 1525 + }, + { + "epoch": 0.01526, + "grad_norm": 0.7118386576342678, + "learning_rate": 0.003, + "loss": 4.2733, + "step": 1526 + }, + { + "epoch": 0.01527, + "grad_norm": 0.6832379742298202, + "learning_rate": 0.003, + "loss": 4.2396, + "step": 1527 + }, + { + "epoch": 0.01528, + "grad_norm": 0.6558129881498832, + "learning_rate": 0.003, + "loss": 4.2558, + "step": 1528 + }, + { + "epoch": 0.01529, + "grad_norm": 0.653598932900876, + "learning_rate": 0.003, + "loss": 4.2506, + "step": 1529 + }, + { + "epoch": 0.0153, + "grad_norm": 0.6481863699832326, + "learning_rate": 0.003, + "loss": 4.2569, + "step": 1530 + }, + { + "epoch": 0.01531, + "grad_norm": 0.7137971422876086, + "learning_rate": 0.003, + "loss": 4.2516, + "step": 1531 + }, + { + "epoch": 0.01532, + "grad_norm": 0.7682370963055534, + "learning_rate": 0.003, + "loss": 4.232, + "step": 1532 + }, + { + "epoch": 0.01533, + "grad_norm": 0.6869067765486859, + "learning_rate": 0.003, + "loss": 4.2552, + "step": 1533 + }, + { + "epoch": 0.01534, + "grad_norm": 0.6763479705568691, + "learning_rate": 0.003, + "loss": 4.251, + "step": 1534 + }, + { + "epoch": 0.01535, + "grad_norm": 0.6278011204652878, + "learning_rate": 0.003, + "loss": 4.2624, + "step": 1535 + }, + { + "epoch": 0.01536, + "grad_norm": 0.5449726408675756, + "learning_rate": 0.003, + "loss": 4.2371, + "step": 1536 + }, + { + "epoch": 0.01537, + "grad_norm": 0.5702070777355899, + "learning_rate": 0.003, + "loss": 4.2348, + "step": 1537 + }, + { + "epoch": 0.01538, + "grad_norm": 0.5568148406111664, + "learning_rate": 0.003, + "loss": 4.2106, + "step": 1538 + }, + { + "epoch": 0.01539, + "grad_norm": 0.5990162223338897, + "learning_rate": 0.003, + "loss": 4.2411, + "step": 1539 + }, + { + "epoch": 0.0154, + "grad_norm": 0.6709932169117396, + "learning_rate": 0.003, + "loss": 4.2471, + "step": 1540 + }, + { + "epoch": 0.01541, + "grad_norm": 0.839479307713138, + "learning_rate": 0.003, + "loss": 4.2333, + "step": 1541 + }, + { + "epoch": 0.01542, + "grad_norm": 0.998609935659052, + "learning_rate": 0.003, + "loss": 4.2723, + "step": 1542 + }, + { + "epoch": 0.01543, + "grad_norm": 0.9175226186557726, + "learning_rate": 0.003, + "loss": 4.2597, + "step": 1543 + }, + { + "epoch": 0.01544, + "grad_norm": 0.9698336862541426, + "learning_rate": 0.003, + "loss": 4.2493, + "step": 1544 + }, + { + "epoch": 0.01545, + "grad_norm": 0.8646283897782253, + "learning_rate": 0.003, + "loss": 4.3133, + "step": 1545 + }, + { + "epoch": 0.01546, + "grad_norm": 0.7489694669979721, + "learning_rate": 0.003, + "loss": 4.2786, + "step": 1546 + }, + { + "epoch": 0.01547, + "grad_norm": 0.7346787237600204, + "learning_rate": 0.003, + "loss": 4.2793, + "step": 1547 + }, + { + "epoch": 0.01548, + "grad_norm": 0.7843295792698941, + "learning_rate": 0.003, + "loss": 4.2451, + "step": 1548 + }, + { + "epoch": 0.01549, + "grad_norm": 0.7547938162635378, + "learning_rate": 0.003, + "loss": 4.2323, + "step": 1549 + }, + { + "epoch": 0.0155, + "grad_norm": 0.8033101017591833, + "learning_rate": 0.003, + "loss": 4.2543, + "step": 1550 + }, + { + "epoch": 0.01551, + "grad_norm": 0.784324716245673, + "learning_rate": 0.003, + "loss": 4.2319, + "step": 1551 + }, + { + "epoch": 0.01552, + "grad_norm": 0.6675659235069369, + "learning_rate": 0.003, + "loss": 4.2485, + "step": 1552 + }, + { + "epoch": 0.01553, + "grad_norm": 0.7200584927165973, + "learning_rate": 0.003, + "loss": 4.2735, + "step": 1553 + }, + { + "epoch": 0.01554, + "grad_norm": 0.833889265994982, + "learning_rate": 0.003, + "loss": 4.259, + "step": 1554 + }, + { + "epoch": 0.01555, + "grad_norm": 0.7292713050790204, + "learning_rate": 0.003, + "loss": 4.2302, + "step": 1555 + }, + { + "epoch": 0.01556, + "grad_norm": 0.6334535318603711, + "learning_rate": 0.003, + "loss": 4.2647, + "step": 1556 + }, + { + "epoch": 0.01557, + "grad_norm": 0.6665206815174863, + "learning_rate": 0.003, + "loss": 4.2396, + "step": 1557 + }, + { + "epoch": 0.01558, + "grad_norm": 0.5467983416304218, + "learning_rate": 0.003, + "loss": 4.2264, + "step": 1558 + }, + { + "epoch": 0.01559, + "grad_norm": 0.5303338136727427, + "learning_rate": 0.003, + "loss": 4.2625, + "step": 1559 + }, + { + "epoch": 0.0156, + "grad_norm": 0.5964427896199266, + "learning_rate": 0.003, + "loss": 4.2526, + "step": 1560 + }, + { + "epoch": 0.01561, + "grad_norm": 0.7487394215839274, + "learning_rate": 0.003, + "loss": 4.2358, + "step": 1561 + }, + { + "epoch": 0.01562, + "grad_norm": 0.8937121416966828, + "learning_rate": 0.003, + "loss": 4.2394, + "step": 1562 + }, + { + "epoch": 0.01563, + "grad_norm": 1.2085214855734254, + "learning_rate": 0.003, + "loss": 4.2602, + "step": 1563 + }, + { + "epoch": 0.01564, + "grad_norm": 0.8696563758847599, + "learning_rate": 0.003, + "loss": 4.284, + "step": 1564 + }, + { + "epoch": 0.01565, + "grad_norm": 0.739740990652482, + "learning_rate": 0.003, + "loss": 4.2566, + "step": 1565 + }, + { + "epoch": 0.01566, + "grad_norm": 0.7600676271323985, + "learning_rate": 0.003, + "loss": 4.2738, + "step": 1566 + }, + { + "epoch": 0.01567, + "grad_norm": 0.7222900608043384, + "learning_rate": 0.003, + "loss": 4.2646, + "step": 1567 + }, + { + "epoch": 0.01568, + "grad_norm": 0.7569225436342659, + "learning_rate": 0.003, + "loss": 4.2466, + "step": 1568 + }, + { + "epoch": 0.01569, + "grad_norm": 0.7441561581750152, + "learning_rate": 0.003, + "loss": 4.252, + "step": 1569 + }, + { + "epoch": 0.0157, + "grad_norm": 0.7580218902280629, + "learning_rate": 0.003, + "loss": 4.2727, + "step": 1570 + }, + { + "epoch": 0.01571, + "grad_norm": 0.771634247408187, + "learning_rate": 0.003, + "loss": 4.2581, + "step": 1571 + }, + { + "epoch": 0.01572, + "grad_norm": 0.95958549182467, + "learning_rate": 0.003, + "loss": 4.2713, + "step": 1572 + }, + { + "epoch": 0.01573, + "grad_norm": 0.9029057415025205, + "learning_rate": 0.003, + "loss": 4.261, + "step": 1573 + }, + { + "epoch": 0.01574, + "grad_norm": 0.8044062509450788, + "learning_rate": 0.003, + "loss": 4.2692, + "step": 1574 + }, + { + "epoch": 0.01575, + "grad_norm": 0.7286639057606195, + "learning_rate": 0.003, + "loss": 4.2305, + "step": 1575 + }, + { + "epoch": 0.01576, + "grad_norm": 0.6821195872107877, + "learning_rate": 0.003, + "loss": 4.2492, + "step": 1576 + }, + { + "epoch": 0.01577, + "grad_norm": 0.6033878670810847, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 1577 + }, + { + "epoch": 0.01578, + "grad_norm": 0.519693589168086, + "learning_rate": 0.003, + "loss": 4.2343, + "step": 1578 + }, + { + "epoch": 0.01579, + "grad_norm": 0.46078437812161954, + "learning_rate": 0.003, + "loss": 4.2271, + "step": 1579 + }, + { + "epoch": 0.0158, + "grad_norm": 0.3811995998581025, + "learning_rate": 0.003, + "loss": 4.2344, + "step": 1580 + }, + { + "epoch": 0.01581, + "grad_norm": 0.40258363386782764, + "learning_rate": 0.003, + "loss": 4.2406, + "step": 1581 + }, + { + "epoch": 0.01582, + "grad_norm": 0.4070479328001984, + "learning_rate": 0.003, + "loss": 4.2222, + "step": 1582 + }, + { + "epoch": 0.01583, + "grad_norm": 0.39481584768858025, + "learning_rate": 0.003, + "loss": 4.2093, + "step": 1583 + }, + { + "epoch": 0.01584, + "grad_norm": 0.4212949759278112, + "learning_rate": 0.003, + "loss": 4.24, + "step": 1584 + }, + { + "epoch": 0.01585, + "grad_norm": 0.6375222350508811, + "learning_rate": 0.003, + "loss": 4.2351, + "step": 1585 + }, + { + "epoch": 0.01586, + "grad_norm": 1.1087267261978884, + "learning_rate": 0.003, + "loss": 4.2389, + "step": 1586 + }, + { + "epoch": 0.01587, + "grad_norm": 1.095427209738268, + "learning_rate": 0.003, + "loss": 4.2576, + "step": 1587 + }, + { + "epoch": 0.01588, + "grad_norm": 0.5974585636629688, + "learning_rate": 0.003, + "loss": 4.2673, + "step": 1588 + }, + { + "epoch": 0.01589, + "grad_norm": 0.6434065905263348, + "learning_rate": 0.003, + "loss": 4.2657, + "step": 1589 + }, + { + "epoch": 0.0159, + "grad_norm": 0.5922765275307262, + "learning_rate": 0.003, + "loss": 4.2507, + "step": 1590 + }, + { + "epoch": 0.01591, + "grad_norm": 0.6559622239667369, + "learning_rate": 0.003, + "loss": 4.2576, + "step": 1591 + }, + { + "epoch": 0.01592, + "grad_norm": 0.5876956364208827, + "learning_rate": 0.003, + "loss": 4.2446, + "step": 1592 + }, + { + "epoch": 0.01593, + "grad_norm": 0.4847713630300587, + "learning_rate": 0.003, + "loss": 4.2428, + "step": 1593 + }, + { + "epoch": 0.01594, + "grad_norm": 0.5605434129471829, + "learning_rate": 0.003, + "loss": 4.2313, + "step": 1594 + }, + { + "epoch": 0.01595, + "grad_norm": 0.5718187801259297, + "learning_rate": 0.003, + "loss": 4.2581, + "step": 1595 + }, + { + "epoch": 0.01596, + "grad_norm": 0.620870673073496, + "learning_rate": 0.003, + "loss": 4.2509, + "step": 1596 + }, + { + "epoch": 0.01597, + "grad_norm": 0.6969552121884817, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 1597 + }, + { + "epoch": 0.01598, + "grad_norm": 0.8094238052393485, + "learning_rate": 0.003, + "loss": 4.2172, + "step": 1598 + }, + { + "epoch": 0.01599, + "grad_norm": 0.7713054312041273, + "learning_rate": 0.003, + "loss": 4.2333, + "step": 1599 + }, + { + "epoch": 0.016, + "grad_norm": 0.7787996781195005, + "learning_rate": 0.003, + "loss": 4.2062, + "step": 1600 + }, + { + "epoch": 0.01601, + "grad_norm": 0.620719741277366, + "learning_rate": 0.003, + "loss": 4.2478, + "step": 1601 + }, + { + "epoch": 0.01602, + "grad_norm": 0.5965590372558249, + "learning_rate": 0.003, + "loss": 4.223, + "step": 1602 + }, + { + "epoch": 0.01603, + "grad_norm": 0.6057948188618382, + "learning_rate": 0.003, + "loss": 4.2386, + "step": 1603 + }, + { + "epoch": 0.01604, + "grad_norm": 0.5859806802733505, + "learning_rate": 0.003, + "loss": 4.2313, + "step": 1604 + }, + { + "epoch": 0.01605, + "grad_norm": 0.5985785202970987, + "learning_rate": 0.003, + "loss": 4.2093, + "step": 1605 + }, + { + "epoch": 0.01606, + "grad_norm": 0.6329923375985832, + "learning_rate": 0.003, + "loss": 4.2131, + "step": 1606 + }, + { + "epoch": 0.01607, + "grad_norm": 0.6777814987430973, + "learning_rate": 0.003, + "loss": 4.2418, + "step": 1607 + }, + { + "epoch": 0.01608, + "grad_norm": 0.6454825679602344, + "learning_rate": 0.003, + "loss": 4.2152, + "step": 1608 + }, + { + "epoch": 0.01609, + "grad_norm": 0.6234331303306465, + "learning_rate": 0.003, + "loss": 4.222, + "step": 1609 + }, + { + "epoch": 0.0161, + "grad_norm": 0.717661852780552, + "learning_rate": 0.003, + "loss": 4.1917, + "step": 1610 + }, + { + "epoch": 0.01611, + "grad_norm": 0.814514408732785, + "learning_rate": 0.003, + "loss": 4.2133, + "step": 1611 + }, + { + "epoch": 0.01612, + "grad_norm": 0.968378243767811, + "learning_rate": 0.003, + "loss": 4.2184, + "step": 1612 + }, + { + "epoch": 0.01613, + "grad_norm": 1.1889382213143287, + "learning_rate": 0.003, + "loss": 4.2747, + "step": 1613 + }, + { + "epoch": 0.01614, + "grad_norm": 0.902755449051569, + "learning_rate": 0.003, + "loss": 4.2894, + "step": 1614 + }, + { + "epoch": 0.01615, + "grad_norm": 0.8743919537144361, + "learning_rate": 0.003, + "loss": 4.2622, + "step": 1615 + }, + { + "epoch": 0.01616, + "grad_norm": 0.8601657492637806, + "learning_rate": 0.003, + "loss": 4.2396, + "step": 1616 + }, + { + "epoch": 0.01617, + "grad_norm": 0.8301577875616364, + "learning_rate": 0.003, + "loss": 4.2333, + "step": 1617 + }, + { + "epoch": 0.01618, + "grad_norm": 0.829332785699801, + "learning_rate": 0.003, + "loss": 4.2387, + "step": 1618 + }, + { + "epoch": 0.01619, + "grad_norm": 0.7047548182139096, + "learning_rate": 0.003, + "loss": 4.2209, + "step": 1619 + }, + { + "epoch": 0.0162, + "grad_norm": 0.7270289420377364, + "learning_rate": 0.003, + "loss": 4.2317, + "step": 1620 + }, + { + "epoch": 0.01621, + "grad_norm": 0.6929887184456018, + "learning_rate": 0.003, + "loss": 4.2257, + "step": 1621 + }, + { + "epoch": 0.01622, + "grad_norm": 0.7471865137769718, + "learning_rate": 0.003, + "loss": 4.2589, + "step": 1622 + }, + { + "epoch": 0.01623, + "grad_norm": 0.8282860007733814, + "learning_rate": 0.003, + "loss": 4.2399, + "step": 1623 + }, + { + "epoch": 0.01624, + "grad_norm": 0.8871911840665138, + "learning_rate": 0.003, + "loss": 4.2581, + "step": 1624 + }, + { + "epoch": 0.01625, + "grad_norm": 1.0259944995678087, + "learning_rate": 0.003, + "loss": 4.2469, + "step": 1625 + }, + { + "epoch": 0.01626, + "grad_norm": 0.937815191301713, + "learning_rate": 0.003, + "loss": 4.2602, + "step": 1626 + }, + { + "epoch": 0.01627, + "grad_norm": 1.064022068457424, + "learning_rate": 0.003, + "loss": 4.2368, + "step": 1627 + }, + { + "epoch": 0.01628, + "grad_norm": 1.0710330182336567, + "learning_rate": 0.003, + "loss": 4.2567, + "step": 1628 + }, + { + "epoch": 0.01629, + "grad_norm": 0.7877325862421115, + "learning_rate": 0.003, + "loss": 4.2414, + "step": 1629 + }, + { + "epoch": 0.0163, + "grad_norm": 0.6743454531988056, + "learning_rate": 0.003, + "loss": 4.2288, + "step": 1630 + }, + { + "epoch": 0.01631, + "grad_norm": 0.701073161255015, + "learning_rate": 0.003, + "loss": 4.2499, + "step": 1631 + }, + { + "epoch": 0.01632, + "grad_norm": 0.6004677978717051, + "learning_rate": 0.003, + "loss": 4.252, + "step": 1632 + }, + { + "epoch": 0.01633, + "grad_norm": 0.5772521307266697, + "learning_rate": 0.003, + "loss": 4.2255, + "step": 1633 + }, + { + "epoch": 0.01634, + "grad_norm": 0.5454153935480286, + "learning_rate": 0.003, + "loss": 4.2406, + "step": 1634 + }, + { + "epoch": 0.01635, + "grad_norm": 0.5154455969479952, + "learning_rate": 0.003, + "loss": 4.218, + "step": 1635 + }, + { + "epoch": 0.01636, + "grad_norm": 0.5267549110628605, + "learning_rate": 0.003, + "loss": 4.2348, + "step": 1636 + }, + { + "epoch": 0.01637, + "grad_norm": 0.5971697216641335, + "learning_rate": 0.003, + "loss": 4.2205, + "step": 1637 + }, + { + "epoch": 0.01638, + "grad_norm": 0.7211767188224439, + "learning_rate": 0.003, + "loss": 4.229, + "step": 1638 + }, + { + "epoch": 0.01639, + "grad_norm": 0.7816125211534669, + "learning_rate": 0.003, + "loss": 4.2548, + "step": 1639 + }, + { + "epoch": 0.0164, + "grad_norm": 0.7389027343982134, + "learning_rate": 0.003, + "loss": 4.2362, + "step": 1640 + }, + { + "epoch": 0.01641, + "grad_norm": 0.5975265072148448, + "learning_rate": 0.003, + "loss": 4.2026, + "step": 1641 + }, + { + "epoch": 0.01642, + "grad_norm": 0.7083476744478381, + "learning_rate": 0.003, + "loss": 4.2507, + "step": 1642 + }, + { + "epoch": 0.01643, + "grad_norm": 0.7791000270261257, + "learning_rate": 0.003, + "loss": 4.2378, + "step": 1643 + }, + { + "epoch": 0.01644, + "grad_norm": 0.9006394984546569, + "learning_rate": 0.003, + "loss": 4.2625, + "step": 1644 + }, + { + "epoch": 0.01645, + "grad_norm": 1.2334686487745912, + "learning_rate": 0.003, + "loss": 4.2236, + "step": 1645 + }, + { + "epoch": 0.01646, + "grad_norm": 0.9932721101220016, + "learning_rate": 0.003, + "loss": 4.263, + "step": 1646 + }, + { + "epoch": 0.01647, + "grad_norm": 0.989511986212608, + "learning_rate": 0.003, + "loss": 4.2576, + "step": 1647 + }, + { + "epoch": 0.01648, + "grad_norm": 0.8142750385583939, + "learning_rate": 0.003, + "loss": 4.23, + "step": 1648 + }, + { + "epoch": 0.01649, + "grad_norm": 0.6967651791258922, + "learning_rate": 0.003, + "loss": 4.2583, + "step": 1649 + }, + { + "epoch": 0.0165, + "grad_norm": 0.7348292229935104, + "learning_rate": 0.003, + "loss": 4.2361, + "step": 1650 + }, + { + "epoch": 0.01651, + "grad_norm": 0.6565673319405235, + "learning_rate": 0.003, + "loss": 4.2402, + "step": 1651 + }, + { + "epoch": 0.01652, + "grad_norm": 0.5674400404701213, + "learning_rate": 0.003, + "loss": 4.1936, + "step": 1652 + }, + { + "epoch": 0.01653, + "grad_norm": 0.5571261023539824, + "learning_rate": 0.003, + "loss": 4.2378, + "step": 1653 + }, + { + "epoch": 0.01654, + "grad_norm": 0.5083863828402352, + "learning_rate": 0.003, + "loss": 4.2229, + "step": 1654 + }, + { + "epoch": 0.01655, + "grad_norm": 0.49820131551804164, + "learning_rate": 0.003, + "loss": 4.23, + "step": 1655 + }, + { + "epoch": 0.01656, + "grad_norm": 0.5201001189396881, + "learning_rate": 0.003, + "loss": 4.2012, + "step": 1656 + }, + { + "epoch": 0.01657, + "grad_norm": 0.6652455003065759, + "learning_rate": 0.003, + "loss": 4.2237, + "step": 1657 + }, + { + "epoch": 0.01658, + "grad_norm": 0.8870121051704852, + "learning_rate": 0.003, + "loss": 4.219, + "step": 1658 + }, + { + "epoch": 0.01659, + "grad_norm": 1.1634698909054708, + "learning_rate": 0.003, + "loss": 4.2361, + "step": 1659 + }, + { + "epoch": 0.0166, + "grad_norm": 0.6579478278546654, + "learning_rate": 0.003, + "loss": 4.2096, + "step": 1660 + }, + { + "epoch": 0.01661, + "grad_norm": 0.5917680182889076, + "learning_rate": 0.003, + "loss": 4.2092, + "step": 1661 + }, + { + "epoch": 0.01662, + "grad_norm": 0.7821000738049482, + "learning_rate": 0.003, + "loss": 4.2171, + "step": 1662 + }, + { + "epoch": 0.01663, + "grad_norm": 0.812312164693218, + "learning_rate": 0.003, + "loss": 4.2422, + "step": 1663 + }, + { + "epoch": 0.01664, + "grad_norm": 0.7053733462479915, + "learning_rate": 0.003, + "loss": 4.2567, + "step": 1664 + }, + { + "epoch": 0.01665, + "grad_norm": 0.713897124959619, + "learning_rate": 0.003, + "loss": 4.2198, + "step": 1665 + }, + { + "epoch": 0.01666, + "grad_norm": 0.6600299709203347, + "learning_rate": 0.003, + "loss": 4.2205, + "step": 1666 + }, + { + "epoch": 0.01667, + "grad_norm": 0.5832085680202569, + "learning_rate": 0.003, + "loss": 4.2321, + "step": 1667 + }, + { + "epoch": 0.01668, + "grad_norm": 0.6589799435966834, + "learning_rate": 0.003, + "loss": 4.2397, + "step": 1668 + }, + { + "epoch": 0.01669, + "grad_norm": 0.7719970182316414, + "learning_rate": 0.003, + "loss": 4.2273, + "step": 1669 + }, + { + "epoch": 0.0167, + "grad_norm": 0.7846205843605817, + "learning_rate": 0.003, + "loss": 4.2184, + "step": 1670 + }, + { + "epoch": 0.01671, + "grad_norm": 0.7635694428992538, + "learning_rate": 0.003, + "loss": 4.2087, + "step": 1671 + }, + { + "epoch": 0.01672, + "grad_norm": 0.7758449881926132, + "learning_rate": 0.003, + "loss": 4.2199, + "step": 1672 + }, + { + "epoch": 0.01673, + "grad_norm": 0.6618290355489082, + "learning_rate": 0.003, + "loss": 4.2134, + "step": 1673 + }, + { + "epoch": 0.01674, + "grad_norm": 0.6727466502726199, + "learning_rate": 0.003, + "loss": 4.232, + "step": 1674 + }, + { + "epoch": 0.01675, + "grad_norm": 0.6690360284678838, + "learning_rate": 0.003, + "loss": 4.2158, + "step": 1675 + }, + { + "epoch": 0.01676, + "grad_norm": 0.6672382539021496, + "learning_rate": 0.003, + "loss": 4.2242, + "step": 1676 + }, + { + "epoch": 0.01677, + "grad_norm": 0.6311361325143124, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 1677 + }, + { + "epoch": 0.01678, + "grad_norm": 0.727198962494881, + "learning_rate": 0.003, + "loss": 4.2299, + "step": 1678 + }, + { + "epoch": 0.01679, + "grad_norm": 0.8433802333808034, + "learning_rate": 0.003, + "loss": 4.2177, + "step": 1679 + }, + { + "epoch": 0.0168, + "grad_norm": 0.8402151525100402, + "learning_rate": 0.003, + "loss": 4.2352, + "step": 1680 + }, + { + "epoch": 0.01681, + "grad_norm": 0.9608253642789906, + "learning_rate": 0.003, + "loss": 4.2433, + "step": 1681 + }, + { + "epoch": 0.01682, + "grad_norm": 1.160024891342376, + "learning_rate": 0.003, + "loss": 4.246, + "step": 1682 + }, + { + "epoch": 0.01683, + "grad_norm": 0.8201452744681862, + "learning_rate": 0.003, + "loss": 4.2539, + "step": 1683 + }, + { + "epoch": 0.01684, + "grad_norm": 0.7434576192749681, + "learning_rate": 0.003, + "loss": 4.2242, + "step": 1684 + }, + { + "epoch": 0.01685, + "grad_norm": 0.8733167238677911, + "learning_rate": 0.003, + "loss": 4.2448, + "step": 1685 + }, + { + "epoch": 0.01686, + "grad_norm": 0.7251849705005492, + "learning_rate": 0.003, + "loss": 4.2342, + "step": 1686 + }, + { + "epoch": 0.01687, + "grad_norm": 0.623355393019925, + "learning_rate": 0.003, + "loss": 4.2258, + "step": 1687 + }, + { + "epoch": 0.01688, + "grad_norm": 0.6485004834811148, + "learning_rate": 0.003, + "loss": 4.2113, + "step": 1688 + }, + { + "epoch": 0.01689, + "grad_norm": 0.6622097332232696, + "learning_rate": 0.003, + "loss": 4.2246, + "step": 1689 + }, + { + "epoch": 0.0169, + "grad_norm": 0.5609920955895745, + "learning_rate": 0.003, + "loss": 4.2095, + "step": 1690 + }, + { + "epoch": 0.01691, + "grad_norm": 0.5554950872814335, + "learning_rate": 0.003, + "loss": 4.2208, + "step": 1691 + }, + { + "epoch": 0.01692, + "grad_norm": 0.578754263387868, + "learning_rate": 0.003, + "loss": 4.2223, + "step": 1692 + }, + { + "epoch": 0.01693, + "grad_norm": 0.54763155586202, + "learning_rate": 0.003, + "loss": 4.2453, + "step": 1693 + }, + { + "epoch": 0.01694, + "grad_norm": 0.5925337152955293, + "learning_rate": 0.003, + "loss": 4.1963, + "step": 1694 + }, + { + "epoch": 0.01695, + "grad_norm": 0.5883008115307695, + "learning_rate": 0.003, + "loss": 4.2285, + "step": 1695 + }, + { + "epoch": 0.01696, + "grad_norm": 0.5777962347479031, + "learning_rate": 0.003, + "loss": 4.2174, + "step": 1696 + }, + { + "epoch": 0.01697, + "grad_norm": 0.6137264663560853, + "learning_rate": 0.003, + "loss": 4.1953, + "step": 1697 + }, + { + "epoch": 0.01698, + "grad_norm": 0.5567779485618047, + "learning_rate": 0.003, + "loss": 4.1887, + "step": 1698 + }, + { + "epoch": 0.01699, + "grad_norm": 0.563692648440794, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 1699 + }, + { + "epoch": 0.017, + "grad_norm": 0.5378862525995165, + "learning_rate": 0.003, + "loss": 4.2165, + "step": 1700 + }, + { + "epoch": 0.01701, + "grad_norm": 0.5131924880718308, + "learning_rate": 0.003, + "loss": 4.1973, + "step": 1701 + }, + { + "epoch": 0.01702, + "grad_norm": 0.5473378733884345, + "learning_rate": 0.003, + "loss": 4.2207, + "step": 1702 + }, + { + "epoch": 0.01703, + "grad_norm": 0.659714976195776, + "learning_rate": 0.003, + "loss": 4.2099, + "step": 1703 + }, + { + "epoch": 0.01704, + "grad_norm": 0.7407833108135735, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 1704 + }, + { + "epoch": 0.01705, + "grad_norm": 0.7567928276503215, + "learning_rate": 0.003, + "loss": 4.2073, + "step": 1705 + }, + { + "epoch": 0.01706, + "grad_norm": 0.7694420146818104, + "learning_rate": 0.003, + "loss": 4.2055, + "step": 1706 + }, + { + "epoch": 0.01707, + "grad_norm": 1.0045884849963176, + "learning_rate": 0.003, + "loss": 4.2374, + "step": 1707 + }, + { + "epoch": 0.01708, + "grad_norm": 1.2575213437630939, + "learning_rate": 0.003, + "loss": 4.2465, + "step": 1708 + }, + { + "epoch": 0.01709, + "grad_norm": 0.848642979076845, + "learning_rate": 0.003, + "loss": 4.2418, + "step": 1709 + }, + { + "epoch": 0.0171, + "grad_norm": 0.7781603648787689, + "learning_rate": 0.003, + "loss": 4.1936, + "step": 1710 + }, + { + "epoch": 0.01711, + "grad_norm": 0.7996923845242195, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 1711 + }, + { + "epoch": 0.01712, + "grad_norm": 0.7302785582736923, + "learning_rate": 0.003, + "loss": 4.2298, + "step": 1712 + }, + { + "epoch": 0.01713, + "grad_norm": 0.7495189068505729, + "learning_rate": 0.003, + "loss": 4.204, + "step": 1713 + }, + { + "epoch": 0.01714, + "grad_norm": 0.68320470368433, + "learning_rate": 0.003, + "loss": 4.2407, + "step": 1714 + }, + { + "epoch": 0.01715, + "grad_norm": 0.744728870764733, + "learning_rate": 0.003, + "loss": 4.2481, + "step": 1715 + }, + { + "epoch": 0.01716, + "grad_norm": 0.7728817520220844, + "learning_rate": 0.003, + "loss": 4.2317, + "step": 1716 + }, + { + "epoch": 0.01717, + "grad_norm": 0.8671487727140617, + "learning_rate": 0.003, + "loss": 4.2664, + "step": 1717 + }, + { + "epoch": 0.01718, + "grad_norm": 0.8286537250448458, + "learning_rate": 0.003, + "loss": 4.237, + "step": 1718 + }, + { + "epoch": 0.01719, + "grad_norm": 0.7719901017400573, + "learning_rate": 0.003, + "loss": 4.2535, + "step": 1719 + }, + { + "epoch": 0.0172, + "grad_norm": 0.7363513590865703, + "learning_rate": 0.003, + "loss": 4.2105, + "step": 1720 + }, + { + "epoch": 0.01721, + "grad_norm": 0.6255496683743654, + "learning_rate": 0.003, + "loss": 4.2258, + "step": 1721 + }, + { + "epoch": 0.01722, + "grad_norm": 0.5928702135655549, + "learning_rate": 0.003, + "loss": 4.2129, + "step": 1722 + }, + { + "epoch": 0.01723, + "grad_norm": 0.6570038276937221, + "learning_rate": 0.003, + "loss": 4.2248, + "step": 1723 + }, + { + "epoch": 0.01724, + "grad_norm": 0.6380856228410711, + "learning_rate": 0.003, + "loss": 4.2241, + "step": 1724 + }, + { + "epoch": 0.01725, + "grad_norm": 0.7477214539729845, + "learning_rate": 0.003, + "loss": 4.2088, + "step": 1725 + }, + { + "epoch": 0.01726, + "grad_norm": 0.7064643502935442, + "learning_rate": 0.003, + "loss": 4.2027, + "step": 1726 + }, + { + "epoch": 0.01727, + "grad_norm": 0.7018895227983273, + "learning_rate": 0.003, + "loss": 4.2082, + "step": 1727 + }, + { + "epoch": 0.01728, + "grad_norm": 0.7268534843228023, + "learning_rate": 0.003, + "loss": 4.2116, + "step": 1728 + }, + { + "epoch": 0.01729, + "grad_norm": 0.8265185231394759, + "learning_rate": 0.003, + "loss": 4.2226, + "step": 1729 + }, + { + "epoch": 0.0173, + "grad_norm": 0.9216458529069881, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 1730 + }, + { + "epoch": 0.01731, + "grad_norm": 1.0984877424905655, + "learning_rate": 0.003, + "loss": 4.2294, + "step": 1731 + }, + { + "epoch": 0.01732, + "grad_norm": 0.9457932448227158, + "learning_rate": 0.003, + "loss": 4.2417, + "step": 1732 + }, + { + "epoch": 0.01733, + "grad_norm": 0.8744510824253744, + "learning_rate": 0.003, + "loss": 4.2124, + "step": 1733 + }, + { + "epoch": 0.01734, + "grad_norm": 0.8464357668562363, + "learning_rate": 0.003, + "loss": 4.2277, + "step": 1734 + }, + { + "epoch": 0.01735, + "grad_norm": 0.7933851274075868, + "learning_rate": 0.003, + "loss": 4.1993, + "step": 1735 + }, + { + "epoch": 0.01736, + "grad_norm": 0.7800431566899245, + "learning_rate": 0.003, + "loss": 4.2002, + "step": 1736 + }, + { + "epoch": 0.01737, + "grad_norm": 0.7049958430065779, + "learning_rate": 0.003, + "loss": 4.2254, + "step": 1737 + }, + { + "epoch": 0.01738, + "grad_norm": 0.7322441940210889, + "learning_rate": 0.003, + "loss": 4.2135, + "step": 1738 + }, + { + "epoch": 0.01739, + "grad_norm": 0.7327690195119888, + "learning_rate": 0.003, + "loss": 4.2297, + "step": 1739 + }, + { + "epoch": 0.0174, + "grad_norm": 0.7530152956649067, + "learning_rate": 0.003, + "loss": 4.231, + "step": 1740 + }, + { + "epoch": 0.01741, + "grad_norm": 0.7681419574064243, + "learning_rate": 0.003, + "loss": 4.2369, + "step": 1741 + }, + { + "epoch": 0.01742, + "grad_norm": 0.8179621872439043, + "learning_rate": 0.003, + "loss": 4.2278, + "step": 1742 + }, + { + "epoch": 0.01743, + "grad_norm": 0.9578047731817773, + "learning_rate": 0.003, + "loss": 4.2432, + "step": 1743 + }, + { + "epoch": 0.01744, + "grad_norm": 1.0853544694502002, + "learning_rate": 0.003, + "loss": 4.2354, + "step": 1744 + }, + { + "epoch": 0.01745, + "grad_norm": 0.901300254011202, + "learning_rate": 0.003, + "loss": 4.1985, + "step": 1745 + }, + { + "epoch": 0.01746, + "grad_norm": 0.7977925606337166, + "learning_rate": 0.003, + "loss": 4.2177, + "step": 1746 + }, + { + "epoch": 0.01747, + "grad_norm": 0.7805003864858971, + "learning_rate": 0.003, + "loss": 4.2153, + "step": 1747 + }, + { + "epoch": 0.01748, + "grad_norm": 0.5958068748685724, + "learning_rate": 0.003, + "loss": 4.2004, + "step": 1748 + }, + { + "epoch": 0.01749, + "grad_norm": 0.5609955740586448, + "learning_rate": 0.003, + "loss": 4.1969, + "step": 1749 + }, + { + "epoch": 0.0175, + "grad_norm": 0.5677742109277946, + "learning_rate": 0.003, + "loss": 4.2121, + "step": 1750 + }, + { + "epoch": 0.01751, + "grad_norm": 0.5067010918467684, + "learning_rate": 0.003, + "loss": 4.2009, + "step": 1751 + }, + { + "epoch": 0.01752, + "grad_norm": 0.4469462107071404, + "learning_rate": 0.003, + "loss": 4.239, + "step": 1752 + }, + { + "epoch": 0.01753, + "grad_norm": 0.4299529922000094, + "learning_rate": 0.003, + "loss": 4.1859, + "step": 1753 + }, + { + "epoch": 0.01754, + "grad_norm": 0.39327472210537173, + "learning_rate": 0.003, + "loss": 4.2141, + "step": 1754 + }, + { + "epoch": 0.01755, + "grad_norm": 0.4833428525390586, + "learning_rate": 0.003, + "loss": 4.1582, + "step": 1755 + }, + { + "epoch": 0.01756, + "grad_norm": 0.5692339112673285, + "learning_rate": 0.003, + "loss": 4.2351, + "step": 1756 + }, + { + "epoch": 0.01757, + "grad_norm": 0.765825485079122, + "learning_rate": 0.003, + "loss": 4.2204, + "step": 1757 + }, + { + "epoch": 0.01758, + "grad_norm": 0.9042836101782444, + "learning_rate": 0.003, + "loss": 4.2384, + "step": 1758 + }, + { + "epoch": 0.01759, + "grad_norm": 0.8742540565152447, + "learning_rate": 0.003, + "loss": 4.2064, + "step": 1759 + }, + { + "epoch": 0.0176, + "grad_norm": 0.7243758331092184, + "learning_rate": 0.003, + "loss": 4.2193, + "step": 1760 + }, + { + "epoch": 0.01761, + "grad_norm": 0.6848453720193391, + "learning_rate": 0.003, + "loss": 4.2074, + "step": 1761 + }, + { + "epoch": 0.01762, + "grad_norm": 0.8704455967545811, + "learning_rate": 0.003, + "loss": 4.2328, + "step": 1762 + }, + { + "epoch": 0.01763, + "grad_norm": 0.8471557315960334, + "learning_rate": 0.003, + "loss": 4.2246, + "step": 1763 + }, + { + "epoch": 0.01764, + "grad_norm": 0.7827279567320375, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 1764 + }, + { + "epoch": 0.01765, + "grad_norm": 0.8790545608575575, + "learning_rate": 0.003, + "loss": 4.2237, + "step": 1765 + }, + { + "epoch": 0.01766, + "grad_norm": 0.763900084595892, + "learning_rate": 0.003, + "loss": 4.2322, + "step": 1766 + }, + { + "epoch": 0.01767, + "grad_norm": 0.7449808930716393, + "learning_rate": 0.003, + "loss": 4.1962, + "step": 1767 + }, + { + "epoch": 0.01768, + "grad_norm": 0.6931774004435453, + "learning_rate": 0.003, + "loss": 4.2017, + "step": 1768 + }, + { + "epoch": 0.01769, + "grad_norm": 0.7426550774463054, + "learning_rate": 0.003, + "loss": 4.2066, + "step": 1769 + }, + { + "epoch": 0.0177, + "grad_norm": 0.7834040480889046, + "learning_rate": 0.003, + "loss": 4.1901, + "step": 1770 + }, + { + "epoch": 0.01771, + "grad_norm": 0.6750175361724853, + "learning_rate": 0.003, + "loss": 4.2195, + "step": 1771 + }, + { + "epoch": 0.01772, + "grad_norm": 0.6376494933849937, + "learning_rate": 0.003, + "loss": 4.1888, + "step": 1772 + }, + { + "epoch": 0.01773, + "grad_norm": 0.6192068250431033, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 1773 + }, + { + "epoch": 0.01774, + "grad_norm": 0.6106095483716042, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 1774 + }, + { + "epoch": 0.01775, + "grad_norm": 0.6196510426340663, + "learning_rate": 0.003, + "loss": 4.1895, + "step": 1775 + }, + { + "epoch": 0.01776, + "grad_norm": 0.7097137120766858, + "learning_rate": 0.003, + "loss": 4.1828, + "step": 1776 + }, + { + "epoch": 0.01777, + "grad_norm": 0.7063319061887237, + "learning_rate": 0.003, + "loss": 4.2044, + "step": 1777 + }, + { + "epoch": 0.01778, + "grad_norm": 0.629621456935323, + "learning_rate": 0.003, + "loss": 4.194, + "step": 1778 + }, + { + "epoch": 0.01779, + "grad_norm": 0.7088930396533265, + "learning_rate": 0.003, + "loss": 4.215, + "step": 1779 + }, + { + "epoch": 0.0178, + "grad_norm": 0.7253919374521124, + "learning_rate": 0.003, + "loss": 4.191, + "step": 1780 + }, + { + "epoch": 0.01781, + "grad_norm": 0.7852702442407208, + "learning_rate": 0.003, + "loss": 4.2044, + "step": 1781 + }, + { + "epoch": 0.01782, + "grad_norm": 0.8394755648372281, + "learning_rate": 0.003, + "loss": 4.2328, + "step": 1782 + }, + { + "epoch": 0.01783, + "grad_norm": 0.6910931858922343, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 1783 + }, + { + "epoch": 0.01784, + "grad_norm": 0.7674520462715927, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 1784 + }, + { + "epoch": 0.01785, + "grad_norm": 0.8199387977488617, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 1785 + }, + { + "epoch": 0.01786, + "grad_norm": 0.9173162484249777, + "learning_rate": 0.003, + "loss": 4.2074, + "step": 1786 + }, + { + "epoch": 0.01787, + "grad_norm": 1.334253988039247, + "learning_rate": 0.003, + "loss": 4.2336, + "step": 1787 + }, + { + "epoch": 0.01788, + "grad_norm": 1.0880162732024794, + "learning_rate": 0.003, + "loss": 4.2464, + "step": 1788 + }, + { + "epoch": 0.01789, + "grad_norm": 0.8768448988019708, + "learning_rate": 0.003, + "loss": 4.2114, + "step": 1789 + }, + { + "epoch": 0.0179, + "grad_norm": 0.9478883510444388, + "learning_rate": 0.003, + "loss": 4.2362, + "step": 1790 + }, + { + "epoch": 0.01791, + "grad_norm": 0.9789788338963192, + "learning_rate": 0.003, + "loss": 4.2475, + "step": 1791 + }, + { + "epoch": 0.01792, + "grad_norm": 0.9379061880413183, + "learning_rate": 0.003, + "loss": 4.2484, + "step": 1792 + }, + { + "epoch": 0.01793, + "grad_norm": 0.9327271640213346, + "learning_rate": 0.003, + "loss": 4.215, + "step": 1793 + }, + { + "epoch": 0.01794, + "grad_norm": 0.7690204695866035, + "learning_rate": 0.003, + "loss": 4.2222, + "step": 1794 + }, + { + "epoch": 0.01795, + "grad_norm": 0.6971902449310058, + "learning_rate": 0.003, + "loss": 4.2172, + "step": 1795 + }, + { + "epoch": 0.01796, + "grad_norm": 0.762901849453012, + "learning_rate": 0.003, + "loss": 4.2347, + "step": 1796 + }, + { + "epoch": 0.01797, + "grad_norm": 0.8440627560282756, + "learning_rate": 0.003, + "loss": 4.2037, + "step": 1797 + }, + { + "epoch": 0.01798, + "grad_norm": 0.8368125917294373, + "learning_rate": 0.003, + "loss": 4.2148, + "step": 1798 + }, + { + "epoch": 0.01799, + "grad_norm": 0.7348763481882639, + "learning_rate": 0.003, + "loss": 4.2247, + "step": 1799 + }, + { + "epoch": 0.018, + "grad_norm": 0.7680359832381647, + "learning_rate": 0.003, + "loss": 4.2184, + "step": 1800 + }, + { + "epoch": 0.01801, + "grad_norm": 0.7404783794030995, + "learning_rate": 0.003, + "loss": 4.2101, + "step": 1801 + }, + { + "epoch": 0.01802, + "grad_norm": 0.6580953290368566, + "learning_rate": 0.003, + "loss": 4.2238, + "step": 1802 + }, + { + "epoch": 0.01803, + "grad_norm": 0.6630631878191348, + "learning_rate": 0.003, + "loss": 4.2284, + "step": 1803 + }, + { + "epoch": 0.01804, + "grad_norm": 0.6396012711111247, + "learning_rate": 0.003, + "loss": 4.1992, + "step": 1804 + }, + { + "epoch": 0.01805, + "grad_norm": 0.5559219636409923, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 1805 + }, + { + "epoch": 0.01806, + "grad_norm": 0.6144421410960899, + "learning_rate": 0.003, + "loss": 4.1763, + "step": 1806 + }, + { + "epoch": 0.01807, + "grad_norm": 0.6146777760728932, + "learning_rate": 0.003, + "loss": 4.2121, + "step": 1807 + }, + { + "epoch": 0.01808, + "grad_norm": 0.6116485891270191, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 1808 + }, + { + "epoch": 0.01809, + "grad_norm": 0.8096274979504547, + "learning_rate": 0.003, + "loss": 4.2275, + "step": 1809 + }, + { + "epoch": 0.0181, + "grad_norm": 1.1102312559563035, + "learning_rate": 0.003, + "loss": 4.2346, + "step": 1810 + }, + { + "epoch": 0.01811, + "grad_norm": 0.8754620418282368, + "learning_rate": 0.003, + "loss": 4.2216, + "step": 1811 + }, + { + "epoch": 0.01812, + "grad_norm": 0.6567434234036816, + "learning_rate": 0.003, + "loss": 4.1947, + "step": 1812 + }, + { + "epoch": 0.01813, + "grad_norm": 0.7375490051000831, + "learning_rate": 0.003, + "loss": 4.2409, + "step": 1813 + }, + { + "epoch": 0.01814, + "grad_norm": 0.7842570681491089, + "learning_rate": 0.003, + "loss": 4.2075, + "step": 1814 + }, + { + "epoch": 0.01815, + "grad_norm": 0.6332278380543743, + "learning_rate": 0.003, + "loss": 4.1891, + "step": 1815 + }, + { + "epoch": 0.01816, + "grad_norm": 0.6578340036445237, + "learning_rate": 0.003, + "loss": 4.2166, + "step": 1816 + }, + { + "epoch": 0.01817, + "grad_norm": 0.6624785828428507, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 1817 + }, + { + "epoch": 0.01818, + "grad_norm": 0.5897502755122268, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 1818 + }, + { + "epoch": 0.01819, + "grad_norm": 0.49357607734419184, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 1819 + }, + { + "epoch": 0.0182, + "grad_norm": 0.5025205629657381, + "learning_rate": 0.003, + "loss": 4.2022, + "step": 1820 + }, + { + "epoch": 0.01821, + "grad_norm": 0.5231764315153676, + "learning_rate": 0.003, + "loss": 4.188, + "step": 1821 + }, + { + "epoch": 0.01822, + "grad_norm": 0.589855961825142, + "learning_rate": 0.003, + "loss": 4.2017, + "step": 1822 + }, + { + "epoch": 0.01823, + "grad_norm": 0.6454531707936206, + "learning_rate": 0.003, + "loss": 4.2159, + "step": 1823 + }, + { + "epoch": 0.01824, + "grad_norm": 0.616450651672451, + "learning_rate": 0.003, + "loss": 4.1987, + "step": 1824 + }, + { + "epoch": 0.01825, + "grad_norm": 0.5393299556773667, + "learning_rate": 0.003, + "loss": 4.199, + "step": 1825 + }, + { + "epoch": 0.01826, + "grad_norm": 0.6162633558651508, + "learning_rate": 0.003, + "loss": 4.2135, + "step": 1826 + }, + { + "epoch": 0.01827, + "grad_norm": 0.6366286833338192, + "learning_rate": 0.003, + "loss": 4.1847, + "step": 1827 + }, + { + "epoch": 0.01828, + "grad_norm": 0.5519308619846083, + "learning_rate": 0.003, + "loss": 4.1935, + "step": 1828 + }, + { + "epoch": 0.01829, + "grad_norm": 0.556700588713125, + "learning_rate": 0.003, + "loss": 4.2083, + "step": 1829 + }, + { + "epoch": 0.0183, + "grad_norm": 0.6225243144725614, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 1830 + }, + { + "epoch": 0.01831, + "grad_norm": 0.8092938095153425, + "learning_rate": 0.003, + "loss": 4.2017, + "step": 1831 + }, + { + "epoch": 0.01832, + "grad_norm": 1.156880999962459, + "learning_rate": 0.003, + "loss": 4.214, + "step": 1832 + }, + { + "epoch": 0.01833, + "grad_norm": 0.8498167035958548, + "learning_rate": 0.003, + "loss": 4.2422, + "step": 1833 + }, + { + "epoch": 0.01834, + "grad_norm": 0.6651811196480657, + "learning_rate": 0.003, + "loss": 4.2008, + "step": 1834 + }, + { + "epoch": 0.01835, + "grad_norm": 0.7438128363368247, + "learning_rate": 0.003, + "loss": 4.205, + "step": 1835 + }, + { + "epoch": 0.01836, + "grad_norm": 0.7712205326044778, + "learning_rate": 0.003, + "loss": 4.2014, + "step": 1836 + }, + { + "epoch": 0.01837, + "grad_norm": 0.7094945295905764, + "learning_rate": 0.003, + "loss": 4.1968, + "step": 1837 + }, + { + "epoch": 0.01838, + "grad_norm": 0.8255996162995448, + "learning_rate": 0.003, + "loss": 4.2049, + "step": 1838 + }, + { + "epoch": 0.01839, + "grad_norm": 0.7001056258854736, + "learning_rate": 0.003, + "loss": 4.2234, + "step": 1839 + }, + { + "epoch": 0.0184, + "grad_norm": 0.7156089343491636, + "learning_rate": 0.003, + "loss": 4.2248, + "step": 1840 + }, + { + "epoch": 0.01841, + "grad_norm": 0.7514284093505675, + "learning_rate": 0.003, + "loss": 4.164, + "step": 1841 + }, + { + "epoch": 0.01842, + "grad_norm": 0.7804042052375482, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 1842 + }, + { + "epoch": 0.01843, + "grad_norm": 0.7741073572881331, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 1843 + }, + { + "epoch": 0.01844, + "grad_norm": 0.8011416312871611, + "learning_rate": 0.003, + "loss": 4.1842, + "step": 1844 + }, + { + "epoch": 0.01845, + "grad_norm": 0.8386720056960224, + "learning_rate": 0.003, + "loss": 4.187, + "step": 1845 + }, + { + "epoch": 0.01846, + "grad_norm": 0.9781723483143008, + "learning_rate": 0.003, + "loss": 4.2242, + "step": 1846 + }, + { + "epoch": 0.01847, + "grad_norm": 0.9173016227544075, + "learning_rate": 0.003, + "loss": 4.2189, + "step": 1847 + }, + { + "epoch": 0.01848, + "grad_norm": 0.961352235410328, + "learning_rate": 0.003, + "loss": 4.2359, + "step": 1848 + }, + { + "epoch": 0.01849, + "grad_norm": 0.9115360913743059, + "learning_rate": 0.003, + "loss": 4.241, + "step": 1849 + }, + { + "epoch": 0.0185, + "grad_norm": 0.949227003474929, + "learning_rate": 0.003, + "loss": 4.2298, + "step": 1850 + }, + { + "epoch": 0.01851, + "grad_norm": 0.756851342238193, + "learning_rate": 0.003, + "loss": 4.2063, + "step": 1851 + }, + { + "epoch": 0.01852, + "grad_norm": 0.7062657388831308, + "learning_rate": 0.003, + "loss": 4.2058, + "step": 1852 + }, + { + "epoch": 0.01853, + "grad_norm": 0.804960952863702, + "learning_rate": 0.003, + "loss": 4.206, + "step": 1853 + }, + { + "epoch": 0.01854, + "grad_norm": 0.8048402157100555, + "learning_rate": 0.003, + "loss": 4.2379, + "step": 1854 + }, + { + "epoch": 0.01855, + "grad_norm": 0.8514933979131342, + "learning_rate": 0.003, + "loss": 4.2236, + "step": 1855 + }, + { + "epoch": 0.01856, + "grad_norm": 0.9293561457561126, + "learning_rate": 0.003, + "loss": 4.2144, + "step": 1856 + }, + { + "epoch": 0.01857, + "grad_norm": 0.8452785281639297, + "learning_rate": 0.003, + "loss": 4.1937, + "step": 1857 + }, + { + "epoch": 0.01858, + "grad_norm": 0.805524803230092, + "learning_rate": 0.003, + "loss": 4.2015, + "step": 1858 + }, + { + "epoch": 0.01859, + "grad_norm": 0.826226040676062, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 1859 + }, + { + "epoch": 0.0186, + "grad_norm": 0.7646190876843869, + "learning_rate": 0.003, + "loss": 4.1808, + "step": 1860 + }, + { + "epoch": 0.01861, + "grad_norm": 0.9398444707205434, + "learning_rate": 0.003, + "loss": 4.229, + "step": 1861 + }, + { + "epoch": 0.01862, + "grad_norm": 1.0799822817783482, + "learning_rate": 0.003, + "loss": 4.208, + "step": 1862 + }, + { + "epoch": 0.01863, + "grad_norm": 1.0663393244093409, + "learning_rate": 0.003, + "loss": 4.2176, + "step": 1863 + }, + { + "epoch": 0.01864, + "grad_norm": 0.9063119158507146, + "learning_rate": 0.003, + "loss": 4.243, + "step": 1864 + }, + { + "epoch": 0.01865, + "grad_norm": 0.8878085186485055, + "learning_rate": 0.003, + "loss": 4.222, + "step": 1865 + }, + { + "epoch": 0.01866, + "grad_norm": 0.8278859311900081, + "learning_rate": 0.003, + "loss": 4.2161, + "step": 1866 + }, + { + "epoch": 0.01867, + "grad_norm": 0.9175811468419192, + "learning_rate": 0.003, + "loss": 4.1983, + "step": 1867 + }, + { + "epoch": 0.01868, + "grad_norm": 0.8195895794655929, + "learning_rate": 0.003, + "loss": 4.2097, + "step": 1868 + }, + { + "epoch": 0.01869, + "grad_norm": 0.6958671652018995, + "learning_rate": 0.003, + "loss": 4.2233, + "step": 1869 + }, + { + "epoch": 0.0187, + "grad_norm": 0.7276701584872959, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 1870 + }, + { + "epoch": 0.01871, + "grad_norm": 0.700252662578374, + "learning_rate": 0.003, + "loss": 4.1886, + "step": 1871 + }, + { + "epoch": 0.01872, + "grad_norm": 0.7936164889803413, + "learning_rate": 0.003, + "loss": 4.2323, + "step": 1872 + }, + { + "epoch": 0.01873, + "grad_norm": 0.8026794723116766, + "learning_rate": 0.003, + "loss": 4.1938, + "step": 1873 + }, + { + "epoch": 0.01874, + "grad_norm": 0.6813841695856067, + "learning_rate": 0.003, + "loss": 4.1939, + "step": 1874 + }, + { + "epoch": 0.01875, + "grad_norm": 0.7703759265763755, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 1875 + }, + { + "epoch": 0.01876, + "grad_norm": 0.7048464208203703, + "learning_rate": 0.003, + "loss": 4.2195, + "step": 1876 + }, + { + "epoch": 0.01877, + "grad_norm": 0.6552231659596246, + "learning_rate": 0.003, + "loss": 4.2204, + "step": 1877 + }, + { + "epoch": 0.01878, + "grad_norm": 0.6594564394551935, + "learning_rate": 0.003, + "loss": 4.2061, + "step": 1878 + }, + { + "epoch": 0.01879, + "grad_norm": 0.6775220505077814, + "learning_rate": 0.003, + "loss": 4.1841, + "step": 1879 + }, + { + "epoch": 0.0188, + "grad_norm": 0.7469198030113778, + "learning_rate": 0.003, + "loss": 4.1948, + "step": 1880 + }, + { + "epoch": 0.01881, + "grad_norm": 0.8052292593043413, + "learning_rate": 0.003, + "loss": 4.1959, + "step": 1881 + }, + { + "epoch": 0.01882, + "grad_norm": 0.9310051602523918, + "learning_rate": 0.003, + "loss": 4.2256, + "step": 1882 + }, + { + "epoch": 0.01883, + "grad_norm": 0.8470067207894152, + "learning_rate": 0.003, + "loss": 4.2057, + "step": 1883 + }, + { + "epoch": 0.01884, + "grad_norm": 0.7163623431526733, + "learning_rate": 0.003, + "loss": 4.1948, + "step": 1884 + }, + { + "epoch": 0.01885, + "grad_norm": 0.7917195791610646, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 1885 + }, + { + "epoch": 0.01886, + "grad_norm": 0.7950048534183088, + "learning_rate": 0.003, + "loss": 4.2171, + "step": 1886 + }, + { + "epoch": 0.01887, + "grad_norm": 0.7209518927340428, + "learning_rate": 0.003, + "loss": 4.2025, + "step": 1887 + }, + { + "epoch": 0.01888, + "grad_norm": 0.633107358042472, + "learning_rate": 0.003, + "loss": 4.1883, + "step": 1888 + }, + { + "epoch": 0.01889, + "grad_norm": 0.6143038001546575, + "learning_rate": 0.003, + "loss": 4.1996, + "step": 1889 + }, + { + "epoch": 0.0189, + "grad_norm": 0.5158052453680656, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 1890 + }, + { + "epoch": 0.01891, + "grad_norm": 0.4633760186396196, + "learning_rate": 0.003, + "loss": 4.2224, + "step": 1891 + }, + { + "epoch": 0.01892, + "grad_norm": 0.4927650862046063, + "learning_rate": 0.003, + "loss": 4.1956, + "step": 1892 + }, + { + "epoch": 0.01893, + "grad_norm": 0.4373024700349087, + "learning_rate": 0.003, + "loss": 4.1773, + "step": 1893 + }, + { + "epoch": 0.01894, + "grad_norm": 0.40335105556509215, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 1894 + }, + { + "epoch": 0.01895, + "grad_norm": 0.3931015972388579, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 1895 + }, + { + "epoch": 0.01896, + "grad_norm": 0.369330374322213, + "learning_rate": 0.003, + "loss": 4.1861, + "step": 1896 + }, + { + "epoch": 0.01897, + "grad_norm": 0.38394762475942135, + "learning_rate": 0.003, + "loss": 4.1841, + "step": 1897 + }, + { + "epoch": 0.01898, + "grad_norm": 0.4616420599151427, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 1898 + }, + { + "epoch": 0.01899, + "grad_norm": 0.6786086644128849, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 1899 + }, + { + "epoch": 0.019, + "grad_norm": 1.1028955308001205, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 1900 + }, + { + "epoch": 0.01901, + "grad_norm": 1.0381658886424407, + "learning_rate": 0.003, + "loss": 4.2416, + "step": 1901 + }, + { + "epoch": 0.01902, + "grad_norm": 0.6228046274507568, + "learning_rate": 0.003, + "loss": 4.202, + "step": 1902 + }, + { + "epoch": 0.01903, + "grad_norm": 0.552687373313921, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 1903 + }, + { + "epoch": 0.01904, + "grad_norm": 0.5642583939104693, + "learning_rate": 0.003, + "loss": 4.2113, + "step": 1904 + }, + { + "epoch": 0.01905, + "grad_norm": 0.4866940659015072, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 1905 + }, + { + "epoch": 0.01906, + "grad_norm": 0.6770303387740269, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 1906 + }, + { + "epoch": 0.01907, + "grad_norm": 0.6799220004643979, + "learning_rate": 0.003, + "loss": 4.199, + "step": 1907 + }, + { + "epoch": 0.01908, + "grad_norm": 0.5738027965979297, + "learning_rate": 0.003, + "loss": 4.2084, + "step": 1908 + }, + { + "epoch": 0.01909, + "grad_norm": 0.5646602374989558, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 1909 + }, + { + "epoch": 0.0191, + "grad_norm": 0.6702075278822099, + "learning_rate": 0.003, + "loss": 4.1866, + "step": 1910 + }, + { + "epoch": 0.01911, + "grad_norm": 0.7731308719042311, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 1911 + }, + { + "epoch": 0.01912, + "grad_norm": 0.6258475347121828, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 1912 + }, + { + "epoch": 0.01913, + "grad_norm": 0.6373334035841824, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 1913 + }, + { + "epoch": 0.01914, + "grad_norm": 0.7497697739080498, + "learning_rate": 0.003, + "loss": 4.212, + "step": 1914 + }, + { + "epoch": 0.01915, + "grad_norm": 0.8443649873420552, + "learning_rate": 0.003, + "loss": 4.204, + "step": 1915 + }, + { + "epoch": 0.01916, + "grad_norm": 0.8327185244516057, + "learning_rate": 0.003, + "loss": 4.1829, + "step": 1916 + }, + { + "epoch": 0.01917, + "grad_norm": 0.8831134466171928, + "learning_rate": 0.003, + "loss": 4.2052, + "step": 1917 + }, + { + "epoch": 0.01918, + "grad_norm": 0.9318795049142128, + "learning_rate": 0.003, + "loss": 4.1944, + "step": 1918 + }, + { + "epoch": 0.01919, + "grad_norm": 0.944334593915859, + "learning_rate": 0.003, + "loss": 4.2049, + "step": 1919 + }, + { + "epoch": 0.0192, + "grad_norm": 0.9365986219494583, + "learning_rate": 0.003, + "loss": 4.2076, + "step": 1920 + }, + { + "epoch": 0.01921, + "grad_norm": 0.9617762005888578, + "learning_rate": 0.003, + "loss": 4.2016, + "step": 1921 + }, + { + "epoch": 0.01922, + "grad_norm": 0.9805293516298391, + "learning_rate": 0.003, + "loss": 4.2181, + "step": 1922 + }, + { + "epoch": 0.01923, + "grad_norm": 0.8498471851851511, + "learning_rate": 0.003, + "loss": 4.2204, + "step": 1923 + }, + { + "epoch": 0.01924, + "grad_norm": 0.9257747078341632, + "learning_rate": 0.003, + "loss": 4.2284, + "step": 1924 + }, + { + "epoch": 0.01925, + "grad_norm": 1.0324220402854314, + "learning_rate": 0.003, + "loss": 4.2182, + "step": 1925 + }, + { + "epoch": 0.01926, + "grad_norm": 0.9745230962825202, + "learning_rate": 0.003, + "loss": 4.2282, + "step": 1926 + }, + { + "epoch": 0.01927, + "grad_norm": 0.7705843813355484, + "learning_rate": 0.003, + "loss": 4.2129, + "step": 1927 + }, + { + "epoch": 0.01928, + "grad_norm": 0.5911790602013839, + "learning_rate": 0.003, + "loss": 4.1789, + "step": 1928 + }, + { + "epoch": 0.01929, + "grad_norm": 0.5782217294801848, + "learning_rate": 0.003, + "loss": 4.1986, + "step": 1929 + }, + { + "epoch": 0.0193, + "grad_norm": 0.5556254348196169, + "learning_rate": 0.003, + "loss": 4.1871, + "step": 1930 + }, + { + "epoch": 0.01931, + "grad_norm": 0.6337897693351197, + "learning_rate": 0.003, + "loss": 4.188, + "step": 1931 + }, + { + "epoch": 0.01932, + "grad_norm": 0.6851141809276902, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 1932 + }, + { + "epoch": 0.01933, + "grad_norm": 0.6532471380653952, + "learning_rate": 0.003, + "loss": 4.1838, + "step": 1933 + }, + { + "epoch": 0.01934, + "grad_norm": 0.660311918680962, + "learning_rate": 0.003, + "loss": 4.1881, + "step": 1934 + }, + { + "epoch": 0.01935, + "grad_norm": 0.6202434802996027, + "learning_rate": 0.003, + "loss": 4.2019, + "step": 1935 + }, + { + "epoch": 0.01936, + "grad_norm": 0.6456697693032586, + "learning_rate": 0.003, + "loss": 4.1829, + "step": 1936 + }, + { + "epoch": 0.01937, + "grad_norm": 0.6945407817649866, + "learning_rate": 0.003, + "loss": 4.1968, + "step": 1937 + }, + { + "epoch": 0.01938, + "grad_norm": 0.7393543695473613, + "learning_rate": 0.003, + "loss": 4.1957, + "step": 1938 + }, + { + "epoch": 0.01939, + "grad_norm": 0.7713579649815137, + "learning_rate": 0.003, + "loss": 4.223, + "step": 1939 + }, + { + "epoch": 0.0194, + "grad_norm": 0.6228971544711813, + "learning_rate": 0.003, + "loss": 4.1928, + "step": 1940 + }, + { + "epoch": 0.01941, + "grad_norm": 0.5726639902946573, + "learning_rate": 0.003, + "loss": 4.1739, + "step": 1941 + }, + { + "epoch": 0.01942, + "grad_norm": 0.6064973298236461, + "learning_rate": 0.003, + "loss": 4.1989, + "step": 1942 + }, + { + "epoch": 0.01943, + "grad_norm": 0.6888737571597997, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 1943 + }, + { + "epoch": 0.01944, + "grad_norm": 0.6708431670479557, + "learning_rate": 0.003, + "loss": 4.2158, + "step": 1944 + }, + { + "epoch": 0.01945, + "grad_norm": 0.5681415129494921, + "learning_rate": 0.003, + "loss": 4.1787, + "step": 1945 + }, + { + "epoch": 0.01946, + "grad_norm": 0.47730421654011235, + "learning_rate": 0.003, + "loss": 4.1716, + "step": 1946 + }, + { + "epoch": 0.01947, + "grad_norm": 0.47116667407771645, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 1947 + }, + { + "epoch": 0.01948, + "grad_norm": 0.6002169491076368, + "learning_rate": 0.003, + "loss": 4.2004, + "step": 1948 + }, + { + "epoch": 0.01949, + "grad_norm": 0.6804506442238769, + "learning_rate": 0.003, + "loss": 4.1961, + "step": 1949 + }, + { + "epoch": 0.0195, + "grad_norm": 0.9031150085350885, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 1950 + }, + { + "epoch": 0.01951, + "grad_norm": 1.055108763514552, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 1951 + }, + { + "epoch": 0.01952, + "grad_norm": 0.7942759742680753, + "learning_rate": 0.003, + "loss": 4.1947, + "step": 1952 + }, + { + "epoch": 0.01953, + "grad_norm": 0.6995419703208098, + "learning_rate": 0.003, + "loss": 4.2007, + "step": 1953 + }, + { + "epoch": 0.01954, + "grad_norm": 0.7323475440015588, + "learning_rate": 0.003, + "loss": 4.2017, + "step": 1954 + }, + { + "epoch": 0.01955, + "grad_norm": 0.654602209622962, + "learning_rate": 0.003, + "loss": 4.1897, + "step": 1955 + }, + { + "epoch": 0.01956, + "grad_norm": 0.6795706239523924, + "learning_rate": 0.003, + "loss": 4.188, + "step": 1956 + }, + { + "epoch": 0.01957, + "grad_norm": 0.822104190978127, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 1957 + }, + { + "epoch": 0.01958, + "grad_norm": 0.7829306267633842, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 1958 + }, + { + "epoch": 0.01959, + "grad_norm": 0.7521187388682534, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 1959 + }, + { + "epoch": 0.0196, + "grad_norm": 0.8699203949276841, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 1960 + }, + { + "epoch": 0.01961, + "grad_norm": 0.9397547951451798, + "learning_rate": 0.003, + "loss": 4.1846, + "step": 1961 + }, + { + "epoch": 0.01962, + "grad_norm": 0.8367517905224471, + "learning_rate": 0.003, + "loss": 4.1848, + "step": 1962 + }, + { + "epoch": 0.01963, + "grad_norm": 0.8914011026895058, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 1963 + }, + { + "epoch": 0.01964, + "grad_norm": 1.069474116819015, + "learning_rate": 0.003, + "loss": 4.2192, + "step": 1964 + }, + { + "epoch": 0.01965, + "grad_norm": 0.9846418582261246, + "learning_rate": 0.003, + "loss": 4.2097, + "step": 1965 + }, + { + "epoch": 0.01966, + "grad_norm": 1.190189959584559, + "learning_rate": 0.003, + "loss": 4.2246, + "step": 1966 + }, + { + "epoch": 0.01967, + "grad_norm": 0.9331071177303962, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 1967 + }, + { + "epoch": 0.01968, + "grad_norm": 0.8236928876485803, + "learning_rate": 0.003, + "loss": 4.2092, + "step": 1968 + }, + { + "epoch": 0.01969, + "grad_norm": 0.9035337012628801, + "learning_rate": 0.003, + "loss": 4.2128, + "step": 1969 + }, + { + "epoch": 0.0197, + "grad_norm": 0.9584839773684666, + "learning_rate": 0.003, + "loss": 4.2139, + "step": 1970 + }, + { + "epoch": 0.01971, + "grad_norm": 1.1976612578146555, + "learning_rate": 0.003, + "loss": 4.2078, + "step": 1971 + }, + { + "epoch": 0.01972, + "grad_norm": 1.03766882598544, + "learning_rate": 0.003, + "loss": 4.216, + "step": 1972 + }, + { + "epoch": 0.01973, + "grad_norm": 1.157490820023187, + "learning_rate": 0.003, + "loss": 4.204, + "step": 1973 + }, + { + "epoch": 0.01974, + "grad_norm": 1.0168705909824025, + "learning_rate": 0.003, + "loss": 4.2365, + "step": 1974 + }, + { + "epoch": 0.01975, + "grad_norm": 0.9807823986551194, + "learning_rate": 0.003, + "loss": 4.2359, + "step": 1975 + }, + { + "epoch": 0.01976, + "grad_norm": 0.7975927738530828, + "learning_rate": 0.003, + "loss": 4.2115, + "step": 1976 + }, + { + "epoch": 0.01977, + "grad_norm": 0.9111601004721311, + "learning_rate": 0.003, + "loss": 4.2503, + "step": 1977 + }, + { + "epoch": 0.01978, + "grad_norm": 0.95585271722792, + "learning_rate": 0.003, + "loss": 4.2456, + "step": 1978 + }, + { + "epoch": 0.01979, + "grad_norm": 0.8967358768483624, + "learning_rate": 0.003, + "loss": 4.2367, + "step": 1979 + }, + { + "epoch": 0.0198, + "grad_norm": 0.9098539062714608, + "learning_rate": 0.003, + "loss": 4.2135, + "step": 1980 + }, + { + "epoch": 0.01981, + "grad_norm": 0.9619207371647327, + "learning_rate": 0.003, + "loss": 4.2036, + "step": 1981 + }, + { + "epoch": 0.01982, + "grad_norm": 1.004618203421798, + "learning_rate": 0.003, + "loss": 4.2311, + "step": 1982 + }, + { + "epoch": 0.01983, + "grad_norm": 0.8778331022367669, + "learning_rate": 0.003, + "loss": 4.2203, + "step": 1983 + }, + { + "epoch": 0.01984, + "grad_norm": 0.7409261134848839, + "learning_rate": 0.003, + "loss": 4.2234, + "step": 1984 + }, + { + "epoch": 0.01985, + "grad_norm": 0.7987127824711159, + "learning_rate": 0.003, + "loss": 4.2126, + "step": 1985 + }, + { + "epoch": 0.01986, + "grad_norm": 0.8488590004969258, + "learning_rate": 0.003, + "loss": 4.2379, + "step": 1986 + }, + { + "epoch": 0.01987, + "grad_norm": 0.856993962490694, + "learning_rate": 0.003, + "loss": 4.1987, + "step": 1987 + }, + { + "epoch": 0.01988, + "grad_norm": 0.6814479531298734, + "learning_rate": 0.003, + "loss": 4.2021, + "step": 1988 + }, + { + "epoch": 0.01989, + "grad_norm": 0.5447842315255877, + "learning_rate": 0.003, + "loss": 4.1871, + "step": 1989 + }, + { + "epoch": 0.0199, + "grad_norm": 0.5760022915320738, + "learning_rate": 0.003, + "loss": 4.1964, + "step": 1990 + }, + { + "epoch": 0.01991, + "grad_norm": 0.5132919071482156, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 1991 + }, + { + "epoch": 0.01992, + "grad_norm": 0.5851245859941666, + "learning_rate": 0.003, + "loss": 4.2128, + "step": 1992 + }, + { + "epoch": 0.01993, + "grad_norm": 0.6267354593691943, + "learning_rate": 0.003, + "loss": 4.1956, + "step": 1993 + }, + { + "epoch": 0.01994, + "grad_norm": 0.7099792076486661, + "learning_rate": 0.003, + "loss": 4.2188, + "step": 1994 + }, + { + "epoch": 0.01995, + "grad_norm": 0.7678935327418717, + "learning_rate": 0.003, + "loss": 4.2111, + "step": 1995 + }, + { + "epoch": 0.01996, + "grad_norm": 0.6394486685805509, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 1996 + }, + { + "epoch": 0.01997, + "grad_norm": 0.4226153497592854, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 1997 + }, + { + "epoch": 0.01998, + "grad_norm": 0.5031225205891672, + "learning_rate": 0.003, + "loss": 4.1844, + "step": 1998 + }, + { + "epoch": 0.01999, + "grad_norm": 0.5395430302746123, + "learning_rate": 0.003, + "loss": 4.1747, + "step": 1999 + }, + { + "epoch": 0.02, + "grad_norm": 0.46234156015204186, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 2000 + }, + { + "epoch": 0.02001, + "grad_norm": 0.4395345274961725, + "learning_rate": 0.003, + "loss": 4.1934, + "step": 2001 + }, + { + "epoch": 0.02002, + "grad_norm": 0.3957661324841875, + "learning_rate": 0.003, + "loss": 4.1863, + "step": 2002 + }, + { + "epoch": 0.02003, + "grad_norm": 0.4711152480437331, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 2003 + }, + { + "epoch": 0.02004, + "grad_norm": 0.546136823152871, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 2004 + }, + { + "epoch": 0.02005, + "grad_norm": 0.654074266015456, + "learning_rate": 0.003, + "loss": 4.1803, + "step": 2005 + }, + { + "epoch": 0.02006, + "grad_norm": 0.710370558494441, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 2006 + }, + { + "epoch": 0.02007, + "grad_norm": 0.7542070025547006, + "learning_rate": 0.003, + "loss": 4.1945, + "step": 2007 + }, + { + "epoch": 0.02008, + "grad_norm": 0.7773827468349818, + "learning_rate": 0.003, + "loss": 4.191, + "step": 2008 + }, + { + "epoch": 0.02009, + "grad_norm": 0.6899352222366664, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 2009 + }, + { + "epoch": 0.0201, + "grad_norm": 0.5614651061397582, + "learning_rate": 0.003, + "loss": 4.1935, + "step": 2010 + }, + { + "epoch": 0.02011, + "grad_norm": 0.6625002839338998, + "learning_rate": 0.003, + "loss": 4.183, + "step": 2011 + }, + { + "epoch": 0.02012, + "grad_norm": 0.6997403670690155, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 2012 + }, + { + "epoch": 0.02013, + "grad_norm": 0.6697515151480535, + "learning_rate": 0.003, + "loss": 4.1896, + "step": 2013 + }, + { + "epoch": 0.02014, + "grad_norm": 0.6386903403073065, + "learning_rate": 0.003, + "loss": 4.1729, + "step": 2014 + }, + { + "epoch": 0.02015, + "grad_norm": 0.5907166723123874, + "learning_rate": 0.003, + "loss": 4.1657, + "step": 2015 + }, + { + "epoch": 0.02016, + "grad_norm": 0.605106268984101, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 2016 + }, + { + "epoch": 0.02017, + "grad_norm": 0.6631806062284518, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 2017 + }, + { + "epoch": 0.02018, + "grad_norm": 0.6347297750042842, + "learning_rate": 0.003, + "loss": 4.1746, + "step": 2018 + }, + { + "epoch": 0.02019, + "grad_norm": 0.8248022848153107, + "learning_rate": 0.003, + "loss": 4.1834, + "step": 2019 + }, + { + "epoch": 0.0202, + "grad_norm": 0.9904970732072179, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 2020 + }, + { + "epoch": 0.02021, + "grad_norm": 0.9086090012812373, + "learning_rate": 0.003, + "loss": 4.2148, + "step": 2021 + }, + { + "epoch": 0.02022, + "grad_norm": 0.9232445860299345, + "learning_rate": 0.003, + "loss": 4.175, + "step": 2022 + }, + { + "epoch": 0.02023, + "grad_norm": 0.8050131783240017, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 2023 + }, + { + "epoch": 0.02024, + "grad_norm": 0.7442373542055678, + "learning_rate": 0.003, + "loss": 4.1717, + "step": 2024 + }, + { + "epoch": 0.02025, + "grad_norm": 0.9832310341451624, + "learning_rate": 0.003, + "loss": 4.2324, + "step": 2025 + }, + { + "epoch": 0.02026, + "grad_norm": 0.9556185585548506, + "learning_rate": 0.003, + "loss": 4.2147, + "step": 2026 + }, + { + "epoch": 0.02027, + "grad_norm": 1.0633653823279028, + "learning_rate": 0.003, + "loss": 4.218, + "step": 2027 + }, + { + "epoch": 0.02028, + "grad_norm": 0.8361708667618777, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 2028 + }, + { + "epoch": 0.02029, + "grad_norm": 0.8621114948206428, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 2029 + }, + { + "epoch": 0.0203, + "grad_norm": 0.7505639644342775, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 2030 + }, + { + "epoch": 0.02031, + "grad_norm": 0.7688539298918926, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 2031 + }, + { + "epoch": 0.02032, + "grad_norm": 0.8008832349814213, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 2032 + }, + { + "epoch": 0.02033, + "grad_norm": 0.7744220160391694, + "learning_rate": 0.003, + "loss": 4.2069, + "step": 2033 + }, + { + "epoch": 0.02034, + "grad_norm": 0.9334621419858093, + "learning_rate": 0.003, + "loss": 4.2063, + "step": 2034 + }, + { + "epoch": 0.02035, + "grad_norm": 1.0003714475706196, + "learning_rate": 0.003, + "loss": 4.1981, + "step": 2035 + }, + { + "epoch": 0.02036, + "grad_norm": 1.0165755074456069, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 2036 + }, + { + "epoch": 0.02037, + "grad_norm": 0.8752656941969562, + "learning_rate": 0.003, + "loss": 4.2133, + "step": 2037 + }, + { + "epoch": 0.02038, + "grad_norm": 0.8030828089902575, + "learning_rate": 0.003, + "loss": 4.1951, + "step": 2038 + }, + { + "epoch": 0.02039, + "grad_norm": 0.6849651287466818, + "learning_rate": 0.003, + "loss": 4.1984, + "step": 2039 + }, + { + "epoch": 0.0204, + "grad_norm": 0.7048992687138306, + "learning_rate": 0.003, + "loss": 4.2015, + "step": 2040 + }, + { + "epoch": 0.02041, + "grad_norm": 0.5846622816322785, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 2041 + }, + { + "epoch": 0.02042, + "grad_norm": 0.579059647971908, + "learning_rate": 0.003, + "loss": 4.159, + "step": 2042 + }, + { + "epoch": 0.02043, + "grad_norm": 0.5320216814522557, + "learning_rate": 0.003, + "loss": 4.2018, + "step": 2043 + }, + { + "epoch": 0.02044, + "grad_norm": 0.5749364817997652, + "learning_rate": 0.003, + "loss": 4.174, + "step": 2044 + }, + { + "epoch": 0.02045, + "grad_norm": 0.5872195224989512, + "learning_rate": 0.003, + "loss": 4.1881, + "step": 2045 + }, + { + "epoch": 0.02046, + "grad_norm": 0.6166345493647961, + "learning_rate": 0.003, + "loss": 4.1634, + "step": 2046 + }, + { + "epoch": 0.02047, + "grad_norm": 0.7223919170905914, + "learning_rate": 0.003, + "loss": 4.2037, + "step": 2047 + }, + { + "epoch": 0.02048, + "grad_norm": 0.7799560828695167, + "learning_rate": 0.003, + "loss": 4.1877, + "step": 2048 + }, + { + "epoch": 0.02049, + "grad_norm": 0.8702209657581298, + "learning_rate": 0.003, + "loss": 4.2131, + "step": 2049 + }, + { + "epoch": 0.0205, + "grad_norm": 0.9375771238974213, + "learning_rate": 0.003, + "loss": 4.1631, + "step": 2050 + }, + { + "epoch": 0.02051, + "grad_norm": 1.1873073578231947, + "learning_rate": 0.003, + "loss": 4.178, + "step": 2051 + }, + { + "epoch": 0.02052, + "grad_norm": 1.03209456773, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 2052 + }, + { + "epoch": 0.02053, + "grad_norm": 1.0555848829046368, + "learning_rate": 0.003, + "loss": 4.1832, + "step": 2053 + }, + { + "epoch": 0.02054, + "grad_norm": 0.9050417954028227, + "learning_rate": 0.003, + "loss": 4.2168, + "step": 2054 + }, + { + "epoch": 0.02055, + "grad_norm": 0.8496945287219897, + "learning_rate": 0.003, + "loss": 4.1765, + "step": 2055 + }, + { + "epoch": 0.02056, + "grad_norm": 0.7540734931162184, + "learning_rate": 0.003, + "loss": 4.17, + "step": 2056 + }, + { + "epoch": 0.02057, + "grad_norm": 0.79729190390271, + "learning_rate": 0.003, + "loss": 4.2076, + "step": 2057 + }, + { + "epoch": 0.02058, + "grad_norm": 0.8092766990194165, + "learning_rate": 0.003, + "loss": 4.1879, + "step": 2058 + }, + { + "epoch": 0.02059, + "grad_norm": 0.787520651985971, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 2059 + }, + { + "epoch": 0.0206, + "grad_norm": 0.8179849805821454, + "learning_rate": 0.003, + "loss": 4.2123, + "step": 2060 + }, + { + "epoch": 0.02061, + "grad_norm": 0.765277725532741, + "learning_rate": 0.003, + "loss": 4.1867, + "step": 2061 + }, + { + "epoch": 0.02062, + "grad_norm": 0.6833167046153735, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 2062 + }, + { + "epoch": 0.02063, + "grad_norm": 0.7852297427280791, + "learning_rate": 0.003, + "loss": 4.1788, + "step": 2063 + }, + { + "epoch": 0.02064, + "grad_norm": 0.7558602646875056, + "learning_rate": 0.003, + "loss": 4.1894, + "step": 2064 + }, + { + "epoch": 0.02065, + "grad_norm": 0.7461513201926369, + "learning_rate": 0.003, + "loss": 4.1686, + "step": 2065 + }, + { + "epoch": 0.02066, + "grad_norm": 0.8279588238364882, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 2066 + }, + { + "epoch": 0.02067, + "grad_norm": 0.7670898564018478, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 2067 + }, + { + "epoch": 0.02068, + "grad_norm": 0.6723403011847539, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 2068 + }, + { + "epoch": 0.02069, + "grad_norm": 0.6336511314262159, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 2069 + }, + { + "epoch": 0.0207, + "grad_norm": 0.70657998006886, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 2070 + }, + { + "epoch": 0.02071, + "grad_norm": 0.7487170820279913, + "learning_rate": 0.003, + "loss": 4.1912, + "step": 2071 + }, + { + "epoch": 0.02072, + "grad_norm": 0.7172563865038417, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 2072 + }, + { + "epoch": 0.02073, + "grad_norm": 0.7815042091167561, + "learning_rate": 0.003, + "loss": 4.1796, + "step": 2073 + }, + { + "epoch": 0.02074, + "grad_norm": 0.7869404752550972, + "learning_rate": 0.003, + "loss": 4.1864, + "step": 2074 + }, + { + "epoch": 0.02075, + "grad_norm": 0.7115380813621578, + "learning_rate": 0.003, + "loss": 4.2091, + "step": 2075 + }, + { + "epoch": 0.02076, + "grad_norm": 0.6271616020332964, + "learning_rate": 0.003, + "loss": 4.199, + "step": 2076 + }, + { + "epoch": 0.02077, + "grad_norm": 0.5945516898335454, + "learning_rate": 0.003, + "loss": 4.1669, + "step": 2077 + }, + { + "epoch": 0.02078, + "grad_norm": 0.5807187113226505, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 2078 + }, + { + "epoch": 0.02079, + "grad_norm": 0.4917157616397027, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 2079 + }, + { + "epoch": 0.0208, + "grad_norm": 0.5388783899386764, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 2080 + }, + { + "epoch": 0.02081, + "grad_norm": 0.6250429891561764, + "learning_rate": 0.003, + "loss": 4.1899, + "step": 2081 + }, + { + "epoch": 0.02082, + "grad_norm": 0.7539524872161132, + "learning_rate": 0.003, + "loss": 4.1879, + "step": 2082 + }, + { + "epoch": 0.02083, + "grad_norm": 0.9021795929932535, + "learning_rate": 0.003, + "loss": 4.189, + "step": 2083 + }, + { + "epoch": 0.02084, + "grad_norm": 0.9892236370325144, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 2084 + }, + { + "epoch": 0.02085, + "grad_norm": 0.9279291435694554, + "learning_rate": 0.003, + "loss": 4.1653, + "step": 2085 + }, + { + "epoch": 0.02086, + "grad_norm": 0.7919688525399468, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 2086 + }, + { + "epoch": 0.02087, + "grad_norm": 0.7965117787009012, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2087 + }, + { + "epoch": 0.02088, + "grad_norm": 0.882456029692519, + "learning_rate": 0.003, + "loss": 4.2097, + "step": 2088 + }, + { + "epoch": 0.02089, + "grad_norm": 0.8843676371103549, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 2089 + }, + { + "epoch": 0.0209, + "grad_norm": 0.9095817089780983, + "learning_rate": 0.003, + "loss": 4.2024, + "step": 2090 + }, + { + "epoch": 0.02091, + "grad_norm": 0.8360707980502976, + "learning_rate": 0.003, + "loss": 4.1884, + "step": 2091 + }, + { + "epoch": 0.02092, + "grad_norm": 0.8954851906550106, + "learning_rate": 0.003, + "loss": 4.189, + "step": 2092 + }, + { + "epoch": 0.02093, + "grad_norm": 0.8146056324857043, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2093 + }, + { + "epoch": 0.02094, + "grad_norm": 0.7558490596404085, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 2094 + }, + { + "epoch": 0.02095, + "grad_norm": 0.7384447141579936, + "learning_rate": 0.003, + "loss": 4.2013, + "step": 2095 + }, + { + "epoch": 0.02096, + "grad_norm": 0.7339651428587286, + "learning_rate": 0.003, + "loss": 4.1917, + "step": 2096 + }, + { + "epoch": 0.02097, + "grad_norm": 0.6671356121778469, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 2097 + }, + { + "epoch": 0.02098, + "grad_norm": 0.653522008057512, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 2098 + }, + { + "epoch": 0.02099, + "grad_norm": 0.6717296847888601, + "learning_rate": 0.003, + "loss": 4.1945, + "step": 2099 + }, + { + "epoch": 0.021, + "grad_norm": 0.6641683049026381, + "learning_rate": 0.003, + "loss": 4.206, + "step": 2100 + }, + { + "epoch": 0.02101, + "grad_norm": 0.8263568144502693, + "learning_rate": 0.003, + "loss": 4.2014, + "step": 2101 + }, + { + "epoch": 0.02102, + "grad_norm": 0.9182856357229072, + "learning_rate": 0.003, + "loss": 4.2116, + "step": 2102 + }, + { + "epoch": 0.02103, + "grad_norm": 0.8775651409452159, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2103 + }, + { + "epoch": 0.02104, + "grad_norm": 0.717248452996101, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 2104 + }, + { + "epoch": 0.02105, + "grad_norm": 0.6147629172749789, + "learning_rate": 0.003, + "loss": 4.1709, + "step": 2105 + }, + { + "epoch": 0.02106, + "grad_norm": 0.5350819274969562, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 2106 + }, + { + "epoch": 0.02107, + "grad_norm": 0.6022860581700974, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 2107 + }, + { + "epoch": 0.02108, + "grad_norm": 0.6033513423948526, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 2108 + }, + { + "epoch": 0.02109, + "grad_norm": 0.6042225440605065, + "learning_rate": 0.003, + "loss": 4.144, + "step": 2109 + }, + { + "epoch": 0.0211, + "grad_norm": 0.560183095906506, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 2110 + }, + { + "epoch": 0.02111, + "grad_norm": 0.5249161566388668, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 2111 + }, + { + "epoch": 0.02112, + "grad_norm": 0.5391312057228763, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 2112 + }, + { + "epoch": 0.02113, + "grad_norm": 0.5260626981895078, + "learning_rate": 0.003, + "loss": 4.1597, + "step": 2113 + }, + { + "epoch": 0.02114, + "grad_norm": 0.5072590511188936, + "learning_rate": 0.003, + "loss": 4.1483, + "step": 2114 + }, + { + "epoch": 0.02115, + "grad_norm": 0.5460141991570172, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 2115 + }, + { + "epoch": 0.02116, + "grad_norm": 0.5749891422597724, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 2116 + }, + { + "epoch": 0.02117, + "grad_norm": 0.6781459947006083, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 2117 + }, + { + "epoch": 0.02118, + "grad_norm": 0.797658364148174, + "learning_rate": 0.003, + "loss": 4.1692, + "step": 2118 + }, + { + "epoch": 0.02119, + "grad_norm": 0.8688435880924688, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 2119 + }, + { + "epoch": 0.0212, + "grad_norm": 0.7833741674522176, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2120 + }, + { + "epoch": 0.02121, + "grad_norm": 0.791532521338522, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 2121 + }, + { + "epoch": 0.02122, + "grad_norm": 0.7913660839492322, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 2122 + }, + { + "epoch": 0.02123, + "grad_norm": 0.7506478823984966, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 2123 + }, + { + "epoch": 0.02124, + "grad_norm": 0.7580020078198468, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 2124 + }, + { + "epoch": 0.02125, + "grad_norm": 0.7006901754929752, + "learning_rate": 0.003, + "loss": 4.1792, + "step": 2125 + }, + { + "epoch": 0.02126, + "grad_norm": 0.7653616945644249, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 2126 + }, + { + "epoch": 0.02127, + "grad_norm": 0.7492928706532664, + "learning_rate": 0.003, + "loss": 4.1859, + "step": 2127 + }, + { + "epoch": 0.02128, + "grad_norm": 0.7795854700268753, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 2128 + }, + { + "epoch": 0.02129, + "grad_norm": 0.9805419044523191, + "learning_rate": 0.003, + "loss": 4.186, + "step": 2129 + }, + { + "epoch": 0.0213, + "grad_norm": 1.0343575534498912, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 2130 + }, + { + "epoch": 0.02131, + "grad_norm": 0.9468793830055084, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 2131 + }, + { + "epoch": 0.02132, + "grad_norm": 0.9021420581140779, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 2132 + }, + { + "epoch": 0.02133, + "grad_norm": 0.8122842455378513, + "learning_rate": 0.003, + "loss": 4.1908, + "step": 2133 + }, + { + "epoch": 0.02134, + "grad_norm": 1.06138323809129, + "learning_rate": 0.003, + "loss": 4.2048, + "step": 2134 + }, + { + "epoch": 0.02135, + "grad_norm": 1.2270031550495697, + "learning_rate": 0.003, + "loss": 4.212, + "step": 2135 + }, + { + "epoch": 0.02136, + "grad_norm": 0.824636655199696, + "learning_rate": 0.003, + "loss": 4.1714, + "step": 2136 + }, + { + "epoch": 0.02137, + "grad_norm": 0.7977636532107003, + "learning_rate": 0.003, + "loss": 4.2074, + "step": 2137 + }, + { + "epoch": 0.02138, + "grad_norm": 0.8175212273276642, + "learning_rate": 0.003, + "loss": 4.2175, + "step": 2138 + }, + { + "epoch": 0.02139, + "grad_norm": 0.8041667304441814, + "learning_rate": 0.003, + "loss": 4.2081, + "step": 2139 + }, + { + "epoch": 0.0214, + "grad_norm": 0.9410719295719912, + "learning_rate": 0.003, + "loss": 4.2078, + "step": 2140 + }, + { + "epoch": 0.02141, + "grad_norm": 0.9128066534114356, + "learning_rate": 0.003, + "loss": 4.1891, + "step": 2141 + }, + { + "epoch": 0.02142, + "grad_norm": 0.7166876786866843, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 2142 + }, + { + "epoch": 0.02143, + "grad_norm": 0.7553648404993994, + "learning_rate": 0.003, + "loss": 4.1879, + "step": 2143 + }, + { + "epoch": 0.02144, + "grad_norm": 0.9693569859499341, + "learning_rate": 0.003, + "loss": 4.1979, + "step": 2144 + }, + { + "epoch": 0.02145, + "grad_norm": 1.005405043081365, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2145 + }, + { + "epoch": 0.02146, + "grad_norm": 0.9174606002788802, + "learning_rate": 0.003, + "loss": 4.1967, + "step": 2146 + }, + { + "epoch": 0.02147, + "grad_norm": 0.7660679608210765, + "learning_rate": 0.003, + "loss": 4.2109, + "step": 2147 + }, + { + "epoch": 0.02148, + "grad_norm": 0.7067244008901828, + "learning_rate": 0.003, + "loss": 4.222, + "step": 2148 + }, + { + "epoch": 0.02149, + "grad_norm": 0.5597922211015285, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 2149 + }, + { + "epoch": 0.0215, + "grad_norm": 0.5887191266363829, + "learning_rate": 0.003, + "loss": 4.1847, + "step": 2150 + }, + { + "epoch": 0.02151, + "grad_norm": 0.5623719175027401, + "learning_rate": 0.003, + "loss": 4.1889, + "step": 2151 + }, + { + "epoch": 0.02152, + "grad_norm": 0.5432643275971011, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 2152 + }, + { + "epoch": 0.02153, + "grad_norm": 0.5447875969469528, + "learning_rate": 0.003, + "loss": 4.1717, + "step": 2153 + }, + { + "epoch": 0.02154, + "grad_norm": 0.64393428597859, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 2154 + }, + { + "epoch": 0.02155, + "grad_norm": 0.836540801508149, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 2155 + }, + { + "epoch": 0.02156, + "grad_norm": 0.9453076222627308, + "learning_rate": 0.003, + "loss": 4.1938, + "step": 2156 + }, + { + "epoch": 0.02157, + "grad_norm": 0.7417036890833684, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 2157 + }, + { + "epoch": 0.02158, + "grad_norm": 0.5957713091881504, + "learning_rate": 0.003, + "loss": 4.1686, + "step": 2158 + }, + { + "epoch": 0.02159, + "grad_norm": 0.6018851123999939, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 2159 + }, + { + "epoch": 0.0216, + "grad_norm": 0.6716376115123553, + "learning_rate": 0.003, + "loss": 4.1831, + "step": 2160 + }, + { + "epoch": 0.02161, + "grad_norm": 0.7819649346338547, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 2161 + }, + { + "epoch": 0.02162, + "grad_norm": 0.8233828232110464, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 2162 + }, + { + "epoch": 0.02163, + "grad_norm": 0.7824568090505709, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 2163 + }, + { + "epoch": 0.02164, + "grad_norm": 0.8597315749146158, + "learning_rate": 0.003, + "loss": 4.1915, + "step": 2164 + }, + { + "epoch": 0.02165, + "grad_norm": 0.7979744235864884, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 2165 + }, + { + "epoch": 0.02166, + "grad_norm": 0.7759936363363831, + "learning_rate": 0.003, + "loss": 4.1947, + "step": 2166 + }, + { + "epoch": 0.02167, + "grad_norm": 0.7097277369414112, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 2167 + }, + { + "epoch": 0.02168, + "grad_norm": 0.6129275664005919, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 2168 + }, + { + "epoch": 0.02169, + "grad_norm": 0.5617202579327312, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 2169 + }, + { + "epoch": 0.0217, + "grad_norm": 0.4729981033232719, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 2170 + }, + { + "epoch": 0.02171, + "grad_norm": 0.5127457916731437, + "learning_rate": 0.003, + "loss": 4.1819, + "step": 2171 + }, + { + "epoch": 0.02172, + "grad_norm": 0.6304303564787295, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 2172 + }, + { + "epoch": 0.02173, + "grad_norm": 0.7235970129567234, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 2173 + }, + { + "epoch": 0.02174, + "grad_norm": 0.9537416454234164, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 2174 + }, + { + "epoch": 0.02175, + "grad_norm": 1.0772501316127283, + "learning_rate": 0.003, + "loss": 4.2082, + "step": 2175 + }, + { + "epoch": 0.02176, + "grad_norm": 0.9320344556958656, + "learning_rate": 0.003, + "loss": 4.1798, + "step": 2176 + }, + { + "epoch": 0.02177, + "grad_norm": 1.2306562612139564, + "learning_rate": 0.003, + "loss": 4.2022, + "step": 2177 + }, + { + "epoch": 0.02178, + "grad_norm": 0.9649462703992993, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 2178 + }, + { + "epoch": 0.02179, + "grad_norm": 0.8364440339114567, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 2179 + }, + { + "epoch": 0.0218, + "grad_norm": 0.8494241782855642, + "learning_rate": 0.003, + "loss": 4.2094, + "step": 2180 + }, + { + "epoch": 0.02181, + "grad_norm": 0.8028341199053393, + "learning_rate": 0.003, + "loss": 4.2103, + "step": 2181 + }, + { + "epoch": 0.02182, + "grad_norm": 0.8789575016731874, + "learning_rate": 0.003, + "loss": 4.1991, + "step": 2182 + }, + { + "epoch": 0.02183, + "grad_norm": 0.7339272838818488, + "learning_rate": 0.003, + "loss": 4.2018, + "step": 2183 + }, + { + "epoch": 0.02184, + "grad_norm": 0.6992146186501603, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 2184 + }, + { + "epoch": 0.02185, + "grad_norm": 0.6859058633396239, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 2185 + }, + { + "epoch": 0.02186, + "grad_norm": 0.671595445918346, + "learning_rate": 0.003, + "loss": 4.186, + "step": 2186 + }, + { + "epoch": 0.02187, + "grad_norm": 0.6535925166183008, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 2187 + }, + { + "epoch": 0.02188, + "grad_norm": 0.6198661745230879, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 2188 + }, + { + "epoch": 0.02189, + "grad_norm": 0.5776255514929863, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 2189 + }, + { + "epoch": 0.0219, + "grad_norm": 0.7433015793364336, + "learning_rate": 0.003, + "loss": 4.1787, + "step": 2190 + }, + { + "epoch": 0.02191, + "grad_norm": 0.8514656528184604, + "learning_rate": 0.003, + "loss": 4.1806, + "step": 2191 + }, + { + "epoch": 0.02192, + "grad_norm": 0.9191452260055095, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 2192 + }, + { + "epoch": 0.02193, + "grad_norm": 0.7405436618584794, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 2193 + }, + { + "epoch": 0.02194, + "grad_norm": 0.7140573914846811, + "learning_rate": 0.003, + "loss": 4.2215, + "step": 2194 + }, + { + "epoch": 0.02195, + "grad_norm": 0.7595143608079454, + "learning_rate": 0.003, + "loss": 4.1859, + "step": 2195 + }, + { + "epoch": 0.02196, + "grad_norm": 0.7549194736868536, + "learning_rate": 0.003, + "loss": 4.1951, + "step": 2196 + }, + { + "epoch": 0.02197, + "grad_norm": 0.6975555957838411, + "learning_rate": 0.003, + "loss": 4.1831, + "step": 2197 + }, + { + "epoch": 0.02198, + "grad_norm": 0.7595090640426224, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2198 + }, + { + "epoch": 0.02199, + "grad_norm": 0.7642539057941399, + "learning_rate": 0.003, + "loss": 4.145, + "step": 2199 + }, + { + "epoch": 0.022, + "grad_norm": 0.7482736596797179, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 2200 + }, + { + "epoch": 0.02201, + "grad_norm": 0.6545018698949331, + "learning_rate": 0.003, + "loss": 4.1824, + "step": 2201 + }, + { + "epoch": 0.02202, + "grad_norm": 0.6269777415335018, + "learning_rate": 0.003, + "loss": 4.1683, + "step": 2202 + }, + { + "epoch": 0.02203, + "grad_norm": 0.7202806939294419, + "learning_rate": 0.003, + "loss": 4.1965, + "step": 2203 + }, + { + "epoch": 0.02204, + "grad_norm": 0.7354393023905447, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 2204 + }, + { + "epoch": 0.02205, + "grad_norm": 0.6658405305239573, + "learning_rate": 0.003, + "loss": 4.1888, + "step": 2205 + }, + { + "epoch": 0.02206, + "grad_norm": 0.5494433484311174, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 2206 + }, + { + "epoch": 0.02207, + "grad_norm": 0.48191881483377974, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 2207 + }, + { + "epoch": 0.02208, + "grad_norm": 0.5055031144775365, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 2208 + }, + { + "epoch": 0.02209, + "grad_norm": 0.5328357421096614, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 2209 + }, + { + "epoch": 0.0221, + "grad_norm": 0.580303567033314, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 2210 + }, + { + "epoch": 0.02211, + "grad_norm": 0.627373370176424, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 2211 + }, + { + "epoch": 0.02212, + "grad_norm": 0.8017723022310775, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 2212 + }, + { + "epoch": 0.02213, + "grad_norm": 0.9755586674162295, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 2213 + }, + { + "epoch": 0.02214, + "grad_norm": 0.8888070506530766, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 2214 + }, + { + "epoch": 0.02215, + "grad_norm": 0.8341073466106758, + "learning_rate": 0.003, + "loss": 4.173, + "step": 2215 + }, + { + "epoch": 0.02216, + "grad_norm": 0.8721558185443227, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 2216 + }, + { + "epoch": 0.02217, + "grad_norm": 0.9058617906836595, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 2217 + }, + { + "epoch": 0.02218, + "grad_norm": 0.8239137730922835, + "learning_rate": 0.003, + "loss": 4.1865, + "step": 2218 + }, + { + "epoch": 0.02219, + "grad_norm": 0.7662188590247174, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 2219 + }, + { + "epoch": 0.0222, + "grad_norm": 0.7276128620873287, + "learning_rate": 0.003, + "loss": 4.1721, + "step": 2220 + }, + { + "epoch": 0.02221, + "grad_norm": 0.8153838699769882, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 2221 + }, + { + "epoch": 0.02222, + "grad_norm": 0.9239982401134911, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 2222 + }, + { + "epoch": 0.02223, + "grad_norm": 0.9248509160902595, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 2223 + }, + { + "epoch": 0.02224, + "grad_norm": 0.8778359853731296, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 2224 + }, + { + "epoch": 0.02225, + "grad_norm": 1.1133790365277945, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 2225 + }, + { + "epoch": 0.02226, + "grad_norm": 0.9357923626997176, + "learning_rate": 0.003, + "loss": 4.1895, + "step": 2226 + }, + { + "epoch": 0.02227, + "grad_norm": 0.8646361690518631, + "learning_rate": 0.003, + "loss": 4.183, + "step": 2227 + }, + { + "epoch": 0.02228, + "grad_norm": 0.938807129740724, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 2228 + }, + { + "epoch": 0.02229, + "grad_norm": 0.8004751977002076, + "learning_rate": 0.003, + "loss": 4.149, + "step": 2229 + }, + { + "epoch": 0.0223, + "grad_norm": 0.7882101721343657, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 2230 + }, + { + "epoch": 0.02231, + "grad_norm": 0.7274707205896823, + "learning_rate": 0.003, + "loss": 4.1896, + "step": 2231 + }, + { + "epoch": 0.02232, + "grad_norm": 0.8291412822325788, + "learning_rate": 0.003, + "loss": 4.2, + "step": 2232 + }, + { + "epoch": 0.02233, + "grad_norm": 0.8063025430085007, + "learning_rate": 0.003, + "loss": 4.166, + "step": 2233 + }, + { + "epoch": 0.02234, + "grad_norm": 0.6911537576664851, + "learning_rate": 0.003, + "loss": 4.1865, + "step": 2234 + }, + { + "epoch": 0.02235, + "grad_norm": 0.7448703016841659, + "learning_rate": 0.003, + "loss": 4.2113, + "step": 2235 + }, + { + "epoch": 0.02236, + "grad_norm": 0.8233234836151326, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 2236 + }, + { + "epoch": 0.02237, + "grad_norm": 1.0093335974280213, + "learning_rate": 0.003, + "loss": 4.2052, + "step": 2237 + }, + { + "epoch": 0.02238, + "grad_norm": 1.1018651279861302, + "learning_rate": 0.003, + "loss": 4.196, + "step": 2238 + }, + { + "epoch": 0.02239, + "grad_norm": 0.7048103914820842, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 2239 + }, + { + "epoch": 0.0224, + "grad_norm": 0.6705866907364436, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 2240 + }, + { + "epoch": 0.02241, + "grad_norm": 0.768034661346772, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 2241 + }, + { + "epoch": 0.02242, + "grad_norm": 0.7013024117659296, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 2242 + }, + { + "epoch": 0.02243, + "grad_norm": 0.8407412837981002, + "learning_rate": 0.003, + "loss": 4.1937, + "step": 2243 + }, + { + "epoch": 0.02244, + "grad_norm": 0.8584973105281422, + "learning_rate": 0.003, + "loss": 4.1731, + "step": 2244 + }, + { + "epoch": 0.02245, + "grad_norm": 0.810849020894934, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 2245 + }, + { + "epoch": 0.02246, + "grad_norm": 0.7441445696162579, + "learning_rate": 0.003, + "loss": 4.191, + "step": 2246 + }, + { + "epoch": 0.02247, + "grad_norm": 0.719895099061681, + "learning_rate": 0.003, + "loss": 4.162, + "step": 2247 + }, + { + "epoch": 0.02248, + "grad_norm": 0.6276605133960077, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 2248 + }, + { + "epoch": 0.02249, + "grad_norm": 0.5820105403717251, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 2249 + }, + { + "epoch": 0.0225, + "grad_norm": 0.494797181910464, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 2250 + }, + { + "epoch": 0.02251, + "grad_norm": 0.45799261155941245, + "learning_rate": 0.003, + "loss": 4.1881, + "step": 2251 + }, + { + "epoch": 0.02252, + "grad_norm": 0.4312165334190654, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 2252 + }, + { + "epoch": 0.02253, + "grad_norm": 0.4089956454491447, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 2253 + }, + { + "epoch": 0.02254, + "grad_norm": 0.41711525641558334, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 2254 + }, + { + "epoch": 0.02255, + "grad_norm": 0.39887312531471364, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 2255 + }, + { + "epoch": 0.02256, + "grad_norm": 0.44922284802782086, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 2256 + }, + { + "epoch": 0.02257, + "grad_norm": 0.43881289908027915, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 2257 + }, + { + "epoch": 0.02258, + "grad_norm": 0.42711858762354266, + "learning_rate": 0.003, + "loss": 4.1569, + "step": 2258 + }, + { + "epoch": 0.02259, + "grad_norm": 0.4761674249696505, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 2259 + }, + { + "epoch": 0.0226, + "grad_norm": 0.6172223177216527, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 2260 + }, + { + "epoch": 0.02261, + "grad_norm": 0.9969811836278653, + "learning_rate": 0.003, + "loss": 4.1626, + "step": 2261 + }, + { + "epoch": 0.02262, + "grad_norm": 1.4207495716773118, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 2262 + }, + { + "epoch": 0.02263, + "grad_norm": 0.5107196284950156, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 2263 + }, + { + "epoch": 0.02264, + "grad_norm": 0.8124315452603879, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 2264 + }, + { + "epoch": 0.02265, + "grad_norm": 0.7127120804095198, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 2265 + }, + { + "epoch": 0.02266, + "grad_norm": 0.5120142831669139, + "learning_rate": 0.003, + "loss": 4.142, + "step": 2266 + }, + { + "epoch": 0.02267, + "grad_norm": 0.842715736214197, + "learning_rate": 0.003, + "loss": 4.1775, + "step": 2267 + }, + { + "epoch": 0.02268, + "grad_norm": 0.8370937405428989, + "learning_rate": 0.003, + "loss": 4.1593, + "step": 2268 + }, + { + "epoch": 0.02269, + "grad_norm": 0.7173617092756458, + "learning_rate": 0.003, + "loss": 4.174, + "step": 2269 + }, + { + "epoch": 0.0227, + "grad_norm": 0.8926880986379259, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 2270 + }, + { + "epoch": 0.02271, + "grad_norm": 0.9394293424718215, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 2271 + }, + { + "epoch": 0.02272, + "grad_norm": 0.9564097717489924, + "learning_rate": 0.003, + "loss": 4.1878, + "step": 2272 + }, + { + "epoch": 0.02273, + "grad_norm": 1.2112473190186792, + "learning_rate": 0.003, + "loss": 4.1998, + "step": 2273 + }, + { + "epoch": 0.02274, + "grad_norm": 0.9744415053378092, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 2274 + }, + { + "epoch": 0.02275, + "grad_norm": 1.0078841940550498, + "learning_rate": 0.003, + "loss": 4.1796, + "step": 2275 + }, + { + "epoch": 0.02276, + "grad_norm": 0.8300538512961302, + "learning_rate": 0.003, + "loss": 4.1696, + "step": 2276 + }, + { + "epoch": 0.02277, + "grad_norm": 0.6643966250880426, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2277 + }, + { + "epoch": 0.02278, + "grad_norm": 0.6190627183036216, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 2278 + }, + { + "epoch": 0.02279, + "grad_norm": 0.551131169249495, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 2279 + }, + { + "epoch": 0.0228, + "grad_norm": 0.5572806608313633, + "learning_rate": 0.003, + "loss": 4.173, + "step": 2280 + }, + { + "epoch": 0.02281, + "grad_norm": 0.630640256246517, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 2281 + }, + { + "epoch": 0.02282, + "grad_norm": 0.667161853354935, + "learning_rate": 0.003, + "loss": 4.181, + "step": 2282 + }, + { + "epoch": 0.02283, + "grad_norm": 0.7804313255347314, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 2283 + }, + { + "epoch": 0.02284, + "grad_norm": 0.8174062173017085, + "learning_rate": 0.003, + "loss": 4.1998, + "step": 2284 + }, + { + "epoch": 0.02285, + "grad_norm": 0.8933970302021632, + "learning_rate": 0.003, + "loss": 4.147, + "step": 2285 + }, + { + "epoch": 0.02286, + "grad_norm": 0.8583328769415308, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 2286 + }, + { + "epoch": 0.02287, + "grad_norm": 0.8734810759184407, + "learning_rate": 0.003, + "loss": 4.1855, + "step": 2287 + }, + { + "epoch": 0.02288, + "grad_norm": 0.8928624064077537, + "learning_rate": 0.003, + "loss": 4.2123, + "step": 2288 + }, + { + "epoch": 0.02289, + "grad_norm": 0.783879288737337, + "learning_rate": 0.003, + "loss": 4.2056, + "step": 2289 + }, + { + "epoch": 0.0229, + "grad_norm": 0.7701279568768081, + "learning_rate": 0.003, + "loss": 4.1809, + "step": 2290 + }, + { + "epoch": 0.02291, + "grad_norm": 0.7353904143222354, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 2291 + }, + { + "epoch": 0.02292, + "grad_norm": 0.5673182764651727, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 2292 + }, + { + "epoch": 0.02293, + "grad_norm": 0.6211445071442583, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 2293 + }, + { + "epoch": 0.02294, + "grad_norm": 0.5897432424358218, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 2294 + }, + { + "epoch": 0.02295, + "grad_norm": 0.6701151196414159, + "learning_rate": 0.003, + "loss": 4.147, + "step": 2295 + }, + { + "epoch": 0.02296, + "grad_norm": 1.0101216354387503, + "learning_rate": 0.003, + "loss": 4.1897, + "step": 2296 + }, + { + "epoch": 0.02297, + "grad_norm": 1.2406081319623274, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 2297 + }, + { + "epoch": 0.02298, + "grad_norm": 0.5988120231143333, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2298 + }, + { + "epoch": 0.02299, + "grad_norm": 0.7795860072160454, + "learning_rate": 0.003, + "loss": 4.1804, + "step": 2299 + }, + { + "epoch": 0.023, + "grad_norm": 0.784959258488215, + "learning_rate": 0.003, + "loss": 4.1984, + "step": 2300 + }, + { + "epoch": 0.02301, + "grad_norm": 0.5879362613709453, + "learning_rate": 0.003, + "loss": 4.157, + "step": 2301 + }, + { + "epoch": 0.02302, + "grad_norm": 0.7106606864619079, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 2302 + }, + { + "epoch": 0.02303, + "grad_norm": 0.7402995000062723, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2303 + }, + { + "epoch": 0.02304, + "grad_norm": 0.8291546359347778, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 2304 + }, + { + "epoch": 0.02305, + "grad_norm": 0.753148855618092, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 2305 + }, + { + "epoch": 0.02306, + "grad_norm": 0.8869246970763861, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 2306 + }, + { + "epoch": 0.02307, + "grad_norm": 0.8925578818662347, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 2307 + }, + { + "epoch": 0.02308, + "grad_norm": 0.9066759173397617, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 2308 + }, + { + "epoch": 0.02309, + "grad_norm": 0.8201889707383639, + "learning_rate": 0.003, + "loss": 4.1952, + "step": 2309 + }, + { + "epoch": 0.0231, + "grad_norm": 0.7904644702923331, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 2310 + }, + { + "epoch": 0.02311, + "grad_norm": 0.7060518287169474, + "learning_rate": 0.003, + "loss": 4.1776, + "step": 2311 + }, + { + "epoch": 0.02312, + "grad_norm": 0.7392874926033767, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 2312 + }, + { + "epoch": 0.02313, + "grad_norm": 0.8538999440765905, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 2313 + }, + { + "epoch": 0.02314, + "grad_norm": 0.9312222422709578, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 2314 + }, + { + "epoch": 0.02315, + "grad_norm": 0.9687214234200097, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 2315 + }, + { + "epoch": 0.02316, + "grad_norm": 0.891697328396861, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 2316 + }, + { + "epoch": 0.02317, + "grad_norm": 0.897934373295407, + "learning_rate": 0.003, + "loss": 4.1753, + "step": 2317 + }, + { + "epoch": 0.02318, + "grad_norm": 0.8051331418091766, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 2318 + }, + { + "epoch": 0.02319, + "grad_norm": 0.703806089932542, + "learning_rate": 0.003, + "loss": 4.1804, + "step": 2319 + }, + { + "epoch": 0.0232, + "grad_norm": 0.7388332818016227, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 2320 + }, + { + "epoch": 0.02321, + "grad_norm": 0.7450344205679207, + "learning_rate": 0.003, + "loss": 4.1692, + "step": 2321 + }, + { + "epoch": 0.02322, + "grad_norm": 0.6660500555618327, + "learning_rate": 0.003, + "loss": 4.1746, + "step": 2322 + }, + { + "epoch": 0.02323, + "grad_norm": 0.6571661630697367, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 2323 + }, + { + "epoch": 0.02324, + "grad_norm": 0.6161646162269099, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 2324 + }, + { + "epoch": 0.02325, + "grad_norm": 0.5919512249258755, + "learning_rate": 0.003, + "loss": 4.1679, + "step": 2325 + }, + { + "epoch": 0.02326, + "grad_norm": 0.7224558633075708, + "learning_rate": 0.003, + "loss": 4.1831, + "step": 2326 + }, + { + "epoch": 0.02327, + "grad_norm": 0.8816178665411202, + "learning_rate": 0.003, + "loss": 4.1759, + "step": 2327 + }, + { + "epoch": 0.02328, + "grad_norm": 0.8699762426391761, + "learning_rate": 0.003, + "loss": 4.176, + "step": 2328 + }, + { + "epoch": 0.02329, + "grad_norm": 0.723960245924297, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 2329 + }, + { + "epoch": 0.0233, + "grad_norm": 0.7628023637683098, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 2330 + }, + { + "epoch": 0.02331, + "grad_norm": 0.7695226471244487, + "learning_rate": 0.003, + "loss": 4.1933, + "step": 2331 + }, + { + "epoch": 0.02332, + "grad_norm": 0.7712606828136085, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 2332 + }, + { + "epoch": 0.02333, + "grad_norm": 0.7181228868994012, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 2333 + }, + { + "epoch": 0.02334, + "grad_norm": 0.6385757918578437, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 2334 + }, + { + "epoch": 0.02335, + "grad_norm": 0.5804913608133354, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 2335 + }, + { + "epoch": 0.02336, + "grad_norm": 0.6406948035907543, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 2336 + }, + { + "epoch": 0.02337, + "grad_norm": 0.7550291811921531, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 2337 + }, + { + "epoch": 0.02338, + "grad_norm": 0.867264165349628, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 2338 + }, + { + "epoch": 0.02339, + "grad_norm": 0.9370869698358649, + "learning_rate": 0.003, + "loss": 4.1596, + "step": 2339 + }, + { + "epoch": 0.0234, + "grad_norm": 1.04114552861465, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 2340 + }, + { + "epoch": 0.02341, + "grad_norm": 0.8419443125713769, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 2341 + }, + { + "epoch": 0.02342, + "grad_norm": 0.7516609856190137, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 2342 + }, + { + "epoch": 0.02343, + "grad_norm": 0.7840218113502367, + "learning_rate": 0.003, + "loss": 4.1597, + "step": 2343 + }, + { + "epoch": 0.02344, + "grad_norm": 0.7106909207560747, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 2344 + }, + { + "epoch": 0.02345, + "grad_norm": 0.695743479378358, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 2345 + }, + { + "epoch": 0.02346, + "grad_norm": 0.7618556711999727, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 2346 + }, + { + "epoch": 0.02347, + "grad_norm": 0.8494779389825688, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 2347 + }, + { + "epoch": 0.02348, + "grad_norm": 0.8819304188600076, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 2348 + }, + { + "epoch": 0.02349, + "grad_norm": 0.801874189029763, + "learning_rate": 0.003, + "loss": 4.1729, + "step": 2349 + }, + { + "epoch": 0.0235, + "grad_norm": 0.7502001194200574, + "learning_rate": 0.003, + "loss": 4.168, + "step": 2350 + }, + { + "epoch": 0.02351, + "grad_norm": 0.7987799293083814, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 2351 + }, + { + "epoch": 0.02352, + "grad_norm": 0.9306705411998242, + "learning_rate": 0.003, + "loss": 4.142, + "step": 2352 + }, + { + "epoch": 0.02353, + "grad_norm": 0.8564501775467426, + "learning_rate": 0.003, + "loss": 4.1759, + "step": 2353 + }, + { + "epoch": 0.02354, + "grad_norm": 0.8345109850380096, + "learning_rate": 0.003, + "loss": 4.1813, + "step": 2354 + }, + { + "epoch": 0.02355, + "grad_norm": 0.7887172745864481, + "learning_rate": 0.003, + "loss": 4.1723, + "step": 2355 + }, + { + "epoch": 0.02356, + "grad_norm": 0.6582873563724112, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 2356 + }, + { + "epoch": 0.02357, + "grad_norm": 0.6737457349000578, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 2357 + }, + { + "epoch": 0.02358, + "grad_norm": 0.7705872627309706, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 2358 + }, + { + "epoch": 0.02359, + "grad_norm": 0.8542556066081018, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 2359 + }, + { + "epoch": 0.0236, + "grad_norm": 0.9842191585501165, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 2360 + }, + { + "epoch": 0.02361, + "grad_norm": 0.9112511096880844, + "learning_rate": 0.003, + "loss": 4.1793, + "step": 2361 + }, + { + "epoch": 0.02362, + "grad_norm": 0.9278874197656558, + "learning_rate": 0.003, + "loss": 4.1712, + "step": 2362 + }, + { + "epoch": 0.02363, + "grad_norm": 0.7482616982968588, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 2363 + }, + { + "epoch": 0.02364, + "grad_norm": 0.5981870727991454, + "learning_rate": 0.003, + "loss": 4.1597, + "step": 2364 + }, + { + "epoch": 0.02365, + "grad_norm": 0.6120042832874062, + "learning_rate": 0.003, + "loss": 4.1788, + "step": 2365 + }, + { + "epoch": 0.02366, + "grad_norm": 0.592965390221233, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 2366 + }, + { + "epoch": 0.02367, + "grad_norm": 0.5967931209483315, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 2367 + }, + { + "epoch": 0.02368, + "grad_norm": 0.6289890460021353, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 2368 + }, + { + "epoch": 0.02369, + "grad_norm": 0.6580991252391354, + "learning_rate": 0.003, + "loss": 4.1441, + "step": 2369 + }, + { + "epoch": 0.0237, + "grad_norm": 0.6945790129958274, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 2370 + }, + { + "epoch": 0.02371, + "grad_norm": 0.7107030962570803, + "learning_rate": 0.003, + "loss": 4.1569, + "step": 2371 + }, + { + "epoch": 0.02372, + "grad_norm": 0.8059296050081108, + "learning_rate": 0.003, + "loss": 4.149, + "step": 2372 + }, + { + "epoch": 0.02373, + "grad_norm": 0.921257026362051, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 2373 + }, + { + "epoch": 0.02374, + "grad_norm": 0.955728704961619, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2374 + }, + { + "epoch": 0.02375, + "grad_norm": 0.7504286137015724, + "learning_rate": 0.003, + "loss": 4.15, + "step": 2375 + }, + { + "epoch": 0.02376, + "grad_norm": 0.7478781963883802, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 2376 + }, + { + "epoch": 0.02377, + "grad_norm": 0.917428121872858, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 2377 + }, + { + "epoch": 0.02378, + "grad_norm": 0.7822980127349399, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 2378 + }, + { + "epoch": 0.02379, + "grad_norm": 0.5863463150077827, + "learning_rate": 0.003, + "loss": 4.136, + "step": 2379 + }, + { + "epoch": 0.0238, + "grad_norm": 0.5340016137538751, + "learning_rate": 0.003, + "loss": 4.109, + "step": 2380 + }, + { + "epoch": 0.02381, + "grad_norm": 0.6128857312553131, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 2381 + }, + { + "epoch": 0.02382, + "grad_norm": 0.5658688824938058, + "learning_rate": 0.003, + "loss": 4.1779, + "step": 2382 + }, + { + "epoch": 0.02383, + "grad_norm": 0.5676379045904221, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 2383 + }, + { + "epoch": 0.02384, + "grad_norm": 0.5238096521489247, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 2384 + }, + { + "epoch": 0.02385, + "grad_norm": 0.5152751390365298, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 2385 + }, + { + "epoch": 0.02386, + "grad_norm": 0.6091514932256747, + "learning_rate": 0.003, + "loss": 4.13, + "step": 2386 + }, + { + "epoch": 0.02387, + "grad_norm": 0.7793599922586826, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 2387 + }, + { + "epoch": 0.02388, + "grad_norm": 0.9352617435026656, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 2388 + }, + { + "epoch": 0.02389, + "grad_norm": 0.9609729763435508, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 2389 + }, + { + "epoch": 0.0239, + "grad_norm": 0.8094018370976185, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 2390 + }, + { + "epoch": 0.02391, + "grad_norm": 0.8506283143692317, + "learning_rate": 0.003, + "loss": 4.1734, + "step": 2391 + }, + { + "epoch": 0.02392, + "grad_norm": 0.9050332115497783, + "learning_rate": 0.003, + "loss": 4.1803, + "step": 2392 + }, + { + "epoch": 0.02393, + "grad_norm": 0.9634890728744777, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 2393 + }, + { + "epoch": 0.02394, + "grad_norm": 0.8882373157340075, + "learning_rate": 0.003, + "loss": 4.1569, + "step": 2394 + }, + { + "epoch": 0.02395, + "grad_norm": 0.7565332153535935, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 2395 + }, + { + "epoch": 0.02396, + "grad_norm": 0.8603458435447788, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 2396 + }, + { + "epoch": 0.02397, + "grad_norm": 0.8261263361309694, + "learning_rate": 0.003, + "loss": 4.169, + "step": 2397 + }, + { + "epoch": 0.02398, + "grad_norm": 0.7794743587492478, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 2398 + }, + { + "epoch": 0.02399, + "grad_norm": 0.8140982417839558, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 2399 + }, + { + "epoch": 0.024, + "grad_norm": 0.9091203275980858, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2400 + }, + { + "epoch": 0.02401, + "grad_norm": 0.9082248326953591, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 2401 + }, + { + "epoch": 0.02402, + "grad_norm": 0.8996640369589112, + "learning_rate": 0.003, + "loss": 4.1798, + "step": 2402 + }, + { + "epoch": 0.02403, + "grad_norm": 0.9493494458526733, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 2403 + }, + { + "epoch": 0.02404, + "grad_norm": 0.9827368493949332, + "learning_rate": 0.003, + "loss": 4.1965, + "step": 2404 + }, + { + "epoch": 0.02405, + "grad_norm": 0.9551905621174913, + "learning_rate": 0.003, + "loss": 4.2021, + "step": 2405 + }, + { + "epoch": 0.02406, + "grad_norm": 0.9364752365255031, + "learning_rate": 0.003, + "loss": 4.1854, + "step": 2406 + }, + { + "epoch": 0.02407, + "grad_norm": 0.9463612550147267, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 2407 + }, + { + "epoch": 0.02408, + "grad_norm": 0.8719994888631714, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 2408 + }, + { + "epoch": 0.02409, + "grad_norm": 0.864310092393717, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 2409 + }, + { + "epoch": 0.0241, + "grad_norm": 0.8973500607718927, + "learning_rate": 0.003, + "loss": 4.1883, + "step": 2410 + }, + { + "epoch": 0.02411, + "grad_norm": 1.1693499034787715, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 2411 + }, + { + "epoch": 0.02412, + "grad_norm": 0.8431604098452435, + "learning_rate": 0.003, + "loss": 4.164, + "step": 2412 + }, + { + "epoch": 0.02413, + "grad_norm": 0.7767055362680262, + "learning_rate": 0.003, + "loss": 4.1597, + "step": 2413 + }, + { + "epoch": 0.02414, + "grad_norm": 0.8330306566206577, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 2414 + }, + { + "epoch": 0.02415, + "grad_norm": 0.8217708926095696, + "learning_rate": 0.003, + "loss": 4.1765, + "step": 2415 + }, + { + "epoch": 0.02416, + "grad_norm": 0.8597994699370953, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 2416 + }, + { + "epoch": 0.02417, + "grad_norm": 0.8398506547738631, + "learning_rate": 0.003, + "loss": 4.1829, + "step": 2417 + }, + { + "epoch": 0.02418, + "grad_norm": 0.7704254084685477, + "learning_rate": 0.003, + "loss": 4.1714, + "step": 2418 + }, + { + "epoch": 0.02419, + "grad_norm": 0.752617109928811, + "learning_rate": 0.003, + "loss": 4.1691, + "step": 2419 + }, + { + "epoch": 0.0242, + "grad_norm": 0.681614467806083, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 2420 + }, + { + "epoch": 0.02421, + "grad_norm": 0.57806703718961, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 2421 + }, + { + "epoch": 0.02422, + "grad_norm": 0.5368829917602504, + "learning_rate": 0.003, + "loss": 4.1904, + "step": 2422 + }, + { + "epoch": 0.02423, + "grad_norm": 0.49926017966447833, + "learning_rate": 0.003, + "loss": 4.1767, + "step": 2423 + }, + { + "epoch": 0.02424, + "grad_norm": 0.44538511553678933, + "learning_rate": 0.003, + "loss": 4.1709, + "step": 2424 + }, + { + "epoch": 0.02425, + "grad_norm": 0.4455623217366007, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 2425 + }, + { + "epoch": 0.02426, + "grad_norm": 0.45034256565793024, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 2426 + }, + { + "epoch": 0.02427, + "grad_norm": 0.4277494354911392, + "learning_rate": 0.003, + "loss": 4.1831, + "step": 2427 + }, + { + "epoch": 0.02428, + "grad_norm": 0.49536930851083055, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 2428 + }, + { + "epoch": 0.02429, + "grad_norm": 0.5349623201813287, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 2429 + }, + { + "epoch": 0.0243, + "grad_norm": 0.6385700511917229, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 2430 + }, + { + "epoch": 0.02431, + "grad_norm": 0.7962007873904552, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 2431 + }, + { + "epoch": 0.02432, + "grad_norm": 0.7552982964141086, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 2432 + }, + { + "epoch": 0.02433, + "grad_norm": 0.5576758209292167, + "learning_rate": 0.003, + "loss": 4.179, + "step": 2433 + }, + { + "epoch": 0.02434, + "grad_norm": 0.5731742455027719, + "learning_rate": 0.003, + "loss": 4.1457, + "step": 2434 + }, + { + "epoch": 0.02435, + "grad_norm": 0.6782357026569522, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 2435 + }, + { + "epoch": 0.02436, + "grad_norm": 0.8071855900106416, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 2436 + }, + { + "epoch": 0.02437, + "grad_norm": 0.9265059909484438, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 2437 + }, + { + "epoch": 0.02438, + "grad_norm": 0.8793214152171194, + "learning_rate": 0.003, + "loss": 4.1644, + "step": 2438 + }, + { + "epoch": 0.02439, + "grad_norm": 0.8832749976596599, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 2439 + }, + { + "epoch": 0.0244, + "grad_norm": 0.9746984844718865, + "learning_rate": 0.003, + "loss": 4.1934, + "step": 2440 + }, + { + "epoch": 0.02441, + "grad_norm": 1.0708150303834312, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 2441 + }, + { + "epoch": 0.02442, + "grad_norm": 1.040411975678819, + "learning_rate": 0.003, + "loss": 4.2014, + "step": 2442 + }, + { + "epoch": 0.02443, + "grad_norm": 1.157601705274513, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 2443 + }, + { + "epoch": 0.02444, + "grad_norm": 0.8998517717034221, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 2444 + }, + { + "epoch": 0.02445, + "grad_norm": 0.7914422604882547, + "learning_rate": 0.003, + "loss": 4.1691, + "step": 2445 + }, + { + "epoch": 0.02446, + "grad_norm": 0.765570533672039, + "learning_rate": 0.003, + "loss": 4.168, + "step": 2446 + }, + { + "epoch": 0.02447, + "grad_norm": 0.7780408804262848, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 2447 + }, + { + "epoch": 0.02448, + "grad_norm": 0.8355663969723807, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 2448 + }, + { + "epoch": 0.02449, + "grad_norm": 0.8911004412381984, + "learning_rate": 0.003, + "loss": 4.173, + "step": 2449 + }, + { + "epoch": 0.0245, + "grad_norm": 0.8721571131136453, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 2450 + }, + { + "epoch": 0.02451, + "grad_norm": 0.8424855685228627, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 2451 + }, + { + "epoch": 0.02452, + "grad_norm": 0.71343486564028, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 2452 + }, + { + "epoch": 0.02453, + "grad_norm": 0.830794139281049, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 2453 + }, + { + "epoch": 0.02454, + "grad_norm": 0.8371964018463887, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 2454 + }, + { + "epoch": 0.02455, + "grad_norm": 0.828173261775671, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 2455 + }, + { + "epoch": 0.02456, + "grad_norm": 0.8111209228027585, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 2456 + }, + { + "epoch": 0.02457, + "grad_norm": 0.7593406743031937, + "learning_rate": 0.003, + "loss": 4.175, + "step": 2457 + }, + { + "epoch": 0.02458, + "grad_norm": 0.6152494042836864, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 2458 + }, + { + "epoch": 0.02459, + "grad_norm": 0.6726714704776762, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 2459 + }, + { + "epoch": 0.0246, + "grad_norm": 0.6849708403280691, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 2460 + }, + { + "epoch": 0.02461, + "grad_norm": 0.6926478544751591, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 2461 + }, + { + "epoch": 0.02462, + "grad_norm": 0.6546290518877314, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 2462 + }, + { + "epoch": 0.02463, + "grad_norm": 0.6160770029093054, + "learning_rate": 0.003, + "loss": 4.1691, + "step": 2463 + }, + { + "epoch": 0.02464, + "grad_norm": 0.5990408684539491, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 2464 + }, + { + "epoch": 0.02465, + "grad_norm": 0.5034148559800348, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 2465 + }, + { + "epoch": 0.02466, + "grad_norm": 0.4860347114545619, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 2466 + }, + { + "epoch": 0.02467, + "grad_norm": 0.4781136004723512, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 2467 + }, + { + "epoch": 0.02468, + "grad_norm": 0.4265902063645912, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 2468 + }, + { + "epoch": 0.02469, + "grad_norm": 0.5508562887056804, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 2469 + }, + { + "epoch": 0.0247, + "grad_norm": 0.8249536754461567, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 2470 + }, + { + "epoch": 0.02471, + "grad_norm": 1.4359640347930054, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 2471 + }, + { + "epoch": 0.02472, + "grad_norm": 0.6769644869683711, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 2472 + }, + { + "epoch": 0.02473, + "grad_norm": 0.7658209871323785, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 2473 + }, + { + "epoch": 0.02474, + "grad_norm": 0.9511612267134097, + "learning_rate": 0.003, + "loss": 4.1669, + "step": 2474 + }, + { + "epoch": 0.02475, + "grad_norm": 0.8892227903889872, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 2475 + }, + { + "epoch": 0.02476, + "grad_norm": 0.9920143400330694, + "learning_rate": 0.003, + "loss": 4.152, + "step": 2476 + }, + { + "epoch": 0.02477, + "grad_norm": 0.9006606413119249, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 2477 + }, + { + "epoch": 0.02478, + "grad_norm": 0.8088690703733978, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 2478 + }, + { + "epoch": 0.02479, + "grad_norm": 0.8182856999501472, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 2479 + }, + { + "epoch": 0.0248, + "grad_norm": 0.7355042387960451, + "learning_rate": 0.003, + "loss": 4.178, + "step": 2480 + }, + { + "epoch": 0.02481, + "grad_norm": 0.6962248057182712, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 2481 + }, + { + "epoch": 0.02482, + "grad_norm": 0.6352228705173556, + "learning_rate": 0.003, + "loss": 4.1666, + "step": 2482 + }, + { + "epoch": 0.02483, + "grad_norm": 0.6906094612236757, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 2483 + }, + { + "epoch": 0.02484, + "grad_norm": 0.66658401463766, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 2484 + }, + { + "epoch": 0.02485, + "grad_norm": 0.6596750859147653, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 2485 + }, + { + "epoch": 0.02486, + "grad_norm": 0.7066148867713012, + "learning_rate": 0.003, + "loss": 4.1929, + "step": 2486 + }, + { + "epoch": 0.02487, + "grad_norm": 0.754767630208959, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 2487 + }, + { + "epoch": 0.02488, + "grad_norm": 0.7178651369815516, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 2488 + }, + { + "epoch": 0.02489, + "grad_norm": 0.7388633013067267, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 2489 + }, + { + "epoch": 0.0249, + "grad_norm": 0.7802405314573639, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 2490 + }, + { + "epoch": 0.02491, + "grad_norm": 0.6722919878141813, + "learning_rate": 0.003, + "loss": 4.147, + "step": 2491 + }, + { + "epoch": 0.02492, + "grad_norm": 0.6231446299505073, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 2492 + }, + { + "epoch": 0.02493, + "grad_norm": 0.6055678621905288, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 2493 + }, + { + "epoch": 0.02494, + "grad_norm": 0.6270135330055409, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 2494 + }, + { + "epoch": 0.02495, + "grad_norm": 0.6525517462302536, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 2495 + }, + { + "epoch": 0.02496, + "grad_norm": 0.7329319033542416, + "learning_rate": 0.003, + "loss": 4.14, + "step": 2496 + }, + { + "epoch": 0.02497, + "grad_norm": 0.9619996867887689, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 2497 + }, + { + "epoch": 0.02498, + "grad_norm": 1.141814403141371, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2498 + }, + { + "epoch": 0.02499, + "grad_norm": 0.9016540268347464, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 2499 + }, + { + "epoch": 0.025, + "grad_norm": 0.9065801171881926, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 2500 + }, + { + "epoch": 0.02501, + "grad_norm": 0.8844338776736929, + "learning_rate": 0.003, + "loss": 4.1765, + "step": 2501 + }, + { + "epoch": 0.02502, + "grad_norm": 0.8046602370535848, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 2502 + }, + { + "epoch": 0.02503, + "grad_norm": 0.7039592913476108, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 2503 + }, + { + "epoch": 0.02504, + "grad_norm": 0.7419132370317111, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 2504 + }, + { + "epoch": 0.02505, + "grad_norm": 0.7606498464541366, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 2505 + }, + { + "epoch": 0.02506, + "grad_norm": 0.769560412914919, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 2506 + }, + { + "epoch": 0.02507, + "grad_norm": 0.8361991263151639, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 2507 + }, + { + "epoch": 0.02508, + "grad_norm": 1.0314436698214402, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 2508 + }, + { + "epoch": 0.02509, + "grad_norm": 1.0808307469983176, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 2509 + }, + { + "epoch": 0.0251, + "grad_norm": 0.9756326578065097, + "learning_rate": 0.003, + "loss": 4.1772, + "step": 2510 + }, + { + "epoch": 0.02511, + "grad_norm": 0.9782444756269447, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2511 + }, + { + "epoch": 0.02512, + "grad_norm": 0.9679666828238646, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 2512 + }, + { + "epoch": 0.02513, + "grad_norm": 0.9141144321737233, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 2513 + }, + { + "epoch": 0.02514, + "grad_norm": 0.870311596191751, + "learning_rate": 0.003, + "loss": 4.1521, + "step": 2514 + }, + { + "epoch": 0.02515, + "grad_norm": 0.718099567068549, + "learning_rate": 0.003, + "loss": 4.134, + "step": 2515 + }, + { + "epoch": 0.02516, + "grad_norm": 0.6485758792814954, + "learning_rate": 0.003, + "loss": 4.153, + "step": 2516 + }, + { + "epoch": 0.02517, + "grad_norm": 0.6631125888262853, + "learning_rate": 0.003, + "loss": 4.1528, + "step": 2517 + }, + { + "epoch": 0.02518, + "grad_norm": 0.62508603256109, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 2518 + }, + { + "epoch": 0.02519, + "grad_norm": 0.6533648513332591, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 2519 + }, + { + "epoch": 0.0252, + "grad_norm": 0.8730883164618287, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 2520 + }, + { + "epoch": 0.02521, + "grad_norm": 1.1508715959612579, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 2521 + }, + { + "epoch": 0.02522, + "grad_norm": 0.8851560060677419, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 2522 + }, + { + "epoch": 0.02523, + "grad_norm": 0.7609503224943213, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2523 + }, + { + "epoch": 0.02524, + "grad_norm": 0.7874722612084045, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 2524 + }, + { + "epoch": 0.02525, + "grad_norm": 0.9497267581795765, + "learning_rate": 0.003, + "loss": 4.1834, + "step": 2525 + }, + { + "epoch": 0.02526, + "grad_norm": 1.1487960136042998, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 2526 + }, + { + "epoch": 0.02527, + "grad_norm": 0.749631180786954, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 2527 + }, + { + "epoch": 0.02528, + "grad_norm": 0.6717432753530606, + "learning_rate": 0.003, + "loss": 4.1816, + "step": 2528 + }, + { + "epoch": 0.02529, + "grad_norm": 0.6421573788772453, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 2529 + }, + { + "epoch": 0.0253, + "grad_norm": 0.7486612963376212, + "learning_rate": 0.003, + "loss": 4.1689, + "step": 2530 + }, + { + "epoch": 0.02531, + "grad_norm": 0.7949983435688287, + "learning_rate": 0.003, + "loss": 4.171, + "step": 2531 + }, + { + "epoch": 0.02532, + "grad_norm": 0.7618621411023159, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 2532 + }, + { + "epoch": 0.02533, + "grad_norm": 0.8123389715874717, + "learning_rate": 0.003, + "loss": 4.1773, + "step": 2533 + }, + { + "epoch": 0.02534, + "grad_norm": 0.808321322590649, + "learning_rate": 0.003, + "loss": 4.1573, + "step": 2534 + }, + { + "epoch": 0.02535, + "grad_norm": 0.7051467031412186, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 2535 + }, + { + "epoch": 0.02536, + "grad_norm": 0.6093069610094953, + "learning_rate": 0.003, + "loss": 4.142, + "step": 2536 + }, + { + "epoch": 0.02537, + "grad_norm": 0.5843395517845005, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 2537 + }, + { + "epoch": 0.02538, + "grad_norm": 0.5674485238862751, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 2538 + }, + { + "epoch": 0.02539, + "grad_norm": 0.5652766007864589, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 2539 + }, + { + "epoch": 0.0254, + "grad_norm": 0.6570254628858063, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 2540 + }, + { + "epoch": 0.02541, + "grad_norm": 0.8206242142708078, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 2541 + }, + { + "epoch": 0.02542, + "grad_norm": 0.9677168967962686, + "learning_rate": 0.003, + "loss": 4.134, + "step": 2542 + }, + { + "epoch": 0.02543, + "grad_norm": 1.024805521305044, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 2543 + }, + { + "epoch": 0.02544, + "grad_norm": 0.7591855178067561, + "learning_rate": 0.003, + "loss": 4.148, + "step": 2544 + }, + { + "epoch": 0.02545, + "grad_norm": 0.6449474624868338, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 2545 + }, + { + "epoch": 0.02546, + "grad_norm": 0.6980586139707194, + "learning_rate": 0.003, + "loss": 4.136, + "step": 2546 + }, + { + "epoch": 0.02547, + "grad_norm": 0.7283106303300475, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 2547 + }, + { + "epoch": 0.02548, + "grad_norm": 0.6850467395404413, + "learning_rate": 0.003, + "loss": 4.1691, + "step": 2548 + }, + { + "epoch": 0.02549, + "grad_norm": 0.7158743790784491, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 2549 + }, + { + "epoch": 0.0255, + "grad_norm": 0.799200990441851, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 2550 + }, + { + "epoch": 0.02551, + "grad_norm": 0.8852033835267075, + "learning_rate": 0.003, + "loss": 4.165, + "step": 2551 + }, + { + "epoch": 0.02552, + "grad_norm": 0.9381299460340671, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 2552 + }, + { + "epoch": 0.02553, + "grad_norm": 0.8787078285398282, + "learning_rate": 0.003, + "loss": 4.2019, + "step": 2553 + }, + { + "epoch": 0.02554, + "grad_norm": 1.0064349811006572, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 2554 + }, + { + "epoch": 0.02555, + "grad_norm": 0.8623285229523936, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 2555 + }, + { + "epoch": 0.02556, + "grad_norm": 0.779791532210884, + "learning_rate": 0.003, + "loss": 4.154, + "step": 2556 + }, + { + "epoch": 0.02557, + "grad_norm": 0.8041071030856266, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 2557 + }, + { + "epoch": 0.02558, + "grad_norm": 0.9046731898466632, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 2558 + }, + { + "epoch": 0.02559, + "grad_norm": 0.9231060963177342, + "learning_rate": 0.003, + "loss": 4.1753, + "step": 2559 + }, + { + "epoch": 0.0256, + "grad_norm": 0.7942918059321874, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 2560 + }, + { + "epoch": 0.02561, + "grad_norm": 0.7182994269745618, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 2561 + }, + { + "epoch": 0.02562, + "grad_norm": 0.7377092426971169, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 2562 + }, + { + "epoch": 0.02563, + "grad_norm": 0.8008807069379792, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 2563 + }, + { + "epoch": 0.02564, + "grad_norm": 0.9654334894945076, + "learning_rate": 0.003, + "loss": 4.1721, + "step": 2564 + }, + { + "epoch": 0.02565, + "grad_norm": 1.1270177489997797, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 2565 + }, + { + "epoch": 0.02566, + "grad_norm": 0.7790702826784099, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 2566 + }, + { + "epoch": 0.02567, + "grad_norm": 0.7253981064969616, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 2567 + }, + { + "epoch": 0.02568, + "grad_norm": 0.7766989633341348, + "learning_rate": 0.003, + "loss": 4.1634, + "step": 2568 + }, + { + "epoch": 0.02569, + "grad_norm": 0.6753943323614341, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 2569 + }, + { + "epoch": 0.0257, + "grad_norm": 0.6053036326649146, + "learning_rate": 0.003, + "loss": 4.142, + "step": 2570 + }, + { + "epoch": 0.02571, + "grad_norm": 0.5412545714840278, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 2571 + }, + { + "epoch": 0.02572, + "grad_norm": 0.618814983550634, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 2572 + }, + { + "epoch": 0.02573, + "grad_norm": 0.6647990779688017, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 2573 + }, + { + "epoch": 0.02574, + "grad_norm": 0.7052848934500006, + "learning_rate": 0.003, + "loss": 4.1737, + "step": 2574 + }, + { + "epoch": 0.02575, + "grad_norm": 0.7610884337795838, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 2575 + }, + { + "epoch": 0.02576, + "grad_norm": 0.6692271813839591, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 2576 + }, + { + "epoch": 0.02577, + "grad_norm": 0.5756800187751938, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 2577 + }, + { + "epoch": 0.02578, + "grad_norm": 0.5445744087322336, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 2578 + }, + { + "epoch": 0.02579, + "grad_norm": 0.5526066768211919, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 2579 + }, + { + "epoch": 0.0258, + "grad_norm": 0.555186498872379, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 2580 + }, + { + "epoch": 0.02581, + "grad_norm": 0.590974511132741, + "learning_rate": 0.003, + "loss": 4.1622, + "step": 2581 + }, + { + "epoch": 0.02582, + "grad_norm": 0.6134714053872987, + "learning_rate": 0.003, + "loss": 4.157, + "step": 2582 + }, + { + "epoch": 0.02583, + "grad_norm": 0.6785456864342533, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 2583 + }, + { + "epoch": 0.02584, + "grad_norm": 0.7692608608699008, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 2584 + }, + { + "epoch": 0.02585, + "grad_norm": 0.8096555829199314, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 2585 + }, + { + "epoch": 0.02586, + "grad_norm": 0.8850658157806096, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 2586 + }, + { + "epoch": 0.02587, + "grad_norm": 1.0102418517134517, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 2587 + }, + { + "epoch": 0.02588, + "grad_norm": 0.9275963803575394, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 2588 + }, + { + "epoch": 0.02589, + "grad_norm": 0.9403619185225363, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 2589 + }, + { + "epoch": 0.0259, + "grad_norm": 1.0078495428012795, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 2590 + }, + { + "epoch": 0.02591, + "grad_norm": 1.2743604633788357, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 2591 + }, + { + "epoch": 0.02592, + "grad_norm": 0.8231355726319857, + "learning_rate": 0.003, + "loss": 4.155, + "step": 2592 + }, + { + "epoch": 0.02593, + "grad_norm": 0.6890041984971179, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 2593 + }, + { + "epoch": 0.02594, + "grad_norm": 0.7454180549344623, + "learning_rate": 0.003, + "loss": 4.1728, + "step": 2594 + }, + { + "epoch": 0.02595, + "grad_norm": 0.6902489836421277, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 2595 + }, + { + "epoch": 0.02596, + "grad_norm": 0.8183952841129141, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 2596 + }, + { + "epoch": 0.02597, + "grad_norm": 0.9211443305771235, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 2597 + }, + { + "epoch": 0.02598, + "grad_norm": 1.0292390798108857, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 2598 + }, + { + "epoch": 0.02599, + "grad_norm": 0.9373288426158787, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 2599 + }, + { + "epoch": 0.026, + "grad_norm": 0.920433494603301, + "learning_rate": 0.003, + "loss": 4.1747, + "step": 2600 + }, + { + "epoch": 0.02601, + "grad_norm": 0.8451864765965912, + "learning_rate": 0.003, + "loss": 4.1776, + "step": 2601 + }, + { + "epoch": 0.02602, + "grad_norm": 0.7170319795923626, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 2602 + }, + { + "epoch": 0.02603, + "grad_norm": 0.6966464803396121, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 2603 + }, + { + "epoch": 0.02604, + "grad_norm": 0.7612566166038212, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 2604 + }, + { + "epoch": 0.02605, + "grad_norm": 0.6464701620756803, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 2605 + }, + { + "epoch": 0.02606, + "grad_norm": 0.5737007841123123, + "learning_rate": 0.003, + "loss": 4.162, + "step": 2606 + }, + { + "epoch": 0.02607, + "grad_norm": 0.6131766565865462, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 2607 + }, + { + "epoch": 0.02608, + "grad_norm": 0.8086035003985383, + "learning_rate": 0.003, + "loss": 4.136, + "step": 2608 + }, + { + "epoch": 0.02609, + "grad_norm": 0.9817909383598199, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 2609 + }, + { + "epoch": 0.0261, + "grad_norm": 1.0720508985841146, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 2610 + }, + { + "epoch": 0.02611, + "grad_norm": 0.8659326470213682, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 2611 + }, + { + "epoch": 0.02612, + "grad_norm": 0.8732525902579722, + "learning_rate": 0.003, + "loss": 4.1648, + "step": 2612 + }, + { + "epoch": 0.02613, + "grad_norm": 0.9874637099516153, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 2613 + }, + { + "epoch": 0.02614, + "grad_norm": 0.906401203843813, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 2614 + }, + { + "epoch": 0.02615, + "grad_norm": 0.6462260762478055, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 2615 + }, + { + "epoch": 0.02616, + "grad_norm": 0.6006231478500065, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 2616 + }, + { + "epoch": 0.02617, + "grad_norm": 0.6102511179138815, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 2617 + }, + { + "epoch": 0.02618, + "grad_norm": 0.6619768894826954, + "learning_rate": 0.003, + "loss": 4.15, + "step": 2618 + }, + { + "epoch": 0.02619, + "grad_norm": 0.7040985692908992, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 2619 + }, + { + "epoch": 0.0262, + "grad_norm": 0.7390370888569476, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 2620 + }, + { + "epoch": 0.02621, + "grad_norm": 0.8595720375733896, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 2621 + }, + { + "epoch": 0.02622, + "grad_norm": 0.8747740097574618, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 2622 + }, + { + "epoch": 0.02623, + "grad_norm": 0.7865985620123697, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 2623 + }, + { + "epoch": 0.02624, + "grad_norm": 0.721745866206175, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 2624 + }, + { + "epoch": 0.02625, + "grad_norm": 0.6369082515266815, + "learning_rate": 0.003, + "loss": 4.1844, + "step": 2625 + }, + { + "epoch": 0.02626, + "grad_norm": 0.6738431853903379, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 2626 + }, + { + "epoch": 0.02627, + "grad_norm": 0.8070717121127621, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 2627 + }, + { + "epoch": 0.02628, + "grad_norm": 0.9052615625467546, + "learning_rate": 0.003, + "loss": 4.1733, + "step": 2628 + }, + { + "epoch": 0.02629, + "grad_norm": 0.9064578842099213, + "learning_rate": 0.003, + "loss": 4.112, + "step": 2629 + }, + { + "epoch": 0.0263, + "grad_norm": 0.9572710825858507, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 2630 + }, + { + "epoch": 0.02631, + "grad_norm": 0.9098415499268271, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2631 + }, + { + "epoch": 0.02632, + "grad_norm": 0.957961730332363, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 2632 + }, + { + "epoch": 0.02633, + "grad_norm": 0.8910675740871853, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 2633 + }, + { + "epoch": 0.02634, + "grad_norm": 0.8894027249202754, + "learning_rate": 0.003, + "loss": 4.1657, + "step": 2634 + }, + { + "epoch": 0.02635, + "grad_norm": 1.0919965764145694, + "learning_rate": 0.003, + "loss": 4.171, + "step": 2635 + }, + { + "epoch": 0.02636, + "grad_norm": 1.101775206670289, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 2636 + }, + { + "epoch": 0.02637, + "grad_norm": 0.9585912224073222, + "learning_rate": 0.003, + "loss": 4.1888, + "step": 2637 + }, + { + "epoch": 0.02638, + "grad_norm": 0.7865200036359159, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 2638 + }, + { + "epoch": 0.02639, + "grad_norm": 0.7526947533075404, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 2639 + }, + { + "epoch": 0.0264, + "grad_norm": 0.8605449721173474, + "learning_rate": 0.003, + "loss": 4.134, + "step": 2640 + }, + { + "epoch": 0.02641, + "grad_norm": 0.9267463548326204, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 2641 + }, + { + "epoch": 0.02642, + "grad_norm": 0.9486962646703484, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 2642 + }, + { + "epoch": 0.02643, + "grad_norm": 0.8844618097245053, + "learning_rate": 0.003, + "loss": 4.15, + "step": 2643 + }, + { + "epoch": 0.02644, + "grad_norm": 0.7967251454981952, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 2644 + }, + { + "epoch": 0.02645, + "grad_norm": 0.8225181053423539, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 2645 + }, + { + "epoch": 0.02646, + "grad_norm": 0.7987479226871795, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 2646 + }, + { + "epoch": 0.02647, + "grad_norm": 0.7821642664442445, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 2647 + }, + { + "epoch": 0.02648, + "grad_norm": 0.9284024850106564, + "learning_rate": 0.003, + "loss": 4.1813, + "step": 2648 + }, + { + "epoch": 0.02649, + "grad_norm": 1.0907008227907322, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 2649 + }, + { + "epoch": 0.0265, + "grad_norm": 1.125793480123405, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 2650 + }, + { + "epoch": 0.02651, + "grad_norm": 0.7578496969583901, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 2651 + }, + { + "epoch": 0.02652, + "grad_norm": 0.7052794942111215, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 2652 + }, + { + "epoch": 0.02653, + "grad_norm": 0.7103196739746442, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 2653 + }, + { + "epoch": 0.02654, + "grad_norm": 0.6925385503676881, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 2654 + }, + { + "epoch": 0.02655, + "grad_norm": 0.781411208540797, + "learning_rate": 0.003, + "loss": 4.147, + "step": 2655 + }, + { + "epoch": 0.02656, + "grad_norm": 0.7579190054446218, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 2656 + }, + { + "epoch": 0.02657, + "grad_norm": 0.7277248228326177, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 2657 + }, + { + "epoch": 0.02658, + "grad_norm": 0.6753248968688912, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 2658 + }, + { + "epoch": 0.02659, + "grad_norm": 0.6796448265031177, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 2659 + }, + { + "epoch": 0.0266, + "grad_norm": 0.6424181529879253, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 2660 + }, + { + "epoch": 0.02661, + "grad_norm": 0.7766396799485783, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 2661 + }, + { + "epoch": 0.02662, + "grad_norm": 0.864583375609625, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 2662 + }, + { + "epoch": 0.02663, + "grad_norm": 0.8284021457779097, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 2663 + }, + { + "epoch": 0.02664, + "grad_norm": 0.7120779790048899, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 2664 + }, + { + "epoch": 0.02665, + "grad_norm": 0.6851733303412342, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2665 + }, + { + "epoch": 0.02666, + "grad_norm": 0.6758809967977054, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 2666 + }, + { + "epoch": 0.02667, + "grad_norm": 0.6361286175404186, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 2667 + }, + { + "epoch": 0.02668, + "grad_norm": 0.5673144703469689, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 2668 + }, + { + "epoch": 0.02669, + "grad_norm": 0.5094068244293505, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 2669 + }, + { + "epoch": 0.0267, + "grad_norm": 0.5324820431990631, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 2670 + }, + { + "epoch": 0.02671, + "grad_norm": 0.6250352979316671, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 2671 + }, + { + "epoch": 0.02672, + "grad_norm": 0.6359363422291939, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 2672 + }, + { + "epoch": 0.02673, + "grad_norm": 0.6102304515358753, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 2673 + }, + { + "epoch": 0.02674, + "grad_norm": 0.7766569602853689, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 2674 + }, + { + "epoch": 0.02675, + "grad_norm": 0.9731715722532488, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 2675 + }, + { + "epoch": 0.02676, + "grad_norm": 1.0370179032200628, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 2676 + }, + { + "epoch": 0.02677, + "grad_norm": 0.7623073096555095, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 2677 + }, + { + "epoch": 0.02678, + "grad_norm": 0.7777692153557114, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 2678 + }, + { + "epoch": 0.02679, + "grad_norm": 0.8733324106472917, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 2679 + }, + { + "epoch": 0.0268, + "grad_norm": 0.8266517145073005, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 2680 + }, + { + "epoch": 0.02681, + "grad_norm": 0.758773812684903, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 2681 + }, + { + "epoch": 0.02682, + "grad_norm": 0.8283252875504241, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 2682 + }, + { + "epoch": 0.02683, + "grad_norm": 0.8404974594062254, + "learning_rate": 0.003, + "loss": 4.1692, + "step": 2683 + }, + { + "epoch": 0.02684, + "grad_norm": 0.818507190423211, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 2684 + }, + { + "epoch": 0.02685, + "grad_norm": 0.7649702763962748, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 2685 + }, + { + "epoch": 0.02686, + "grad_norm": 0.7611427665837982, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 2686 + }, + { + "epoch": 0.02687, + "grad_norm": 0.9628527455690736, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 2687 + }, + { + "epoch": 0.02688, + "grad_norm": 0.8564202799506135, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 2688 + }, + { + "epoch": 0.02689, + "grad_norm": 0.8557838178212854, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 2689 + }, + { + "epoch": 0.0269, + "grad_norm": 0.9118623509339678, + "learning_rate": 0.003, + "loss": 4.164, + "step": 2690 + }, + { + "epoch": 0.02691, + "grad_norm": 0.8301540595445771, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2691 + }, + { + "epoch": 0.02692, + "grad_norm": 0.6228920440524642, + "learning_rate": 0.003, + "loss": 4.1516, + "step": 2692 + }, + { + "epoch": 0.02693, + "grad_norm": 0.6859665731725182, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 2693 + }, + { + "epoch": 0.02694, + "grad_norm": 0.7138959075984226, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 2694 + }, + { + "epoch": 0.02695, + "grad_norm": 0.7527603952329793, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 2695 + }, + { + "epoch": 0.02696, + "grad_norm": 0.7052829196693109, + "learning_rate": 0.003, + "loss": 4.144, + "step": 2696 + }, + { + "epoch": 0.02697, + "grad_norm": 0.6398105779035383, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 2697 + }, + { + "epoch": 0.02698, + "grad_norm": 0.7194666389048138, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 2698 + }, + { + "epoch": 0.02699, + "grad_norm": 0.7132487376498455, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 2699 + }, + { + "epoch": 0.027, + "grad_norm": 0.656063779780492, + "learning_rate": 0.003, + "loss": 4.124, + "step": 2700 + }, + { + "epoch": 0.02701, + "grad_norm": 0.629455348788436, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 2701 + }, + { + "epoch": 0.02702, + "grad_norm": 0.7466120152268881, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 2702 + }, + { + "epoch": 0.02703, + "grad_norm": 0.777360526823959, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 2703 + }, + { + "epoch": 0.02704, + "grad_norm": 0.8782075840804512, + "learning_rate": 0.003, + "loss": 4.11, + "step": 2704 + }, + { + "epoch": 0.02705, + "grad_norm": 0.8925534153148256, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 2705 + }, + { + "epoch": 0.02706, + "grad_norm": 0.9726075369304501, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 2706 + }, + { + "epoch": 0.02707, + "grad_norm": 0.9439770827114989, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 2707 + }, + { + "epoch": 0.02708, + "grad_norm": 0.8038542754161909, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 2708 + }, + { + "epoch": 0.02709, + "grad_norm": 0.8203620131302694, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 2709 + }, + { + "epoch": 0.0271, + "grad_norm": 0.8724888263473524, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 2710 + }, + { + "epoch": 0.02711, + "grad_norm": 0.8653954860429884, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 2711 + }, + { + "epoch": 0.02712, + "grad_norm": 0.8020656899710369, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 2712 + }, + { + "epoch": 0.02713, + "grad_norm": 0.8083253547021919, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 2713 + }, + { + "epoch": 0.02714, + "grad_norm": 0.8690387355838295, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 2714 + }, + { + "epoch": 0.02715, + "grad_norm": 1.0593958700668973, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 2715 + }, + { + "epoch": 0.02716, + "grad_norm": 1.0574588180443436, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 2716 + }, + { + "epoch": 0.02717, + "grad_norm": 0.9164750014025292, + "learning_rate": 0.003, + "loss": 4.147, + "step": 2717 + }, + { + "epoch": 0.02718, + "grad_norm": 0.9137711318115498, + "learning_rate": 0.003, + "loss": 4.1696, + "step": 2718 + }, + { + "epoch": 0.02719, + "grad_norm": 0.7747921431969758, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 2719 + }, + { + "epoch": 0.0272, + "grad_norm": 0.8252071711518221, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 2720 + }, + { + "epoch": 0.02721, + "grad_norm": 0.9010417786683242, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 2721 + }, + { + "epoch": 0.02722, + "grad_norm": 1.0430685055558186, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 2722 + }, + { + "epoch": 0.02723, + "grad_norm": 0.8338446086253062, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 2723 + }, + { + "epoch": 0.02724, + "grad_norm": 0.7926892406317515, + "learning_rate": 0.003, + "loss": 4.1604, + "step": 2724 + }, + { + "epoch": 0.02725, + "grad_norm": 0.9476051066490954, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 2725 + }, + { + "epoch": 0.02726, + "grad_norm": 0.9718868135312356, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 2726 + }, + { + "epoch": 0.02727, + "grad_norm": 0.8794969874731017, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 2727 + }, + { + "epoch": 0.02728, + "grad_norm": 0.7500620026401651, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 2728 + }, + { + "epoch": 0.02729, + "grad_norm": 0.682111991166883, + "learning_rate": 0.003, + "loss": 4.172, + "step": 2729 + }, + { + "epoch": 0.0273, + "grad_norm": 0.5938416056877477, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 2730 + }, + { + "epoch": 0.02731, + "grad_norm": 0.5014324118447245, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 2731 + }, + { + "epoch": 0.02732, + "grad_norm": 0.5566218730286413, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 2732 + }, + { + "epoch": 0.02733, + "grad_norm": 0.656514130835437, + "learning_rate": 0.003, + "loss": 4.16, + "step": 2733 + }, + { + "epoch": 0.02734, + "grad_norm": 0.7092083749191737, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 2734 + }, + { + "epoch": 0.02735, + "grad_norm": 0.7490351435778966, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 2735 + }, + { + "epoch": 0.02736, + "grad_norm": 0.8057410040143573, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 2736 + }, + { + "epoch": 0.02737, + "grad_norm": 0.7272389058341057, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 2737 + }, + { + "epoch": 0.02738, + "grad_norm": 0.7181903456270696, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 2738 + }, + { + "epoch": 0.02739, + "grad_norm": 0.8790015592136898, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 2739 + }, + { + "epoch": 0.0274, + "grad_norm": 1.0911668104806873, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 2740 + }, + { + "epoch": 0.02741, + "grad_norm": 1.0149366352031834, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2741 + }, + { + "epoch": 0.02742, + "grad_norm": 0.8648703789133807, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 2742 + }, + { + "epoch": 0.02743, + "grad_norm": 0.8509616129669461, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 2743 + }, + { + "epoch": 0.02744, + "grad_norm": 0.7514267415016833, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 2744 + }, + { + "epoch": 0.02745, + "grad_norm": 0.666188856053761, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 2745 + }, + { + "epoch": 0.02746, + "grad_norm": 0.6204707613465867, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 2746 + }, + { + "epoch": 0.02747, + "grad_norm": 0.5745978206046833, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 2747 + }, + { + "epoch": 0.02748, + "grad_norm": 0.5095037008788326, + "learning_rate": 0.003, + "loss": 4.106, + "step": 2748 + }, + { + "epoch": 0.02749, + "grad_norm": 0.4918061860336399, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 2749 + }, + { + "epoch": 0.0275, + "grad_norm": 0.4689633122885494, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 2750 + }, + { + "epoch": 0.02751, + "grad_norm": 0.5240834818389507, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2751 + }, + { + "epoch": 0.02752, + "grad_norm": 0.45737584283965704, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 2752 + }, + { + "epoch": 0.02753, + "grad_norm": 0.4854554116683414, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 2753 + }, + { + "epoch": 0.02754, + "grad_norm": 0.5451937888698382, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 2754 + }, + { + "epoch": 0.02755, + "grad_norm": 0.6902713768999703, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 2755 + }, + { + "epoch": 0.02756, + "grad_norm": 0.9172618359251131, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 2756 + }, + { + "epoch": 0.02757, + "grad_norm": 1.264093714132492, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 2757 + }, + { + "epoch": 0.02758, + "grad_norm": 0.6764873385501415, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 2758 + }, + { + "epoch": 0.02759, + "grad_norm": 0.6735199753323651, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 2759 + }, + { + "epoch": 0.0276, + "grad_norm": 0.7869359581922244, + "learning_rate": 0.003, + "loss": 4.143, + "step": 2760 + }, + { + "epoch": 0.02761, + "grad_norm": 0.8232479826802365, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 2761 + }, + { + "epoch": 0.02762, + "grad_norm": 0.8094991367198255, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 2762 + }, + { + "epoch": 0.02763, + "grad_norm": 0.808838083900847, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 2763 + }, + { + "epoch": 0.02764, + "grad_norm": 0.9499840304387701, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 2764 + }, + { + "epoch": 0.02765, + "grad_norm": 1.1661515206508581, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 2765 + }, + { + "epoch": 0.02766, + "grad_norm": 0.8834816047449258, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 2766 + }, + { + "epoch": 0.02767, + "grad_norm": 0.9679955840143722, + "learning_rate": 0.003, + "loss": 4.1737, + "step": 2767 + }, + { + "epoch": 0.02768, + "grad_norm": 1.0122550261242687, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 2768 + }, + { + "epoch": 0.02769, + "grad_norm": 0.936589272381051, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 2769 + }, + { + "epoch": 0.0277, + "grad_norm": 0.8812342595154814, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 2770 + }, + { + "epoch": 0.02771, + "grad_norm": 0.8385099378903244, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 2771 + }, + { + "epoch": 0.02772, + "grad_norm": 0.8615548972194946, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 2772 + }, + { + "epoch": 0.02773, + "grad_norm": 0.7470925818947043, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 2773 + }, + { + "epoch": 0.02774, + "grad_norm": 0.7432890220276933, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 2774 + }, + { + "epoch": 0.02775, + "grad_norm": 0.714251751808306, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 2775 + }, + { + "epoch": 0.02776, + "grad_norm": 0.6696685091947796, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 2776 + }, + { + "epoch": 0.02777, + "grad_norm": 0.6601977907009066, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 2777 + }, + { + "epoch": 0.02778, + "grad_norm": 0.6363902696248916, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 2778 + }, + { + "epoch": 0.02779, + "grad_norm": 0.7347310050156719, + "learning_rate": 0.003, + "loss": 4.126, + "step": 2779 + }, + { + "epoch": 0.0278, + "grad_norm": 0.8193092944005641, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 2780 + }, + { + "epoch": 0.02781, + "grad_norm": 0.9505772704924963, + "learning_rate": 0.003, + "loss": 4.113, + "step": 2781 + }, + { + "epoch": 0.02782, + "grad_norm": 1.09253758346844, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 2782 + }, + { + "epoch": 0.02783, + "grad_norm": 0.7922623009085619, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 2783 + }, + { + "epoch": 0.02784, + "grad_norm": 0.7730396671741616, + "learning_rate": 0.003, + "loss": 4.136, + "step": 2784 + }, + { + "epoch": 0.02785, + "grad_norm": 0.8111585358679776, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 2785 + }, + { + "epoch": 0.02786, + "grad_norm": 0.7983751779273013, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 2786 + }, + { + "epoch": 0.02787, + "grad_norm": 0.8522172170925804, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 2787 + }, + { + "epoch": 0.02788, + "grad_norm": 0.8952452301797756, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 2788 + }, + { + "epoch": 0.02789, + "grad_norm": 0.9315897488998124, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 2789 + }, + { + "epoch": 0.0279, + "grad_norm": 0.7715461952206646, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 2790 + }, + { + "epoch": 0.02791, + "grad_norm": 0.7419347495262759, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 2791 + }, + { + "epoch": 0.02792, + "grad_norm": 0.804466637477434, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 2792 + }, + { + "epoch": 0.02793, + "grad_norm": 0.9499312139027115, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 2793 + }, + { + "epoch": 0.02794, + "grad_norm": 1.079651373139261, + "learning_rate": 0.003, + "loss": 4.1503, + "step": 2794 + }, + { + "epoch": 0.02795, + "grad_norm": 0.7784790735501392, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 2795 + }, + { + "epoch": 0.02796, + "grad_norm": 0.7016883329409419, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 2796 + }, + { + "epoch": 0.02797, + "grad_norm": 0.8527425055953903, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 2797 + }, + { + "epoch": 0.02798, + "grad_norm": 0.8344419810312547, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 2798 + }, + { + "epoch": 0.02799, + "grad_norm": 0.8118333253983723, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 2799 + }, + { + "epoch": 0.028, + "grad_norm": 0.8276964023163034, + "learning_rate": 0.003, + "loss": 4.154, + "step": 2800 + }, + { + "epoch": 0.02801, + "grad_norm": 0.8829595698308325, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 2801 + }, + { + "epoch": 0.02802, + "grad_norm": 0.9178695537376129, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 2802 + }, + { + "epoch": 0.02803, + "grad_norm": 0.773306657979122, + "learning_rate": 0.003, + "loss": 4.1566, + "step": 2803 + }, + { + "epoch": 0.02804, + "grad_norm": 0.6894758923555496, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 2804 + }, + { + "epoch": 0.02805, + "grad_norm": 0.6917993567356406, + "learning_rate": 0.003, + "loss": 4.1566, + "step": 2805 + }, + { + "epoch": 0.02806, + "grad_norm": 0.6387085680780493, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 2806 + }, + { + "epoch": 0.02807, + "grad_norm": 0.580517980379294, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 2807 + }, + { + "epoch": 0.02808, + "grad_norm": 0.4980072648857084, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 2808 + }, + { + "epoch": 0.02809, + "grad_norm": 0.5292621050422022, + "learning_rate": 0.003, + "loss": 4.1503, + "step": 2809 + }, + { + "epoch": 0.0281, + "grad_norm": 0.5662997833368786, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 2810 + }, + { + "epoch": 0.02811, + "grad_norm": 0.6056304842841205, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 2811 + }, + { + "epoch": 0.02812, + "grad_norm": 0.6510291438212791, + "learning_rate": 0.003, + "loss": 4.131, + "step": 2812 + }, + { + "epoch": 0.02813, + "grad_norm": 0.6856589618404986, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 2813 + }, + { + "epoch": 0.02814, + "grad_norm": 0.7345790506941087, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 2814 + }, + { + "epoch": 0.02815, + "grad_norm": 0.7816359381581716, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 2815 + }, + { + "epoch": 0.02816, + "grad_norm": 0.8925647710841813, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 2816 + }, + { + "epoch": 0.02817, + "grad_norm": 1.169228788902705, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 2817 + }, + { + "epoch": 0.02818, + "grad_norm": 0.9618528564502903, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 2818 + }, + { + "epoch": 0.02819, + "grad_norm": 1.0178604687945154, + "learning_rate": 0.003, + "loss": 4.15, + "step": 2819 + }, + { + "epoch": 0.0282, + "grad_norm": 0.8729007051855019, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 2820 + }, + { + "epoch": 0.02821, + "grad_norm": 0.7594720998624757, + "learning_rate": 0.003, + "loss": 4.1557, + "step": 2821 + }, + { + "epoch": 0.02822, + "grad_norm": 0.7764641169305468, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 2822 + }, + { + "epoch": 0.02823, + "grad_norm": 0.79863443820284, + "learning_rate": 0.003, + "loss": 4.154, + "step": 2823 + }, + { + "epoch": 0.02824, + "grad_norm": 0.7182103089348308, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 2824 + }, + { + "epoch": 0.02825, + "grad_norm": 0.7529407236815215, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 2825 + }, + { + "epoch": 0.02826, + "grad_norm": 0.8056220195019022, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 2826 + }, + { + "epoch": 0.02827, + "grad_norm": 0.8893025905667793, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 2827 + }, + { + "epoch": 0.02828, + "grad_norm": 0.9458155387571202, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 2828 + }, + { + "epoch": 0.02829, + "grad_norm": 1.0816962232960836, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 2829 + }, + { + "epoch": 0.0283, + "grad_norm": 0.85593992308399, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 2830 + }, + { + "epoch": 0.02831, + "grad_norm": 0.8103547032302357, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 2831 + }, + { + "epoch": 0.02832, + "grad_norm": 1.0004568820848851, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 2832 + }, + { + "epoch": 0.02833, + "grad_norm": 1.2257623869393093, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 2833 + }, + { + "epoch": 0.02834, + "grad_norm": 0.8959695287844174, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 2834 + }, + { + "epoch": 0.02835, + "grad_norm": 0.9046726088007764, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 2835 + }, + { + "epoch": 0.02836, + "grad_norm": 0.9388179555413667, + "learning_rate": 0.003, + "loss": 4.1634, + "step": 2836 + }, + { + "epoch": 0.02837, + "grad_norm": 0.8876864866399585, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 2837 + }, + { + "epoch": 0.02838, + "grad_norm": 0.93231141838983, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 2838 + }, + { + "epoch": 0.02839, + "grad_norm": 0.8358991135319658, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 2839 + }, + { + "epoch": 0.0284, + "grad_norm": 0.8834561842613636, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 2840 + }, + { + "epoch": 0.02841, + "grad_norm": 0.9006412651182929, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 2841 + }, + { + "epoch": 0.02842, + "grad_norm": 0.9446862919101724, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 2842 + }, + { + "epoch": 0.02843, + "grad_norm": 0.8098223520337203, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 2843 + }, + { + "epoch": 0.02844, + "grad_norm": 0.7322010149090498, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 2844 + }, + { + "epoch": 0.02845, + "grad_norm": 0.6471809615510085, + "learning_rate": 0.003, + "loss": 4.126, + "step": 2845 + }, + { + "epoch": 0.02846, + "grad_norm": 0.6878791837952257, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 2846 + }, + { + "epoch": 0.02847, + "grad_norm": 0.7696793006326048, + "learning_rate": 0.003, + "loss": 4.1788, + "step": 2847 + }, + { + "epoch": 0.02848, + "grad_norm": 0.8913190904442756, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 2848 + }, + { + "epoch": 0.02849, + "grad_norm": 1.018118471142102, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 2849 + }, + { + "epoch": 0.0285, + "grad_norm": 0.9082110603942541, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 2850 + }, + { + "epoch": 0.02851, + "grad_norm": 0.822250823175019, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 2851 + }, + { + "epoch": 0.02852, + "grad_norm": 0.6052901464954267, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 2852 + }, + { + "epoch": 0.02853, + "grad_norm": 0.5735422812437921, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 2853 + }, + { + "epoch": 0.02854, + "grad_norm": 0.5429930679913405, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 2854 + }, + { + "epoch": 0.02855, + "grad_norm": 0.5339097730360199, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 2855 + }, + { + "epoch": 0.02856, + "grad_norm": 0.5860774205400787, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 2856 + }, + { + "epoch": 0.02857, + "grad_norm": 0.6433890086945638, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 2857 + }, + { + "epoch": 0.02858, + "grad_norm": 0.7879326106492056, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 2858 + }, + { + "epoch": 0.02859, + "grad_norm": 0.8693860304588211, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 2859 + }, + { + "epoch": 0.0286, + "grad_norm": 0.8079095294087945, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 2860 + }, + { + "epoch": 0.02861, + "grad_norm": 0.6656764279644372, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 2861 + }, + { + "epoch": 0.02862, + "grad_norm": 0.6696791424270172, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 2862 + }, + { + "epoch": 0.02863, + "grad_norm": 0.7769070451321298, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 2863 + }, + { + "epoch": 0.02864, + "grad_norm": 0.8224098496945286, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 2864 + }, + { + "epoch": 0.02865, + "grad_norm": 0.8864904617013009, + "learning_rate": 0.003, + "loss": 4.1583, + "step": 2865 + }, + { + "epoch": 0.02866, + "grad_norm": 0.9583148110901643, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 2866 + }, + { + "epoch": 0.02867, + "grad_norm": 0.8904807738138028, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 2867 + }, + { + "epoch": 0.02868, + "grad_norm": 0.9293551806467387, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 2868 + }, + { + "epoch": 0.02869, + "grad_norm": 1.0457657175233683, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 2869 + }, + { + "epoch": 0.0287, + "grad_norm": 1.0953444171616145, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 2870 + }, + { + "epoch": 0.02871, + "grad_norm": 0.7587885970460945, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 2871 + }, + { + "epoch": 0.02872, + "grad_norm": 0.6777165872502906, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 2872 + }, + { + "epoch": 0.02873, + "grad_norm": 0.8924775793037929, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 2873 + }, + { + "epoch": 0.02874, + "grad_norm": 0.9191095226023763, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 2874 + }, + { + "epoch": 0.02875, + "grad_norm": 1.0057201785085328, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2875 + }, + { + "epoch": 0.02876, + "grad_norm": 0.9084894421524936, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 2876 + }, + { + "epoch": 0.02877, + "grad_norm": 0.9095891721863432, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 2877 + }, + { + "epoch": 0.02878, + "grad_norm": 0.9172527890033496, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 2878 + }, + { + "epoch": 0.02879, + "grad_norm": 0.8805567763002417, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 2879 + }, + { + "epoch": 0.0288, + "grad_norm": 0.9996123727255034, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 2880 + }, + { + "epoch": 0.02881, + "grad_norm": 1.046330847202451, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 2881 + }, + { + "epoch": 0.02882, + "grad_norm": 0.772117255675699, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 2882 + }, + { + "epoch": 0.02883, + "grad_norm": 0.7008407026763371, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 2883 + }, + { + "epoch": 0.02884, + "grad_norm": 0.778782483892839, + "learning_rate": 0.003, + "loss": 4.177, + "step": 2884 + }, + { + "epoch": 0.02885, + "grad_norm": 0.7648524574296616, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 2885 + }, + { + "epoch": 0.02886, + "grad_norm": 0.6944709668733141, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 2886 + }, + { + "epoch": 0.02887, + "grad_norm": 0.6701833801380009, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 2887 + }, + { + "epoch": 0.02888, + "grad_norm": 0.6522446873326756, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 2888 + }, + { + "epoch": 0.02889, + "grad_norm": 0.5695684026681817, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 2889 + }, + { + "epoch": 0.0289, + "grad_norm": 0.5387645552557849, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 2890 + }, + { + "epoch": 0.02891, + "grad_norm": 0.6068073035716243, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 2891 + }, + { + "epoch": 0.02892, + "grad_norm": 0.5642060774681064, + "learning_rate": 0.003, + "loss": 4.109, + "step": 2892 + }, + { + "epoch": 0.02893, + "grad_norm": 0.64295806093738, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 2893 + }, + { + "epoch": 0.02894, + "grad_norm": 0.7755044700890192, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 2894 + }, + { + "epoch": 0.02895, + "grad_norm": 0.7810569175434396, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2895 + }, + { + "epoch": 0.02896, + "grad_norm": 0.5425405989527838, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 2896 + }, + { + "epoch": 0.02897, + "grad_norm": 0.469070597819537, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 2897 + }, + { + "epoch": 0.02898, + "grad_norm": 0.5139238004799116, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 2898 + }, + { + "epoch": 0.02899, + "grad_norm": 0.6544046694256337, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 2899 + }, + { + "epoch": 0.029, + "grad_norm": 0.7180257646745092, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 2900 + }, + { + "epoch": 0.02901, + "grad_norm": 0.6895214504299034, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 2901 + }, + { + "epoch": 0.02902, + "grad_norm": 0.6936366493807695, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 2902 + }, + { + "epoch": 0.02903, + "grad_norm": 0.7680000086672168, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 2903 + }, + { + "epoch": 0.02904, + "grad_norm": 1.0523944817643214, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 2904 + }, + { + "epoch": 0.02905, + "grad_norm": 1.051471635673892, + "learning_rate": 0.003, + "loss": 4.1582, + "step": 2905 + }, + { + "epoch": 0.02906, + "grad_norm": 0.8511171595587916, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 2906 + }, + { + "epoch": 0.02907, + "grad_norm": 0.8224344496311664, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 2907 + }, + { + "epoch": 0.02908, + "grad_norm": 0.6999459176877906, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 2908 + }, + { + "epoch": 0.02909, + "grad_norm": 0.819486070194941, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 2909 + }, + { + "epoch": 0.0291, + "grad_norm": 1.066905831023775, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 2910 + }, + { + "epoch": 0.02911, + "grad_norm": 0.993942719913462, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 2911 + }, + { + "epoch": 0.02912, + "grad_norm": 0.9395324885369105, + "learning_rate": 0.003, + "loss": 4.114, + "step": 2912 + }, + { + "epoch": 0.02913, + "grad_norm": 0.9563813373046008, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 2913 + }, + { + "epoch": 0.02914, + "grad_norm": 1.204534731933291, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 2914 + }, + { + "epoch": 0.02915, + "grad_norm": 0.9555386893347948, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 2915 + }, + { + "epoch": 0.02916, + "grad_norm": 1.0824011254780639, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 2916 + }, + { + "epoch": 0.02917, + "grad_norm": 1.0942512414631864, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 2917 + }, + { + "epoch": 0.02918, + "grad_norm": 0.9071341774299415, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 2918 + }, + { + "epoch": 0.02919, + "grad_norm": 0.9246562992704038, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 2919 + }, + { + "epoch": 0.0292, + "grad_norm": 0.8971677038672106, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 2920 + }, + { + "epoch": 0.02921, + "grad_norm": 0.926961421602634, + "learning_rate": 0.003, + "loss": 4.1553, + "step": 2921 + }, + { + "epoch": 0.02922, + "grad_norm": 1.0848027072453201, + "learning_rate": 0.003, + "loss": 4.161, + "step": 2922 + }, + { + "epoch": 0.02923, + "grad_norm": 0.912142779077174, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 2923 + }, + { + "epoch": 0.02924, + "grad_norm": 0.9865161740552381, + "learning_rate": 0.003, + "loss": 4.1648, + "step": 2924 + }, + { + "epoch": 0.02925, + "grad_norm": 1.0463105360491507, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 2925 + }, + { + "epoch": 0.02926, + "grad_norm": 0.9582122548230526, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 2926 + }, + { + "epoch": 0.02927, + "grad_norm": 0.8403311811917105, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 2927 + }, + { + "epoch": 0.02928, + "grad_norm": 0.7854536393801775, + "learning_rate": 0.003, + "loss": 4.1695, + "step": 2928 + }, + { + "epoch": 0.02929, + "grad_norm": 0.7778022366151639, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 2929 + }, + { + "epoch": 0.0293, + "grad_norm": 0.7271690214257167, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 2930 + }, + { + "epoch": 0.02931, + "grad_norm": 0.782379273286375, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 2931 + }, + { + "epoch": 0.02932, + "grad_norm": 0.7974568484352181, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 2932 + }, + { + "epoch": 0.02933, + "grad_norm": 0.7828005432560315, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 2933 + }, + { + "epoch": 0.02934, + "grad_norm": 0.6614858211514127, + "learning_rate": 0.003, + "loss": 4.14, + "step": 2934 + }, + { + "epoch": 0.02935, + "grad_norm": 0.5873856004708764, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 2935 + }, + { + "epoch": 0.02936, + "grad_norm": 0.578823462032298, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 2936 + }, + { + "epoch": 0.02937, + "grad_norm": 0.6078930790219389, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 2937 + }, + { + "epoch": 0.02938, + "grad_norm": 0.7146509387847667, + "learning_rate": 0.003, + "loss": 4.139, + "step": 2938 + }, + { + "epoch": 0.02939, + "grad_norm": 0.8250738871743942, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 2939 + }, + { + "epoch": 0.0294, + "grad_norm": 0.8696648935923786, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 2940 + }, + { + "epoch": 0.02941, + "grad_norm": 0.9090371360813282, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 2941 + }, + { + "epoch": 0.02942, + "grad_norm": 0.8939350957264195, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 2942 + }, + { + "epoch": 0.02943, + "grad_norm": 0.7023666348770743, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 2943 + }, + { + "epoch": 0.02944, + "grad_norm": 0.7067049448574164, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 2944 + }, + { + "epoch": 0.02945, + "grad_norm": 0.6812738232543556, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 2945 + }, + { + "epoch": 0.02946, + "grad_norm": 0.6818360632430761, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 2946 + }, + { + "epoch": 0.02947, + "grad_norm": 0.6754875767553201, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 2947 + }, + { + "epoch": 0.02948, + "grad_norm": 0.6113651211997226, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 2948 + }, + { + "epoch": 0.02949, + "grad_norm": 0.5812056588270224, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 2949 + }, + { + "epoch": 0.0295, + "grad_norm": 0.6282215017847426, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 2950 + }, + { + "epoch": 0.02951, + "grad_norm": 0.7817687295389203, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 2951 + }, + { + "epoch": 0.02952, + "grad_norm": 1.0587479164041036, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 2952 + }, + { + "epoch": 0.02953, + "grad_norm": 0.9887356887399129, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 2953 + }, + { + "epoch": 0.02954, + "grad_norm": 0.7805100959975674, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 2954 + }, + { + "epoch": 0.02955, + "grad_norm": 0.6651525603429118, + "learning_rate": 0.003, + "loss": 4.138, + "step": 2955 + }, + { + "epoch": 0.02956, + "grad_norm": 0.7699019886627679, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 2956 + }, + { + "epoch": 0.02957, + "grad_norm": 0.816443587627682, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 2957 + }, + { + "epoch": 0.02958, + "grad_norm": 0.7961877462112285, + "learning_rate": 0.003, + "loss": 4.1553, + "step": 2958 + }, + { + "epoch": 0.02959, + "grad_norm": 0.7550243768723249, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 2959 + }, + { + "epoch": 0.0296, + "grad_norm": 0.7723125453491877, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 2960 + }, + { + "epoch": 0.02961, + "grad_norm": 0.8583607117371619, + "learning_rate": 0.003, + "loss": 4.127, + "step": 2961 + }, + { + "epoch": 0.02962, + "grad_norm": 0.8691287903706768, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 2962 + }, + { + "epoch": 0.02963, + "grad_norm": 0.8786112930172387, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 2963 + }, + { + "epoch": 0.02964, + "grad_norm": 0.9167891542829973, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 2964 + }, + { + "epoch": 0.02965, + "grad_norm": 0.9419888897278245, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 2965 + }, + { + "epoch": 0.02966, + "grad_norm": 1.1891655776183911, + "learning_rate": 0.003, + "loss": 4.1783, + "step": 2966 + }, + { + "epoch": 0.02967, + "grad_norm": 1.0455487688038185, + "learning_rate": 0.003, + "loss": 4.1592, + "step": 2967 + }, + { + "epoch": 0.02968, + "grad_norm": 1.0164630073257412, + "learning_rate": 0.003, + "loss": 4.155, + "step": 2968 + }, + { + "epoch": 0.02969, + "grad_norm": 1.0185954831664046, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 2969 + }, + { + "epoch": 0.0297, + "grad_norm": 1.043969964180144, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 2970 + }, + { + "epoch": 0.02971, + "grad_norm": 0.8397020648022002, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 2971 + }, + { + "epoch": 0.02972, + "grad_norm": 0.6978014599627598, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 2972 + }, + { + "epoch": 0.02973, + "grad_norm": 0.7307935837042818, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 2973 + }, + { + "epoch": 0.02974, + "grad_norm": 0.8618322772785041, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 2974 + }, + { + "epoch": 0.02975, + "grad_norm": 0.9050029913314735, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 2975 + }, + { + "epoch": 0.02976, + "grad_norm": 0.8860984382067472, + "learning_rate": 0.003, + "loss": 4.143, + "step": 2976 + }, + { + "epoch": 0.02977, + "grad_norm": 0.6971184630795387, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 2977 + }, + { + "epoch": 0.02978, + "grad_norm": 0.7418889082376986, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 2978 + }, + { + "epoch": 0.02979, + "grad_norm": 0.7888236204090991, + "learning_rate": 0.003, + "loss": 4.133, + "step": 2979 + }, + { + "epoch": 0.0298, + "grad_norm": 0.7867568676249742, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 2980 + }, + { + "epoch": 0.02981, + "grad_norm": 0.7844697697074997, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 2981 + }, + { + "epoch": 0.02982, + "grad_norm": 0.6873427232122642, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 2982 + }, + { + "epoch": 0.02983, + "grad_norm": 0.7086227659108, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 2983 + }, + { + "epoch": 0.02984, + "grad_norm": 0.7859850298348623, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 2984 + }, + { + "epoch": 0.02985, + "grad_norm": 0.7729797989684547, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 2985 + }, + { + "epoch": 0.02986, + "grad_norm": 0.6844032687054661, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 2986 + }, + { + "epoch": 0.02987, + "grad_norm": 0.5762042827564682, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 2987 + }, + { + "epoch": 0.02988, + "grad_norm": 0.5620617182133862, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 2988 + }, + { + "epoch": 0.02989, + "grad_norm": 0.5894029162100797, + "learning_rate": 0.003, + "loss": 4.132, + "step": 2989 + }, + { + "epoch": 0.0299, + "grad_norm": 0.5837394519047691, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 2990 + }, + { + "epoch": 0.02991, + "grad_norm": 0.5606386580984015, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 2991 + }, + { + "epoch": 0.02992, + "grad_norm": 0.6147897819301783, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 2992 + }, + { + "epoch": 0.02993, + "grad_norm": 0.6621349604754609, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 2993 + }, + { + "epoch": 0.02994, + "grad_norm": 0.8082428932736804, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 2994 + }, + { + "epoch": 0.02995, + "grad_norm": 0.9615508716688156, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 2995 + }, + { + "epoch": 0.02996, + "grad_norm": 0.9867828831321791, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 2996 + }, + { + "epoch": 0.02997, + "grad_norm": 0.6984846126670015, + "learning_rate": 0.003, + "loss": 4.112, + "step": 2997 + }, + { + "epoch": 0.02998, + "grad_norm": 0.7004936715134856, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 2998 + }, + { + "epoch": 0.02999, + "grad_norm": 0.8087922821030329, + "learning_rate": 0.003, + "loss": 4.106, + "step": 2999 + }, + { + "epoch": 0.03, + "grad_norm": 0.8187211297372818, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 3000 + }, + { + "epoch": 0.03001, + "grad_norm": 0.7682209061752201, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 3001 + }, + { + "epoch": 0.03002, + "grad_norm": 0.7895156191994268, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 3002 + }, + { + "epoch": 0.03003, + "grad_norm": 0.9473169165490495, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 3003 + }, + { + "epoch": 0.03004, + "grad_norm": 0.9772459835659143, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 3004 + }, + { + "epoch": 0.03005, + "grad_norm": 0.9271870841526649, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 3005 + }, + { + "epoch": 0.03006, + "grad_norm": 0.81719364295006, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3006 + }, + { + "epoch": 0.03007, + "grad_norm": 0.8963225985425357, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 3007 + }, + { + "epoch": 0.03008, + "grad_norm": 0.7818424581508167, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 3008 + }, + { + "epoch": 0.03009, + "grad_norm": 0.8175800868725163, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 3009 + }, + { + "epoch": 0.0301, + "grad_norm": 0.8915634122617913, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 3010 + }, + { + "epoch": 0.03011, + "grad_norm": 0.9916010748888695, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 3011 + }, + { + "epoch": 0.03012, + "grad_norm": 0.9009840506519443, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3012 + }, + { + "epoch": 0.03013, + "grad_norm": 0.8990400357883044, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 3013 + }, + { + "epoch": 0.03014, + "grad_norm": 0.8701803132962008, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 3014 + }, + { + "epoch": 0.03015, + "grad_norm": 0.9059467708321113, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 3015 + }, + { + "epoch": 0.03016, + "grad_norm": 0.8139710481414894, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 3016 + }, + { + "epoch": 0.03017, + "grad_norm": 0.9003169401719128, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 3017 + }, + { + "epoch": 0.03018, + "grad_norm": 0.9148332662761703, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 3018 + }, + { + "epoch": 0.03019, + "grad_norm": 0.8097272342713991, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 3019 + }, + { + "epoch": 0.0302, + "grad_norm": 0.8482835840133595, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 3020 + }, + { + "epoch": 0.03021, + "grad_norm": 0.7875221358593745, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 3021 + }, + { + "epoch": 0.03022, + "grad_norm": 0.792143621479002, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 3022 + }, + { + "epoch": 0.03023, + "grad_norm": 0.8897124061645492, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 3023 + }, + { + "epoch": 0.03024, + "grad_norm": 1.016846130536088, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 3024 + }, + { + "epoch": 0.03025, + "grad_norm": 0.9813514493806819, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 3025 + }, + { + "epoch": 0.03026, + "grad_norm": 0.9856440178637326, + "learning_rate": 0.003, + "loss": 4.119, + "step": 3026 + }, + { + "epoch": 0.03027, + "grad_norm": 1.0170462610037305, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 3027 + }, + { + "epoch": 0.03028, + "grad_norm": 0.9963929882829512, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 3028 + }, + { + "epoch": 0.03029, + "grad_norm": 1.1230470374601593, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 3029 + }, + { + "epoch": 0.0303, + "grad_norm": 0.9952469868078255, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 3030 + }, + { + "epoch": 0.03031, + "grad_norm": 0.7974194139158707, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 3031 + }, + { + "epoch": 0.03032, + "grad_norm": 0.813594268367442, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 3032 + }, + { + "epoch": 0.03033, + "grad_norm": 0.8000080591024091, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 3033 + }, + { + "epoch": 0.03034, + "grad_norm": 0.8702456358651927, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 3034 + }, + { + "epoch": 0.03035, + "grad_norm": 0.972910117224573, + "learning_rate": 0.003, + "loss": 4.1701, + "step": 3035 + }, + { + "epoch": 0.03036, + "grad_norm": 0.946589194340261, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 3036 + }, + { + "epoch": 0.03037, + "grad_norm": 1.0013129497511943, + "learning_rate": 0.003, + "loss": 4.13, + "step": 3037 + }, + { + "epoch": 0.03038, + "grad_norm": 1.0504414618218458, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 3038 + }, + { + "epoch": 0.03039, + "grad_norm": 0.9552159328342077, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 3039 + }, + { + "epoch": 0.0304, + "grad_norm": 0.8125546290590606, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 3040 + }, + { + "epoch": 0.03041, + "grad_norm": 0.7829978292483236, + "learning_rate": 0.003, + "loss": 4.142, + "step": 3041 + }, + { + "epoch": 0.03042, + "grad_norm": 0.7611930258666472, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 3042 + }, + { + "epoch": 0.03043, + "grad_norm": 0.7149717720566953, + "learning_rate": 0.003, + "loss": 4.118, + "step": 3043 + }, + { + "epoch": 0.03044, + "grad_norm": 0.7624230406515938, + "learning_rate": 0.003, + "loss": 4.12, + "step": 3044 + }, + { + "epoch": 0.03045, + "grad_norm": 0.8408403467918055, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3045 + }, + { + "epoch": 0.03046, + "grad_norm": 0.7851334779365589, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 3046 + }, + { + "epoch": 0.03047, + "grad_norm": 0.7122809857507961, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 3047 + }, + { + "epoch": 0.03048, + "grad_norm": 0.7964799855759337, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 3048 + }, + { + "epoch": 0.03049, + "grad_norm": 0.887239863638981, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 3049 + }, + { + "epoch": 0.0305, + "grad_norm": 0.9999215882595172, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 3050 + }, + { + "epoch": 0.03051, + "grad_norm": 1.0055098406963119, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 3051 + }, + { + "epoch": 0.03052, + "grad_norm": 0.9173091708203194, + "learning_rate": 0.003, + "loss": 4.146, + "step": 3052 + }, + { + "epoch": 0.03053, + "grad_norm": 0.8629407073466706, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 3053 + }, + { + "epoch": 0.03054, + "grad_norm": 0.7744384974895352, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 3054 + }, + { + "epoch": 0.03055, + "grad_norm": 0.6670797287560947, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 3055 + }, + { + "epoch": 0.03056, + "grad_norm": 0.658467407296622, + "learning_rate": 0.003, + "loss": 4.134, + "step": 3056 + }, + { + "epoch": 0.03057, + "grad_norm": 0.6556069512553365, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 3057 + }, + { + "epoch": 0.03058, + "grad_norm": 0.5717711332151091, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 3058 + }, + { + "epoch": 0.03059, + "grad_norm": 0.493409526384602, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 3059 + }, + { + "epoch": 0.0306, + "grad_norm": 0.5082424754021986, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 3060 + }, + { + "epoch": 0.03061, + "grad_norm": 0.4898096783004193, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 3061 + }, + { + "epoch": 0.03062, + "grad_norm": 0.5061719030039263, + "learning_rate": 0.003, + "loss": 4.144, + "step": 3062 + }, + { + "epoch": 0.03063, + "grad_norm": 0.6337804215976139, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 3063 + }, + { + "epoch": 0.03064, + "grad_norm": 0.7363596095895648, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 3064 + }, + { + "epoch": 0.03065, + "grad_norm": 0.7804431388454032, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 3065 + }, + { + "epoch": 0.03066, + "grad_norm": 0.8048441855814717, + "learning_rate": 0.003, + "loss": 4.14, + "step": 3066 + }, + { + "epoch": 0.03067, + "grad_norm": 0.936660465754291, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 3067 + }, + { + "epoch": 0.03068, + "grad_norm": 1.199111088765608, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 3068 + }, + { + "epoch": 0.03069, + "grad_norm": 1.024080179977825, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 3069 + }, + { + "epoch": 0.0307, + "grad_norm": 0.9087822982161698, + "learning_rate": 0.003, + "loss": 4.125, + "step": 3070 + }, + { + "epoch": 0.03071, + "grad_norm": 0.773169782429931, + "learning_rate": 0.003, + "loss": 4.121, + "step": 3071 + }, + { + "epoch": 0.03072, + "grad_norm": 0.7385627562002322, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 3072 + }, + { + "epoch": 0.03073, + "grad_norm": 0.8590738611383513, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 3073 + }, + { + "epoch": 0.03074, + "grad_norm": 0.7460812381201545, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 3074 + }, + { + "epoch": 0.03075, + "grad_norm": 0.9060557817564706, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 3075 + }, + { + "epoch": 0.03076, + "grad_norm": 0.9352871058483179, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 3076 + }, + { + "epoch": 0.03077, + "grad_norm": 0.8648002758853153, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 3077 + }, + { + "epoch": 0.03078, + "grad_norm": 0.7596804078061197, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 3078 + }, + { + "epoch": 0.03079, + "grad_norm": 0.7757175523238533, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 3079 + }, + { + "epoch": 0.0308, + "grad_norm": 0.8209364936522558, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 3080 + }, + { + "epoch": 0.03081, + "grad_norm": 0.8477812326348159, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 3081 + }, + { + "epoch": 0.03082, + "grad_norm": 0.8013577673715985, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 3082 + }, + { + "epoch": 0.03083, + "grad_norm": 0.7134216891930999, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 3083 + }, + { + "epoch": 0.03084, + "grad_norm": 0.7257190995117091, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 3084 + }, + { + "epoch": 0.03085, + "grad_norm": 0.8088783726503086, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 3085 + }, + { + "epoch": 0.03086, + "grad_norm": 0.8932889669749948, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 3086 + }, + { + "epoch": 0.03087, + "grad_norm": 0.9271283171010463, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 3087 + }, + { + "epoch": 0.03088, + "grad_norm": 1.0610338130344832, + "learning_rate": 0.003, + "loss": 4.1534, + "step": 3088 + }, + { + "epoch": 0.03089, + "grad_norm": 0.9082960513513103, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 3089 + }, + { + "epoch": 0.0309, + "grad_norm": 0.8382419044332593, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 3090 + }, + { + "epoch": 0.03091, + "grad_norm": 0.7721347908154246, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 3091 + }, + { + "epoch": 0.03092, + "grad_norm": 0.5770289909808661, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 3092 + }, + { + "epoch": 0.03093, + "grad_norm": 0.6048583748826258, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 3093 + }, + { + "epoch": 0.03094, + "grad_norm": 0.5039228690347607, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 3094 + }, + { + "epoch": 0.03095, + "grad_norm": 0.4932619648402882, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 3095 + }, + { + "epoch": 0.03096, + "grad_norm": 0.44699630710548827, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 3096 + }, + { + "epoch": 0.03097, + "grad_norm": 0.42800774052172613, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 3097 + }, + { + "epoch": 0.03098, + "grad_norm": 0.5150974229730493, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3098 + }, + { + "epoch": 0.03099, + "grad_norm": 0.6808104155412404, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 3099 + }, + { + "epoch": 0.031, + "grad_norm": 1.1124941338393186, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 3100 + }, + { + "epoch": 0.03101, + "grad_norm": 1.0220222334230853, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 3101 + }, + { + "epoch": 0.03102, + "grad_norm": 0.6635321325690858, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 3102 + }, + { + "epoch": 0.03103, + "grad_norm": 0.7879413374703326, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 3103 + }, + { + "epoch": 0.03104, + "grad_norm": 1.0336622472450878, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 3104 + }, + { + "epoch": 0.03105, + "grad_norm": 0.8225297439648239, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 3105 + }, + { + "epoch": 0.03106, + "grad_norm": 0.9200113560338348, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 3106 + }, + { + "epoch": 0.03107, + "grad_norm": 0.8693837130689374, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 3107 + }, + { + "epoch": 0.03108, + "grad_norm": 0.7481880861084276, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 3108 + }, + { + "epoch": 0.03109, + "grad_norm": 0.7161901524718123, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 3109 + }, + { + "epoch": 0.0311, + "grad_norm": 0.7717771970106354, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 3110 + }, + { + "epoch": 0.03111, + "grad_norm": 0.7918589418817704, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 3111 + }, + { + "epoch": 0.03112, + "grad_norm": 0.8112295219654945, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 3112 + }, + { + "epoch": 0.03113, + "grad_norm": 0.8724658411048615, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 3113 + }, + { + "epoch": 0.03114, + "grad_norm": 1.0040711142290948, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 3114 + }, + { + "epoch": 0.03115, + "grad_norm": 1.2108105696861353, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 3115 + }, + { + "epoch": 0.03116, + "grad_norm": 0.9386810822379548, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 3116 + }, + { + "epoch": 0.03117, + "grad_norm": 0.7805193907050799, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 3117 + }, + { + "epoch": 0.03118, + "grad_norm": 0.9100639956456111, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 3118 + }, + { + "epoch": 0.03119, + "grad_norm": 0.9031494351027202, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 3119 + }, + { + "epoch": 0.0312, + "grad_norm": 0.9920122086984918, + "learning_rate": 0.003, + "loss": 4.154, + "step": 3120 + }, + { + "epoch": 0.03121, + "grad_norm": 1.24679349546566, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 3121 + }, + { + "epoch": 0.03122, + "grad_norm": 0.8914719736220875, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 3122 + }, + { + "epoch": 0.03123, + "grad_norm": 0.7579991423334604, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 3123 + }, + { + "epoch": 0.03124, + "grad_norm": 0.7621871567104618, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 3124 + }, + { + "epoch": 0.03125, + "grad_norm": 0.8062917259017726, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3125 + }, + { + "epoch": 0.03126, + "grad_norm": 0.833712338155859, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 3126 + }, + { + "epoch": 0.03127, + "grad_norm": 0.8998971933566047, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 3127 + }, + { + "epoch": 0.03128, + "grad_norm": 0.9380372829323406, + "learning_rate": 0.003, + "loss": 4.126, + "step": 3128 + }, + { + "epoch": 0.03129, + "grad_norm": 0.9521516498255859, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 3129 + }, + { + "epoch": 0.0313, + "grad_norm": 0.8951772330842159, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 3130 + }, + { + "epoch": 0.03131, + "grad_norm": 0.7755621391919874, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 3131 + }, + { + "epoch": 0.03132, + "grad_norm": 0.8129807814076891, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 3132 + }, + { + "epoch": 0.03133, + "grad_norm": 0.9303049881214197, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 3133 + }, + { + "epoch": 0.03134, + "grad_norm": 0.9415721268309895, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 3134 + }, + { + "epoch": 0.03135, + "grad_norm": 0.9104183542988432, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 3135 + }, + { + "epoch": 0.03136, + "grad_norm": 0.9100236222536682, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 3136 + }, + { + "epoch": 0.03137, + "grad_norm": 0.8798690651639479, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 3137 + }, + { + "epoch": 0.03138, + "grad_norm": 0.7903392795080382, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 3138 + }, + { + "epoch": 0.03139, + "grad_norm": 0.7667246215003086, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 3139 + }, + { + "epoch": 0.0314, + "grad_norm": 0.7544018295001089, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 3140 + }, + { + "epoch": 0.03141, + "grad_norm": 0.6716524130054017, + "learning_rate": 0.003, + "loss": 4.154, + "step": 3141 + }, + { + "epoch": 0.03142, + "grad_norm": 0.6382378457514613, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 3142 + }, + { + "epoch": 0.03143, + "grad_norm": 0.7526757290066967, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 3143 + }, + { + "epoch": 0.03144, + "grad_norm": 0.938489146233584, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 3144 + }, + { + "epoch": 0.03145, + "grad_norm": 1.0493538668141305, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 3145 + }, + { + "epoch": 0.03146, + "grad_norm": 0.9048504750556206, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 3146 + }, + { + "epoch": 0.03147, + "grad_norm": 0.9152634263998236, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 3147 + }, + { + "epoch": 0.03148, + "grad_norm": 0.878607908387141, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 3148 + }, + { + "epoch": 0.03149, + "grad_norm": 0.6944713264838394, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 3149 + }, + { + "epoch": 0.0315, + "grad_norm": 0.6811030653109197, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 3150 + }, + { + "epoch": 0.03151, + "grad_norm": 0.7166494028947639, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 3151 + }, + { + "epoch": 0.03152, + "grad_norm": 0.7180332750616866, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 3152 + }, + { + "epoch": 0.03153, + "grad_norm": 0.8734661321822886, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 3153 + }, + { + "epoch": 0.03154, + "grad_norm": 0.9738879775883073, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 3154 + }, + { + "epoch": 0.03155, + "grad_norm": 0.8618576527781467, + "learning_rate": 0.003, + "loss": 4.127, + "step": 3155 + }, + { + "epoch": 0.03156, + "grad_norm": 0.7300214083601039, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 3156 + }, + { + "epoch": 0.03157, + "grad_norm": 0.6260105425384149, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 3157 + }, + { + "epoch": 0.03158, + "grad_norm": 0.682384994546209, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 3158 + }, + { + "epoch": 0.03159, + "grad_norm": 0.7718937672456789, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3159 + }, + { + "epoch": 0.0316, + "grad_norm": 0.6586793613223971, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 3160 + }, + { + "epoch": 0.03161, + "grad_norm": 0.6187219141678427, + "learning_rate": 0.003, + "loss": 4.133, + "step": 3161 + }, + { + "epoch": 0.03162, + "grad_norm": 0.7732103677060469, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 3162 + }, + { + "epoch": 0.03163, + "grad_norm": 0.9681852590593839, + "learning_rate": 0.003, + "loss": 4.129, + "step": 3163 + }, + { + "epoch": 0.03164, + "grad_norm": 0.983612124215567, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 3164 + }, + { + "epoch": 0.03165, + "grad_norm": 0.8828238997897765, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 3165 + }, + { + "epoch": 0.03166, + "grad_norm": 0.9082253980909, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3166 + }, + { + "epoch": 0.03167, + "grad_norm": 0.8883995374606223, + "learning_rate": 0.003, + "loss": 4.11, + "step": 3167 + }, + { + "epoch": 0.03168, + "grad_norm": 0.8814873185877611, + "learning_rate": 0.003, + "loss": 4.1367, + "step": 3168 + }, + { + "epoch": 0.03169, + "grad_norm": 0.8358140037799937, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 3169 + }, + { + "epoch": 0.0317, + "grad_norm": 0.9091839519331014, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 3170 + }, + { + "epoch": 0.03171, + "grad_norm": 0.7528676805145117, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 3171 + }, + { + "epoch": 0.03172, + "grad_norm": 0.6184044517416146, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3172 + }, + { + "epoch": 0.03173, + "grad_norm": 0.7308731124601682, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 3173 + }, + { + "epoch": 0.03174, + "grad_norm": 0.8588565114301099, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 3174 + }, + { + "epoch": 0.03175, + "grad_norm": 1.2162741354980293, + "learning_rate": 0.003, + "loss": 4.1527, + "step": 3175 + }, + { + "epoch": 0.03176, + "grad_norm": 0.9397594133644137, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 3176 + }, + { + "epoch": 0.03177, + "grad_norm": 0.9189809109610272, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 3177 + }, + { + "epoch": 0.03178, + "grad_norm": 0.8047272875332512, + "learning_rate": 0.003, + "loss": 4.1657, + "step": 3178 + }, + { + "epoch": 0.03179, + "grad_norm": 0.932150949680668, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 3179 + }, + { + "epoch": 0.0318, + "grad_norm": 0.9592877689147434, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 3180 + }, + { + "epoch": 0.03181, + "grad_norm": 0.9595136242555029, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 3181 + }, + { + "epoch": 0.03182, + "grad_norm": 0.9565837686401868, + "learning_rate": 0.003, + "loss": 4.142, + "step": 3182 + }, + { + "epoch": 0.03183, + "grad_norm": 0.9741938675445059, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 3183 + }, + { + "epoch": 0.03184, + "grad_norm": 0.9857805137742832, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 3184 + }, + { + "epoch": 0.03185, + "grad_norm": 0.9176062885733086, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 3185 + }, + { + "epoch": 0.03186, + "grad_norm": 0.8884334076590751, + "learning_rate": 0.003, + "loss": 4.1367, + "step": 3186 + }, + { + "epoch": 0.03187, + "grad_norm": 0.8658125563673901, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 3187 + }, + { + "epoch": 0.03188, + "grad_norm": 0.8999051178552305, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 3188 + }, + { + "epoch": 0.03189, + "grad_norm": 0.7859126574348619, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 3189 + }, + { + "epoch": 0.0319, + "grad_norm": 0.8606531029730619, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 3190 + }, + { + "epoch": 0.03191, + "grad_norm": 0.9918147629707751, + "learning_rate": 0.003, + "loss": 4.149, + "step": 3191 + }, + { + "epoch": 0.03192, + "grad_norm": 1.0056390969187974, + "learning_rate": 0.003, + "loss": 4.132, + "step": 3192 + }, + { + "epoch": 0.03193, + "grad_norm": 0.8860275310714892, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 3193 + }, + { + "epoch": 0.03194, + "grad_norm": 0.9173243755728898, + "learning_rate": 0.003, + "loss": 4.1534, + "step": 3194 + }, + { + "epoch": 0.03195, + "grad_norm": 0.8392077257777418, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 3195 + }, + { + "epoch": 0.03196, + "grad_norm": 0.7352847590623333, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 3196 + }, + { + "epoch": 0.03197, + "grad_norm": 0.715497912918627, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 3197 + }, + { + "epoch": 0.03198, + "grad_norm": 0.6376443032868672, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 3198 + }, + { + "epoch": 0.03199, + "grad_norm": 0.5811704510301426, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 3199 + }, + { + "epoch": 0.032, + "grad_norm": 0.5381272704276707, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3200 + }, + { + "epoch": 0.03201, + "grad_norm": 0.5120575482476429, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 3201 + }, + { + "epoch": 0.03202, + "grad_norm": 0.5958579649120576, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3202 + }, + { + "epoch": 0.03203, + "grad_norm": 0.7890366994930119, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 3203 + }, + { + "epoch": 0.03204, + "grad_norm": 0.9077676416657651, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 3204 + }, + { + "epoch": 0.03205, + "grad_norm": 0.754000654803963, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 3205 + }, + { + "epoch": 0.03206, + "grad_norm": 0.600502725716816, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 3206 + }, + { + "epoch": 0.03207, + "grad_norm": 0.6048683504208053, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 3207 + }, + { + "epoch": 0.03208, + "grad_norm": 0.7129517255787963, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 3208 + }, + { + "epoch": 0.03209, + "grad_norm": 0.8158420961472652, + "learning_rate": 0.003, + "loss": 4.1516, + "step": 3209 + }, + { + "epoch": 0.0321, + "grad_norm": 0.7005791064338225, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 3210 + }, + { + "epoch": 0.03211, + "grad_norm": 0.6098399298288999, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 3211 + }, + { + "epoch": 0.03212, + "grad_norm": 0.6538561751290226, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 3212 + }, + { + "epoch": 0.03213, + "grad_norm": 0.5790618052777453, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 3213 + }, + { + "epoch": 0.03214, + "grad_norm": 0.6875003076630353, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 3214 + }, + { + "epoch": 0.03215, + "grad_norm": 0.799314724709517, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 3215 + }, + { + "epoch": 0.03216, + "grad_norm": 0.8890107745271224, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 3216 + }, + { + "epoch": 0.03217, + "grad_norm": 0.7725727639400009, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3217 + }, + { + "epoch": 0.03218, + "grad_norm": 0.7089931384963014, + "learning_rate": 0.003, + "loss": 4.096, + "step": 3218 + }, + { + "epoch": 0.03219, + "grad_norm": 0.755051345678096, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 3219 + }, + { + "epoch": 0.0322, + "grad_norm": 0.9521310432872768, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 3220 + }, + { + "epoch": 0.03221, + "grad_norm": 0.8616808278278768, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 3221 + }, + { + "epoch": 0.03222, + "grad_norm": 0.873993324764959, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 3222 + }, + { + "epoch": 0.03223, + "grad_norm": 0.8264507288043698, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 3223 + }, + { + "epoch": 0.03224, + "grad_norm": 0.8289275480200801, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 3224 + }, + { + "epoch": 0.03225, + "grad_norm": 0.7514528620660997, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3225 + }, + { + "epoch": 0.03226, + "grad_norm": 0.7197038673889892, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 3226 + }, + { + "epoch": 0.03227, + "grad_norm": 0.8892719165128083, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 3227 + }, + { + "epoch": 0.03228, + "grad_norm": 0.9570294472379692, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 3228 + }, + { + "epoch": 0.03229, + "grad_norm": 0.9375751530879919, + "learning_rate": 0.003, + "loss": 4.12, + "step": 3229 + }, + { + "epoch": 0.0323, + "grad_norm": 0.9755322805010285, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3230 + }, + { + "epoch": 0.03231, + "grad_norm": 1.2740697032530537, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 3231 + }, + { + "epoch": 0.03232, + "grad_norm": 0.8205568087154713, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3232 + }, + { + "epoch": 0.03233, + "grad_norm": 0.7719838274157066, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 3233 + }, + { + "epoch": 0.03234, + "grad_norm": 0.7880067935874491, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 3234 + }, + { + "epoch": 0.03235, + "grad_norm": 0.8025776008313478, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 3235 + }, + { + "epoch": 0.03236, + "grad_norm": 1.0153689767602323, + "learning_rate": 0.003, + "loss": 4.124, + "step": 3236 + }, + { + "epoch": 0.03237, + "grad_norm": 1.0960896554896313, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 3237 + }, + { + "epoch": 0.03238, + "grad_norm": 0.9653438379897337, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 3238 + }, + { + "epoch": 0.03239, + "grad_norm": 0.9477216778661465, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 3239 + }, + { + "epoch": 0.0324, + "grad_norm": 1.0000644526629416, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 3240 + }, + { + "epoch": 0.03241, + "grad_norm": 0.9623250673117251, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 3241 + }, + { + "epoch": 0.03242, + "grad_norm": 1.00080805991717, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 3242 + }, + { + "epoch": 0.03243, + "grad_norm": 0.864922721509246, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 3243 + }, + { + "epoch": 0.03244, + "grad_norm": 0.7896866879394736, + "learning_rate": 0.003, + "loss": 4.13, + "step": 3244 + }, + { + "epoch": 0.03245, + "grad_norm": 0.723475186710102, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 3245 + }, + { + "epoch": 0.03246, + "grad_norm": 0.6786037554737262, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3246 + }, + { + "epoch": 0.03247, + "grad_norm": 0.6637920051672874, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 3247 + }, + { + "epoch": 0.03248, + "grad_norm": 0.7161721376620257, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 3248 + }, + { + "epoch": 0.03249, + "grad_norm": 0.7558986889375228, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 3249 + }, + { + "epoch": 0.0325, + "grad_norm": 0.7122471010381649, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 3250 + }, + { + "epoch": 0.03251, + "grad_norm": 0.7576764147111074, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 3251 + }, + { + "epoch": 0.03252, + "grad_norm": 0.6662542349208606, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 3252 + }, + { + "epoch": 0.03253, + "grad_norm": 0.6584169138349598, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 3253 + }, + { + "epoch": 0.03254, + "grad_norm": 0.7740732689052414, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 3254 + }, + { + "epoch": 0.03255, + "grad_norm": 0.9473846923391861, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 3255 + }, + { + "epoch": 0.03256, + "grad_norm": 0.929085241706198, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 3256 + }, + { + "epoch": 0.03257, + "grad_norm": 0.9501158349657576, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 3257 + }, + { + "epoch": 0.03258, + "grad_norm": 0.9141127030789927, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 3258 + }, + { + "epoch": 0.03259, + "grad_norm": 0.7903443980069681, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 3259 + }, + { + "epoch": 0.0326, + "grad_norm": 0.7186252333798928, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 3260 + }, + { + "epoch": 0.03261, + "grad_norm": 0.6176139155287416, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 3261 + }, + { + "epoch": 0.03262, + "grad_norm": 0.5770619992009712, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 3262 + }, + { + "epoch": 0.03263, + "grad_norm": 0.6179754292775074, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 3263 + }, + { + "epoch": 0.03264, + "grad_norm": 0.7950184256183296, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 3264 + }, + { + "epoch": 0.03265, + "grad_norm": 0.8260324885629711, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 3265 + }, + { + "epoch": 0.03266, + "grad_norm": 0.9331143722743607, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 3266 + }, + { + "epoch": 0.03267, + "grad_norm": 0.8951624994125074, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 3267 + }, + { + "epoch": 0.03268, + "grad_norm": 0.8052096070419401, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 3268 + }, + { + "epoch": 0.03269, + "grad_norm": 0.8166220426949835, + "learning_rate": 0.003, + "loss": 4.115, + "step": 3269 + }, + { + "epoch": 0.0327, + "grad_norm": 0.7018775720106071, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 3270 + }, + { + "epoch": 0.03271, + "grad_norm": 0.7111861211006557, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 3271 + }, + { + "epoch": 0.03272, + "grad_norm": 0.8446103418486255, + "learning_rate": 0.003, + "loss": 4.111, + "step": 3272 + }, + { + "epoch": 0.03273, + "grad_norm": 0.9056741913837429, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 3273 + }, + { + "epoch": 0.03274, + "grad_norm": 0.8569785882714194, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 3274 + }, + { + "epoch": 0.03275, + "grad_norm": 0.8267745539239577, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 3275 + }, + { + "epoch": 0.03276, + "grad_norm": 0.8644593844634266, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 3276 + }, + { + "epoch": 0.03277, + "grad_norm": 1.0535041439341994, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 3277 + }, + { + "epoch": 0.03278, + "grad_norm": 0.993769735898057, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3278 + }, + { + "epoch": 0.03279, + "grad_norm": 1.1362817862452346, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 3279 + }, + { + "epoch": 0.0328, + "grad_norm": 0.785895471391003, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 3280 + }, + { + "epoch": 0.03281, + "grad_norm": 0.8032123764058076, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 3281 + }, + { + "epoch": 0.03282, + "grad_norm": 0.9387436206649247, + "learning_rate": 0.003, + "loss": 4.103, + "step": 3282 + }, + { + "epoch": 0.03283, + "grad_norm": 1.1369187077903837, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 3283 + }, + { + "epoch": 0.03284, + "grad_norm": 0.9173031886562234, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 3284 + }, + { + "epoch": 0.03285, + "grad_norm": 1.0358430479898526, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 3285 + }, + { + "epoch": 0.03286, + "grad_norm": 0.8561133575195394, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 3286 + }, + { + "epoch": 0.03287, + "grad_norm": 0.8329432797036997, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 3287 + }, + { + "epoch": 0.03288, + "grad_norm": 0.8334315378634501, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 3288 + }, + { + "epoch": 0.03289, + "grad_norm": 0.6838577814234701, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3289 + }, + { + "epoch": 0.0329, + "grad_norm": 0.6781901884347332, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 3290 + }, + { + "epoch": 0.03291, + "grad_norm": 0.6974761035595561, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 3291 + }, + { + "epoch": 0.03292, + "grad_norm": 0.8026120729452656, + "learning_rate": 0.003, + "loss": 4.093, + "step": 3292 + }, + { + "epoch": 0.03293, + "grad_norm": 0.8409575609279364, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3293 + }, + { + "epoch": 0.03294, + "grad_norm": 0.9672433152587747, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 3294 + }, + { + "epoch": 0.03295, + "grad_norm": 0.952646103687869, + "learning_rate": 0.003, + "loss": 4.117, + "step": 3295 + }, + { + "epoch": 0.03296, + "grad_norm": 0.898170814180958, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3296 + }, + { + "epoch": 0.03297, + "grad_norm": 1.021226648982521, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 3297 + }, + { + "epoch": 0.03298, + "grad_norm": 1.0948474288454675, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 3298 + }, + { + "epoch": 0.03299, + "grad_norm": 1.0192347868788103, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 3299 + }, + { + "epoch": 0.033, + "grad_norm": 1.0495723694795078, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 3300 + }, + { + "epoch": 0.03301, + "grad_norm": 1.0851645957051337, + "learning_rate": 0.003, + "loss": 4.155, + "step": 3301 + }, + { + "epoch": 0.03302, + "grad_norm": 0.9231732440912226, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 3302 + }, + { + "epoch": 0.03303, + "grad_norm": 0.8311956950138633, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 3303 + }, + { + "epoch": 0.03304, + "grad_norm": 0.7259197244925034, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 3304 + }, + { + "epoch": 0.03305, + "grad_norm": 0.6328132146169181, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 3305 + }, + { + "epoch": 0.03306, + "grad_norm": 0.680683566379763, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 3306 + }, + { + "epoch": 0.03307, + "grad_norm": 0.6927470498017374, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 3307 + }, + { + "epoch": 0.03308, + "grad_norm": 0.6675628417853611, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 3308 + }, + { + "epoch": 0.03309, + "grad_norm": 0.7202067492998604, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 3309 + }, + { + "epoch": 0.0331, + "grad_norm": 0.7699244574031957, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 3310 + }, + { + "epoch": 0.03311, + "grad_norm": 0.8693826092494591, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 3311 + }, + { + "epoch": 0.03312, + "grad_norm": 1.172876733274895, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 3312 + }, + { + "epoch": 0.03313, + "grad_norm": 0.9783748676175467, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 3313 + }, + { + "epoch": 0.03314, + "grad_norm": 0.9677078544866878, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 3314 + }, + { + "epoch": 0.03315, + "grad_norm": 0.8922546596507663, + "learning_rate": 0.003, + "loss": 4.131, + "step": 3315 + }, + { + "epoch": 0.03316, + "grad_norm": 0.8601480724971649, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 3316 + }, + { + "epoch": 0.03317, + "grad_norm": 0.8689703079288349, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3317 + }, + { + "epoch": 0.03318, + "grad_norm": 0.8404981974903554, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 3318 + }, + { + "epoch": 0.03319, + "grad_norm": 0.7816646536926355, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 3319 + }, + { + "epoch": 0.0332, + "grad_norm": 0.7686980103144088, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 3320 + }, + { + "epoch": 0.03321, + "grad_norm": 0.6688242043093231, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 3321 + }, + { + "epoch": 0.03322, + "grad_norm": 0.6171612564014062, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 3322 + }, + { + "epoch": 0.03323, + "grad_norm": 0.5970300363358997, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 3323 + }, + { + "epoch": 0.03324, + "grad_norm": 0.5953328227942196, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 3324 + }, + { + "epoch": 0.03325, + "grad_norm": 0.6169531392155312, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 3325 + }, + { + "epoch": 0.03326, + "grad_norm": 0.7413959337275615, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 3326 + }, + { + "epoch": 0.03327, + "grad_norm": 0.9232832965259093, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 3327 + }, + { + "epoch": 0.03328, + "grad_norm": 1.0485483129958542, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 3328 + }, + { + "epoch": 0.03329, + "grad_norm": 0.8420326285667852, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 3329 + }, + { + "epoch": 0.0333, + "grad_norm": 0.7052108452174394, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 3330 + }, + { + "epoch": 0.03331, + "grad_norm": 0.728092857943501, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 3331 + }, + { + "epoch": 0.03332, + "grad_norm": 0.9200654034817243, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 3332 + }, + { + "epoch": 0.03333, + "grad_norm": 0.9758232395574222, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 3333 + }, + { + "epoch": 0.03334, + "grad_norm": 1.0419984825931974, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 3334 + }, + { + "epoch": 0.03335, + "grad_norm": 0.9936868073161941, + "learning_rate": 0.003, + "loss": 4.1367, + "step": 3335 + }, + { + "epoch": 0.03336, + "grad_norm": 0.9188784917760547, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 3336 + }, + { + "epoch": 0.03337, + "grad_norm": 0.9676171269128566, + "learning_rate": 0.003, + "loss": 4.113, + "step": 3337 + }, + { + "epoch": 0.03338, + "grad_norm": 1.2398133112982033, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 3338 + }, + { + "epoch": 0.03339, + "grad_norm": 0.8757441136351082, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 3339 + }, + { + "epoch": 0.0334, + "grad_norm": 0.803431258859875, + "learning_rate": 0.003, + "loss": 4.102, + "step": 3340 + }, + { + "epoch": 0.03341, + "grad_norm": 0.7508144599659284, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 3341 + }, + { + "epoch": 0.03342, + "grad_norm": 0.6682938205970289, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 3342 + }, + { + "epoch": 0.03343, + "grad_norm": 0.6286049563460856, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 3343 + }, + { + "epoch": 0.03344, + "grad_norm": 0.7479131859288655, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 3344 + }, + { + "epoch": 0.03345, + "grad_norm": 0.9317587885861235, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3345 + }, + { + "epoch": 0.03346, + "grad_norm": 1.1810549914532213, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 3346 + }, + { + "epoch": 0.03347, + "grad_norm": 0.803326085791172, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 3347 + }, + { + "epoch": 0.03348, + "grad_norm": 0.6682684868194113, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 3348 + }, + { + "epoch": 0.03349, + "grad_norm": 0.7128883398636897, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 3349 + }, + { + "epoch": 0.0335, + "grad_norm": 0.8166001203873546, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 3350 + }, + { + "epoch": 0.03351, + "grad_norm": 0.8984169700598436, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 3351 + }, + { + "epoch": 0.03352, + "grad_norm": 0.891385033781036, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 3352 + }, + { + "epoch": 0.03353, + "grad_norm": 0.7436503547090926, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 3353 + }, + { + "epoch": 0.03354, + "grad_norm": 0.7000314128832471, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 3354 + }, + { + "epoch": 0.03355, + "grad_norm": 0.8076872538085662, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 3355 + }, + { + "epoch": 0.03356, + "grad_norm": 0.9065676166442113, + "learning_rate": 0.003, + "loss": 4.141, + "step": 3356 + }, + { + "epoch": 0.03357, + "grad_norm": 0.8283445403636854, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 3357 + }, + { + "epoch": 0.03358, + "grad_norm": 0.711728954112842, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 3358 + }, + { + "epoch": 0.03359, + "grad_norm": 0.6700966992589463, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 3359 + }, + { + "epoch": 0.0336, + "grad_norm": 0.6990021919801547, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 3360 + }, + { + "epoch": 0.03361, + "grad_norm": 0.6512206263292598, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 3361 + }, + { + "epoch": 0.03362, + "grad_norm": 0.6471137948913391, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 3362 + }, + { + "epoch": 0.03363, + "grad_norm": 0.6486978718096742, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 3363 + }, + { + "epoch": 0.03364, + "grad_norm": 0.7056083808952703, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 3364 + }, + { + "epoch": 0.03365, + "grad_norm": 0.7889254390774209, + "learning_rate": 0.003, + "loss": 4.096, + "step": 3365 + }, + { + "epoch": 0.03366, + "grad_norm": 0.8858274543513762, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 3366 + }, + { + "epoch": 0.03367, + "grad_norm": 0.8996550932306419, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 3367 + }, + { + "epoch": 0.03368, + "grad_norm": 0.9535511924749507, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 3368 + }, + { + "epoch": 0.03369, + "grad_norm": 1.025280504328262, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 3369 + }, + { + "epoch": 0.0337, + "grad_norm": 0.9606870288728299, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 3370 + }, + { + "epoch": 0.03371, + "grad_norm": 0.911846692093779, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 3371 + }, + { + "epoch": 0.03372, + "grad_norm": 0.8892950330733644, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 3372 + }, + { + "epoch": 0.03373, + "grad_norm": 0.9613355807854816, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 3373 + }, + { + "epoch": 0.03374, + "grad_norm": 0.9293495359333047, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 3374 + }, + { + "epoch": 0.03375, + "grad_norm": 0.9945254272923405, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 3375 + }, + { + "epoch": 0.03376, + "grad_norm": 0.9458311049733065, + "learning_rate": 0.003, + "loss": 4.085, + "step": 3376 + }, + { + "epoch": 0.03377, + "grad_norm": 1.0143873093094964, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 3377 + }, + { + "epoch": 0.03378, + "grad_norm": 0.9867149119212901, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 3378 + }, + { + "epoch": 0.03379, + "grad_norm": 0.909410152436646, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 3379 + }, + { + "epoch": 0.0338, + "grad_norm": 0.9582033813869726, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 3380 + }, + { + "epoch": 0.03381, + "grad_norm": 1.0515732546439378, + "learning_rate": 0.003, + "loss": 4.1483, + "step": 3381 + }, + { + "epoch": 0.03382, + "grad_norm": 0.9000543481782362, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 3382 + }, + { + "epoch": 0.03383, + "grad_norm": 0.7851905547422763, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 3383 + }, + { + "epoch": 0.03384, + "grad_norm": 0.7425344508528795, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 3384 + }, + { + "epoch": 0.03385, + "grad_norm": 0.7242449162586684, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 3385 + }, + { + "epoch": 0.03386, + "grad_norm": 0.7178134850877214, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 3386 + }, + { + "epoch": 0.03387, + "grad_norm": 0.6034717016906057, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 3387 + }, + { + "epoch": 0.03388, + "grad_norm": 0.6250385791221151, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 3388 + }, + { + "epoch": 0.03389, + "grad_norm": 0.5925814922250926, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 3389 + }, + { + "epoch": 0.0339, + "grad_norm": 0.6938565572265246, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 3390 + }, + { + "epoch": 0.03391, + "grad_norm": 0.8588428651999951, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 3391 + }, + { + "epoch": 0.03392, + "grad_norm": 1.0828213831544788, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 3392 + }, + { + "epoch": 0.03393, + "grad_norm": 0.9139268583774075, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 3393 + }, + { + "epoch": 0.03394, + "grad_norm": 0.8962526152154793, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 3394 + }, + { + "epoch": 0.03395, + "grad_norm": 0.9550321739497196, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 3395 + }, + { + "epoch": 0.03396, + "grad_norm": 0.9058142633647185, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 3396 + }, + { + "epoch": 0.03397, + "grad_norm": 1.030198089545334, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 3397 + }, + { + "epoch": 0.03398, + "grad_norm": 0.9817159965257135, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 3398 + }, + { + "epoch": 0.03399, + "grad_norm": 0.7537874005656849, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 3399 + }, + { + "epoch": 0.034, + "grad_norm": 0.7534398090951075, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 3400 + }, + { + "epoch": 0.03401, + "grad_norm": 0.846711011216498, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 3401 + }, + { + "epoch": 0.03402, + "grad_norm": 1.0304564422923, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 3402 + }, + { + "epoch": 0.03403, + "grad_norm": 0.8955445182551958, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 3403 + }, + { + "epoch": 0.03404, + "grad_norm": 0.8683725321318734, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 3404 + }, + { + "epoch": 0.03405, + "grad_norm": 0.8700430489216084, + "learning_rate": 0.003, + "loss": 4.126, + "step": 3405 + }, + { + "epoch": 0.03406, + "grad_norm": 0.856851032889231, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 3406 + }, + { + "epoch": 0.03407, + "grad_norm": 0.8583304198024593, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 3407 + }, + { + "epoch": 0.03408, + "grad_norm": 0.8259652751248255, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 3408 + }, + { + "epoch": 0.03409, + "grad_norm": 0.9359241584231073, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 3409 + }, + { + "epoch": 0.0341, + "grad_norm": 0.938170801150728, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 3410 + }, + { + "epoch": 0.03411, + "grad_norm": 0.9444575938733144, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 3411 + }, + { + "epoch": 0.03412, + "grad_norm": 0.9224218450030172, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 3412 + }, + { + "epoch": 0.03413, + "grad_norm": 0.9312506396494133, + "learning_rate": 0.003, + "loss": 4.1534, + "step": 3413 + }, + { + "epoch": 0.03414, + "grad_norm": 0.9761972676777384, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 3414 + }, + { + "epoch": 0.03415, + "grad_norm": 1.189613033448552, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 3415 + }, + { + "epoch": 0.03416, + "grad_norm": 0.8255201977611025, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 3416 + }, + { + "epoch": 0.03417, + "grad_norm": 0.8947733285821016, + "learning_rate": 0.003, + "loss": 4.125, + "step": 3417 + }, + { + "epoch": 0.03418, + "grad_norm": 0.7984406024805614, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 3418 + }, + { + "epoch": 0.03419, + "grad_norm": 0.6630379610145436, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 3419 + }, + { + "epoch": 0.0342, + "grad_norm": 0.5975770128227218, + "learning_rate": 0.003, + "loss": 4.12, + "step": 3420 + }, + { + "epoch": 0.03421, + "grad_norm": 0.6059481665025913, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 3421 + }, + { + "epoch": 0.03422, + "grad_norm": 0.6007015602316165, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 3422 + }, + { + "epoch": 0.03423, + "grad_norm": 0.6213374488193623, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 3423 + }, + { + "epoch": 0.03424, + "grad_norm": 0.7357379283476251, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 3424 + }, + { + "epoch": 0.03425, + "grad_norm": 1.1388627740722068, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 3425 + }, + { + "epoch": 0.03426, + "grad_norm": 1.2676165376349322, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 3426 + }, + { + "epoch": 0.03427, + "grad_norm": 0.8591095580169765, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 3427 + }, + { + "epoch": 0.03428, + "grad_norm": 0.7637078879713047, + "learning_rate": 0.003, + "loss": 4.096, + "step": 3428 + }, + { + "epoch": 0.03429, + "grad_norm": 0.7598510773275634, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 3429 + }, + { + "epoch": 0.0343, + "grad_norm": 0.7189854251191673, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3430 + }, + { + "epoch": 0.03431, + "grad_norm": 0.6960848938203806, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 3431 + }, + { + "epoch": 0.03432, + "grad_norm": 0.7905143093567869, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 3432 + }, + { + "epoch": 0.03433, + "grad_norm": 0.6763604081182782, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 3433 + }, + { + "epoch": 0.03434, + "grad_norm": 0.6584657241744457, + "learning_rate": 0.003, + "loss": 4.098, + "step": 3434 + }, + { + "epoch": 0.03435, + "grad_norm": 0.6146074495577768, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 3435 + }, + { + "epoch": 0.03436, + "grad_norm": 0.6821260061437763, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 3436 + }, + { + "epoch": 0.03437, + "grad_norm": 0.7002392953868479, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 3437 + }, + { + "epoch": 0.03438, + "grad_norm": 0.6865303492632201, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 3438 + }, + { + "epoch": 0.03439, + "grad_norm": 0.6862765139756455, + "learning_rate": 0.003, + "loss": 4.109, + "step": 3439 + }, + { + "epoch": 0.0344, + "grad_norm": 0.7374953588444623, + "learning_rate": 0.003, + "loss": 4.085, + "step": 3440 + }, + { + "epoch": 0.03441, + "grad_norm": 0.8034314810869642, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 3441 + }, + { + "epoch": 0.03442, + "grad_norm": 0.9713109980816932, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 3442 + }, + { + "epoch": 0.03443, + "grad_norm": 1.2417830843481585, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 3443 + }, + { + "epoch": 0.03444, + "grad_norm": 0.9221286248093234, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 3444 + }, + { + "epoch": 0.03445, + "grad_norm": 0.95846171622353, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 3445 + }, + { + "epoch": 0.03446, + "grad_norm": 1.0528525400170592, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 3446 + }, + { + "epoch": 0.03447, + "grad_norm": 1.0620743717942975, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 3447 + }, + { + "epoch": 0.03448, + "grad_norm": 1.0787590252225165, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 3448 + }, + { + "epoch": 0.03449, + "grad_norm": 1.078404883653542, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 3449 + }, + { + "epoch": 0.0345, + "grad_norm": 1.1540054715604005, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 3450 + }, + { + "epoch": 0.03451, + "grad_norm": 0.8053901369440304, + "learning_rate": 0.003, + "loss": 4.102, + "step": 3451 + }, + { + "epoch": 0.03452, + "grad_norm": 0.8242631664830717, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 3452 + }, + { + "epoch": 0.03453, + "grad_norm": 0.8626897904731716, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 3453 + }, + { + "epoch": 0.03454, + "grad_norm": 0.8237109468250721, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 3454 + }, + { + "epoch": 0.03455, + "grad_norm": 0.804980006239357, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 3455 + }, + { + "epoch": 0.03456, + "grad_norm": 0.729280374546535, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 3456 + }, + { + "epoch": 0.03457, + "grad_norm": 0.7776253504611971, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 3457 + }, + { + "epoch": 0.03458, + "grad_norm": 0.7417498419116878, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 3458 + }, + { + "epoch": 0.03459, + "grad_norm": 0.6919191119122285, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 3459 + }, + { + "epoch": 0.0346, + "grad_norm": 0.6964165299460008, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 3460 + }, + { + "epoch": 0.03461, + "grad_norm": 0.715480299476076, + "learning_rate": 0.003, + "loss": 4.109, + "step": 3461 + }, + { + "epoch": 0.03462, + "grad_norm": 0.7536303091133582, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 3462 + }, + { + "epoch": 0.03463, + "grad_norm": 0.7082893489121429, + "learning_rate": 0.003, + "loss": 4.087, + "step": 3463 + }, + { + "epoch": 0.03464, + "grad_norm": 0.6417430899292896, + "learning_rate": 0.003, + "loss": 4.099, + "step": 3464 + }, + { + "epoch": 0.03465, + "grad_norm": 0.6337345648035622, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 3465 + }, + { + "epoch": 0.03466, + "grad_norm": 0.6853326694383307, + "learning_rate": 0.003, + "loss": 4.107, + "step": 3466 + }, + { + "epoch": 0.03467, + "grad_norm": 0.6969053686300763, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 3467 + }, + { + "epoch": 0.03468, + "grad_norm": 0.7399108221307881, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 3468 + }, + { + "epoch": 0.03469, + "grad_norm": 0.8787869181598735, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 3469 + }, + { + "epoch": 0.0347, + "grad_norm": 1.0101947797716855, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 3470 + }, + { + "epoch": 0.03471, + "grad_norm": 1.1334674964796345, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 3471 + }, + { + "epoch": 0.03472, + "grad_norm": 0.8368294781332597, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 3472 + }, + { + "epoch": 0.03473, + "grad_norm": 0.8757089713498055, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 3473 + }, + { + "epoch": 0.03474, + "grad_norm": 0.8138215418551573, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 3474 + }, + { + "epoch": 0.03475, + "grad_norm": 0.9122035240346946, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 3475 + }, + { + "epoch": 0.03476, + "grad_norm": 1.0107344329113028, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 3476 + }, + { + "epoch": 0.03477, + "grad_norm": 1.0671405774179539, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 3477 + }, + { + "epoch": 0.03478, + "grad_norm": 1.011473839771263, + "learning_rate": 0.003, + "loss": 4.128, + "step": 3478 + }, + { + "epoch": 0.03479, + "grad_norm": 1.0143839143452251, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 3479 + }, + { + "epoch": 0.0348, + "grad_norm": 1.01956672356136, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 3480 + }, + { + "epoch": 0.03481, + "grad_norm": 0.8544300973866115, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 3481 + }, + { + "epoch": 0.03482, + "grad_norm": 1.0322476593810674, + "learning_rate": 0.003, + "loss": 4.133, + "step": 3482 + }, + { + "epoch": 0.03483, + "grad_norm": 1.1510621333444928, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3483 + }, + { + "epoch": 0.03484, + "grad_norm": 0.929657295041821, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 3484 + }, + { + "epoch": 0.03485, + "grad_norm": 0.9553453027677797, + "learning_rate": 0.003, + "loss": 4.146, + "step": 3485 + }, + { + "epoch": 0.03486, + "grad_norm": 0.9940865088692804, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 3486 + }, + { + "epoch": 0.03487, + "grad_norm": 0.9250098589178641, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 3487 + }, + { + "epoch": 0.03488, + "grad_norm": 0.9460673526858474, + "learning_rate": 0.003, + "loss": 4.111, + "step": 3488 + }, + { + "epoch": 0.03489, + "grad_norm": 0.9687600332195162, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 3489 + }, + { + "epoch": 0.0349, + "grad_norm": 1.0784208838544622, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 3490 + }, + { + "epoch": 0.03491, + "grad_norm": 0.9346645940080963, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 3491 + }, + { + "epoch": 0.03492, + "grad_norm": 0.8257240466605443, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 3492 + }, + { + "epoch": 0.03493, + "grad_norm": 0.6651031888705147, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 3493 + }, + { + "epoch": 0.03494, + "grad_norm": 0.6641292093348049, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 3494 + }, + { + "epoch": 0.03495, + "grad_norm": 0.7091189628573692, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 3495 + }, + { + "epoch": 0.03496, + "grad_norm": 0.6860863869815901, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 3496 + }, + { + "epoch": 0.03497, + "grad_norm": 0.6717227669489184, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 3497 + }, + { + "epoch": 0.03498, + "grad_norm": 0.6858788932767997, + "learning_rate": 0.003, + "loss": 4.098, + "step": 3498 + }, + { + "epoch": 0.03499, + "grad_norm": 0.6840439194259218, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 3499 + }, + { + "epoch": 0.035, + "grad_norm": 0.7718701545173198, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 3500 + }, + { + "epoch": 0.03501, + "grad_norm": 0.9467030386692938, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 3501 + }, + { + "epoch": 0.03502, + "grad_norm": 1.1344054626766904, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 3502 + }, + { + "epoch": 0.03503, + "grad_norm": 0.8227576423331606, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 3503 + }, + { + "epoch": 0.03504, + "grad_norm": 0.7907161997976049, + "learning_rate": 0.003, + "loss": 4.103, + "step": 3504 + }, + { + "epoch": 0.03505, + "grad_norm": 0.7949755989311121, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3505 + }, + { + "epoch": 0.03506, + "grad_norm": 0.8242114597731752, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 3506 + }, + { + "epoch": 0.03507, + "grad_norm": 0.827615657879298, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 3507 + }, + { + "epoch": 0.03508, + "grad_norm": 0.8269261240893123, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 3508 + }, + { + "epoch": 0.03509, + "grad_norm": 0.8828534433530547, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 3509 + }, + { + "epoch": 0.0351, + "grad_norm": 1.0949297929244133, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 3510 + }, + { + "epoch": 0.03511, + "grad_norm": 0.8474490719439868, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 3511 + }, + { + "epoch": 0.03512, + "grad_norm": 0.6104860735283346, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 3512 + }, + { + "epoch": 0.03513, + "grad_norm": 0.6050004310996162, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 3513 + }, + { + "epoch": 0.03514, + "grad_norm": 0.6520473843771001, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 3514 + }, + { + "epoch": 0.03515, + "grad_norm": 0.6055869327734182, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 3515 + }, + { + "epoch": 0.03516, + "grad_norm": 0.6086175138721427, + "learning_rate": 0.003, + "loss": 4.136, + "step": 3516 + }, + { + "epoch": 0.03517, + "grad_norm": 0.6697942755212305, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 3517 + }, + { + "epoch": 0.03518, + "grad_norm": 0.7287441549864627, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 3518 + }, + { + "epoch": 0.03519, + "grad_norm": 0.6795543307624995, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 3519 + }, + { + "epoch": 0.0352, + "grad_norm": 0.5583215165075127, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 3520 + }, + { + "epoch": 0.03521, + "grad_norm": 0.6145999630971761, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 3521 + }, + { + "epoch": 0.03522, + "grad_norm": 0.6235299979345595, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 3522 + }, + { + "epoch": 0.03523, + "grad_norm": 0.6982274043250706, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 3523 + }, + { + "epoch": 0.03524, + "grad_norm": 0.8135221205032699, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 3524 + }, + { + "epoch": 0.03525, + "grad_norm": 0.9550932431193345, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 3525 + }, + { + "epoch": 0.03526, + "grad_norm": 0.9769559861681798, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 3526 + }, + { + "epoch": 0.03527, + "grad_norm": 0.9916459776325028, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 3527 + }, + { + "epoch": 0.03528, + "grad_norm": 1.161858614363555, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 3528 + }, + { + "epoch": 0.03529, + "grad_norm": 0.8840893570988297, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 3529 + }, + { + "epoch": 0.0353, + "grad_norm": 0.9045833698319836, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 3530 + }, + { + "epoch": 0.03531, + "grad_norm": 1.0824064864090153, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3531 + }, + { + "epoch": 0.03532, + "grad_norm": 1.04041676924932, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 3532 + }, + { + "epoch": 0.03533, + "grad_norm": 1.0971254119119307, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 3533 + }, + { + "epoch": 0.03534, + "grad_norm": 0.9892749396028874, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3534 + }, + { + "epoch": 0.03535, + "grad_norm": 0.8985518501036314, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 3535 + }, + { + "epoch": 0.03536, + "grad_norm": 0.7993862708923696, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 3536 + }, + { + "epoch": 0.03537, + "grad_norm": 0.8735445800029786, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 3537 + }, + { + "epoch": 0.03538, + "grad_norm": 0.9507196424596396, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 3538 + }, + { + "epoch": 0.03539, + "grad_norm": 0.9320860917835899, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 3539 + }, + { + "epoch": 0.0354, + "grad_norm": 1.0666914195552315, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 3540 + }, + { + "epoch": 0.03541, + "grad_norm": 0.9666553633993341, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 3541 + }, + { + "epoch": 0.03542, + "grad_norm": 1.075546280507871, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 3542 + }, + { + "epoch": 0.03543, + "grad_norm": 0.8581109533371994, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 3543 + }, + { + "epoch": 0.03544, + "grad_norm": 0.8504542868770893, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 3544 + }, + { + "epoch": 0.03545, + "grad_norm": 0.8221740518533669, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 3545 + }, + { + "epoch": 0.03546, + "grad_norm": 0.856238149878472, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 3546 + }, + { + "epoch": 0.03547, + "grad_norm": 0.8136205837732846, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3547 + }, + { + "epoch": 0.03548, + "grad_norm": 0.9120415275886503, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3548 + }, + { + "epoch": 0.03549, + "grad_norm": 0.8741924577842357, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 3549 + }, + { + "epoch": 0.0355, + "grad_norm": 0.964690101454288, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 3550 + }, + { + "epoch": 0.03551, + "grad_norm": 1.1078408326241436, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 3551 + }, + { + "epoch": 0.03552, + "grad_norm": 0.8276304700566389, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 3552 + }, + { + "epoch": 0.03553, + "grad_norm": 0.6988104369223903, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 3553 + }, + { + "epoch": 0.03554, + "grad_norm": 0.8099473741524069, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3554 + }, + { + "epoch": 0.03555, + "grad_norm": 0.7657626977957569, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 3555 + }, + { + "epoch": 0.03556, + "grad_norm": 0.6884161752719029, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 3556 + }, + { + "epoch": 0.03557, + "grad_norm": 0.6257341319138348, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 3557 + }, + { + "epoch": 0.03558, + "grad_norm": 0.6075327774193832, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 3558 + }, + { + "epoch": 0.03559, + "grad_norm": 0.7848274408146922, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 3559 + }, + { + "epoch": 0.0356, + "grad_norm": 1.0374578386942082, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 3560 + }, + { + "epoch": 0.03561, + "grad_norm": 1.0758010083557943, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 3561 + }, + { + "epoch": 0.03562, + "grad_norm": 0.7322873944575906, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 3562 + }, + { + "epoch": 0.03563, + "grad_norm": 0.7971284325210378, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 3563 + }, + { + "epoch": 0.03564, + "grad_norm": 1.0246051569009555, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 3564 + }, + { + "epoch": 0.03565, + "grad_norm": 1.2267527206427153, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 3565 + }, + { + "epoch": 0.03566, + "grad_norm": 0.8984275051065349, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 3566 + }, + { + "epoch": 0.03567, + "grad_norm": 0.9814381465438691, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3567 + }, + { + "epoch": 0.03568, + "grad_norm": 0.862338747629593, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 3568 + }, + { + "epoch": 0.03569, + "grad_norm": 0.8507776080373626, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3569 + }, + { + "epoch": 0.0357, + "grad_norm": 0.8450119567501945, + "learning_rate": 0.003, + "loss": 4.108, + "step": 3570 + }, + { + "epoch": 0.03571, + "grad_norm": 0.8474288365898862, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 3571 + }, + { + "epoch": 0.03572, + "grad_norm": 0.8860990170054822, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 3572 + }, + { + "epoch": 0.03573, + "grad_norm": 0.9378112912832053, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 3573 + }, + { + "epoch": 0.03574, + "grad_norm": 0.7793011201012979, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 3574 + }, + { + "epoch": 0.03575, + "grad_norm": 0.6347551630020188, + "learning_rate": 0.003, + "loss": 4.099, + "step": 3575 + }, + { + "epoch": 0.03576, + "grad_norm": 0.6273033455713914, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 3576 + }, + { + "epoch": 0.03577, + "grad_norm": 0.5948625546551696, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 3577 + }, + { + "epoch": 0.03578, + "grad_norm": 0.5805191268794814, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3578 + }, + { + "epoch": 0.03579, + "grad_norm": 0.5381920460033986, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 3579 + }, + { + "epoch": 0.0358, + "grad_norm": 0.7081167212858254, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 3580 + }, + { + "epoch": 0.03581, + "grad_norm": 0.7386340619770309, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 3581 + }, + { + "epoch": 0.03582, + "grad_norm": 0.7587714591488295, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 3582 + }, + { + "epoch": 0.03583, + "grad_norm": 0.7113793632761408, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 3583 + }, + { + "epoch": 0.03584, + "grad_norm": 0.7188054852978065, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 3584 + }, + { + "epoch": 0.03585, + "grad_norm": 0.8853261674987007, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 3585 + }, + { + "epoch": 0.03586, + "grad_norm": 0.92627880930667, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 3586 + }, + { + "epoch": 0.03587, + "grad_norm": 0.9331834952096867, + "learning_rate": 0.003, + "loss": 4.105, + "step": 3587 + }, + { + "epoch": 0.03588, + "grad_norm": 1.056355930051097, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 3588 + }, + { + "epoch": 0.03589, + "grad_norm": 0.8554656185849556, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 3589 + }, + { + "epoch": 0.0359, + "grad_norm": 0.7792232208377492, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 3590 + }, + { + "epoch": 0.03591, + "grad_norm": 0.9342912310042636, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 3591 + }, + { + "epoch": 0.03592, + "grad_norm": 1.0503229079193128, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 3592 + }, + { + "epoch": 0.03593, + "grad_norm": 0.984608248076784, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 3593 + }, + { + "epoch": 0.03594, + "grad_norm": 0.9992015208659966, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 3594 + }, + { + "epoch": 0.03595, + "grad_norm": 0.9486124739849024, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 3595 + }, + { + "epoch": 0.03596, + "grad_norm": 0.8044450679529417, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 3596 + }, + { + "epoch": 0.03597, + "grad_norm": 0.7529472117586564, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 3597 + }, + { + "epoch": 0.03598, + "grad_norm": 0.8016519620578548, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 3598 + }, + { + "epoch": 0.03599, + "grad_norm": 0.8552258797660491, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 3599 + }, + { + "epoch": 0.036, + "grad_norm": 1.122743201613801, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 3600 + }, + { + "epoch": 0.03601, + "grad_norm": 0.9558780975359691, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 3601 + }, + { + "epoch": 0.03602, + "grad_norm": 1.0399279121526195, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 3602 + }, + { + "epoch": 0.03603, + "grad_norm": 1.1551062763334126, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 3603 + }, + { + "epoch": 0.03604, + "grad_norm": 1.0548506296076239, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 3604 + }, + { + "epoch": 0.03605, + "grad_norm": 1.001983402921064, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 3605 + }, + { + "epoch": 0.03606, + "grad_norm": 1.0312319108923618, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 3606 + }, + { + "epoch": 0.03607, + "grad_norm": 0.8219603208628181, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 3607 + }, + { + "epoch": 0.03608, + "grad_norm": 0.6498677677486123, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 3608 + }, + { + "epoch": 0.03609, + "grad_norm": 0.7278852617397139, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 3609 + }, + { + "epoch": 0.0361, + "grad_norm": 0.6666501580008998, + "learning_rate": 0.003, + "loss": 4.126, + "step": 3610 + }, + { + "epoch": 0.03611, + "grad_norm": 0.6972920943122692, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 3611 + }, + { + "epoch": 0.03612, + "grad_norm": 0.8293279199659314, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 3612 + }, + { + "epoch": 0.03613, + "grad_norm": 0.8280160598898497, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 3613 + }, + { + "epoch": 0.03614, + "grad_norm": 0.8586511775157879, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 3614 + }, + { + "epoch": 0.03615, + "grad_norm": 0.8273149005356012, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 3615 + }, + { + "epoch": 0.03616, + "grad_norm": 0.7395998497715127, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 3616 + }, + { + "epoch": 0.03617, + "grad_norm": 0.6904839766505554, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 3617 + }, + { + "epoch": 0.03618, + "grad_norm": 0.7602021438934125, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 3618 + }, + { + "epoch": 0.03619, + "grad_norm": 0.9970287144190367, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 3619 + }, + { + "epoch": 0.0362, + "grad_norm": 1.0227512354316275, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 3620 + }, + { + "epoch": 0.03621, + "grad_norm": 0.8480434130474432, + "learning_rate": 0.003, + "loss": 4.112, + "step": 3621 + }, + { + "epoch": 0.03622, + "grad_norm": 0.9271509010196743, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 3622 + }, + { + "epoch": 0.03623, + "grad_norm": 1.022948927798352, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 3623 + }, + { + "epoch": 0.03624, + "grad_norm": 0.9265641312425992, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 3624 + }, + { + "epoch": 0.03625, + "grad_norm": 0.7813804518175715, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3625 + }, + { + "epoch": 0.03626, + "grad_norm": 0.7288040056584648, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 3626 + }, + { + "epoch": 0.03627, + "grad_norm": 0.7898367328534155, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 3627 + }, + { + "epoch": 0.03628, + "grad_norm": 0.7391500282099692, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 3628 + }, + { + "epoch": 0.03629, + "grad_norm": 0.7244163565447427, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 3629 + }, + { + "epoch": 0.0363, + "grad_norm": 0.8351668360599707, + "learning_rate": 0.003, + "loss": 4.108, + "step": 3630 + }, + { + "epoch": 0.03631, + "grad_norm": 1.1144221832925818, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 3631 + }, + { + "epoch": 0.03632, + "grad_norm": 0.9360712274389132, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 3632 + }, + { + "epoch": 0.03633, + "grad_norm": 0.9876936083254465, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 3633 + }, + { + "epoch": 0.03634, + "grad_norm": 0.955681965536761, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 3634 + }, + { + "epoch": 0.03635, + "grad_norm": 0.9023651977998233, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 3635 + }, + { + "epoch": 0.03636, + "grad_norm": 0.8274154329119018, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 3636 + }, + { + "epoch": 0.03637, + "grad_norm": 0.9898592340376765, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 3637 + }, + { + "epoch": 0.03638, + "grad_norm": 0.961405507097599, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 3638 + }, + { + "epoch": 0.03639, + "grad_norm": 1.0204413055029042, + "learning_rate": 0.003, + "loss": 4.145, + "step": 3639 + }, + { + "epoch": 0.0364, + "grad_norm": 0.9674075028743515, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 3640 + }, + { + "epoch": 0.03641, + "grad_norm": 1.0768346328712466, + "learning_rate": 0.003, + "loss": 4.125, + "step": 3641 + }, + { + "epoch": 0.03642, + "grad_norm": 0.9224514479840723, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 3642 + }, + { + "epoch": 0.03643, + "grad_norm": 1.0336265072437074, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 3643 + }, + { + "epoch": 0.03644, + "grad_norm": 0.8524724447399462, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 3644 + }, + { + "epoch": 0.03645, + "grad_norm": 0.746873753702558, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 3645 + }, + { + "epoch": 0.03646, + "grad_norm": 0.8800991852578832, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3646 + }, + { + "epoch": 0.03647, + "grad_norm": 0.9340999264767859, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 3647 + }, + { + "epoch": 0.03648, + "grad_norm": 0.9312651206865153, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 3648 + }, + { + "epoch": 0.03649, + "grad_norm": 0.9165122392124831, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 3649 + }, + { + "epoch": 0.0365, + "grad_norm": 0.956526663147831, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 3650 + }, + { + "epoch": 0.03651, + "grad_norm": 0.8484726129203977, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 3651 + }, + { + "epoch": 0.03652, + "grad_norm": 0.7894465078154086, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 3652 + }, + { + "epoch": 0.03653, + "grad_norm": 0.6892303382737016, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 3653 + }, + { + "epoch": 0.03654, + "grad_norm": 0.6682166624701149, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 3654 + }, + { + "epoch": 0.03655, + "grad_norm": 0.6644222696456153, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 3655 + }, + { + "epoch": 0.03656, + "grad_norm": 0.6972374084601054, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 3656 + }, + { + "epoch": 0.03657, + "grad_norm": 0.6757860783494947, + "learning_rate": 0.003, + "loss": 4.071, + "step": 3657 + }, + { + "epoch": 0.03658, + "grad_norm": 0.7083328895452327, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 3658 + }, + { + "epoch": 0.03659, + "grad_norm": 0.7718946593018728, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 3659 + }, + { + "epoch": 0.0366, + "grad_norm": 0.7376732318109632, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 3660 + }, + { + "epoch": 0.03661, + "grad_norm": 0.8048594254747962, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 3661 + }, + { + "epoch": 0.03662, + "grad_norm": 0.9041684527522658, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 3662 + }, + { + "epoch": 0.03663, + "grad_norm": 0.8857606200128569, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 3663 + }, + { + "epoch": 0.03664, + "grad_norm": 0.9256759603717064, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 3664 + }, + { + "epoch": 0.03665, + "grad_norm": 0.9238563202936576, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 3665 + }, + { + "epoch": 0.03666, + "grad_norm": 0.8203365130003366, + "learning_rate": 0.003, + "loss": 4.074, + "step": 3666 + }, + { + "epoch": 0.03667, + "grad_norm": 0.9024471801382937, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 3667 + }, + { + "epoch": 0.03668, + "grad_norm": 1.0552967059744773, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 3668 + }, + { + "epoch": 0.03669, + "grad_norm": 1.0820390277779695, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 3669 + }, + { + "epoch": 0.0367, + "grad_norm": 0.9454674065362177, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3670 + }, + { + "epoch": 0.03671, + "grad_norm": 1.0059330853811892, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 3671 + }, + { + "epoch": 0.03672, + "grad_norm": 1.2723317844660584, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 3672 + }, + { + "epoch": 0.03673, + "grad_norm": 0.8050766891971283, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 3673 + }, + { + "epoch": 0.03674, + "grad_norm": 0.8682507056177615, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 3674 + }, + { + "epoch": 0.03675, + "grad_norm": 1.0276133687150726, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 3675 + }, + { + "epoch": 0.03676, + "grad_norm": 1.0466449018255517, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 3676 + }, + { + "epoch": 0.03677, + "grad_norm": 0.9830229711282336, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 3677 + }, + { + "epoch": 0.03678, + "grad_norm": 0.9217069489116423, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3678 + }, + { + "epoch": 0.03679, + "grad_norm": 1.0822177384710947, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 3679 + }, + { + "epoch": 0.0368, + "grad_norm": 1.0907486819128838, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 3680 + }, + { + "epoch": 0.03681, + "grad_norm": 0.8227793104650016, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 3681 + }, + { + "epoch": 0.03682, + "grad_norm": 0.7808126016328544, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3682 + }, + { + "epoch": 0.03683, + "grad_norm": 0.7174400556511353, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 3683 + }, + { + "epoch": 0.03684, + "grad_norm": 0.6775479041332162, + "learning_rate": 0.003, + "loss": 4.104, + "step": 3684 + }, + { + "epoch": 0.03685, + "grad_norm": 0.6420088354614908, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 3685 + }, + { + "epoch": 0.03686, + "grad_norm": 0.598867465128322, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 3686 + }, + { + "epoch": 0.03687, + "grad_norm": 0.6115731068077563, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 3687 + }, + { + "epoch": 0.03688, + "grad_norm": 0.7067511764658732, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 3688 + }, + { + "epoch": 0.03689, + "grad_norm": 0.7020192080483855, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 3689 + }, + { + "epoch": 0.0369, + "grad_norm": 0.6267089096242754, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 3690 + }, + { + "epoch": 0.03691, + "grad_norm": 0.5456674081937926, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 3691 + }, + { + "epoch": 0.03692, + "grad_norm": 0.6082919076452641, + "learning_rate": 0.003, + "loss": 4.127, + "step": 3692 + }, + { + "epoch": 0.03693, + "grad_norm": 0.7939340117807796, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 3693 + }, + { + "epoch": 0.03694, + "grad_norm": 1.0013591289075594, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3694 + }, + { + "epoch": 0.03695, + "grad_norm": 1.022061020358787, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 3695 + }, + { + "epoch": 0.03696, + "grad_norm": 0.6767445565263814, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 3696 + }, + { + "epoch": 0.03697, + "grad_norm": 0.6543502581869196, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 3697 + }, + { + "epoch": 0.03698, + "grad_norm": 0.898978343569856, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 3698 + }, + { + "epoch": 0.03699, + "grad_norm": 0.8905816524859695, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 3699 + }, + { + "epoch": 0.037, + "grad_norm": 0.7721797452363818, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 3700 + }, + { + "epoch": 0.03701, + "grad_norm": 0.9150903630097929, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 3701 + }, + { + "epoch": 0.03702, + "grad_norm": 0.846227592035782, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 3702 + }, + { + "epoch": 0.03703, + "grad_norm": 0.797196639545616, + "learning_rate": 0.003, + "loss": 4.105, + "step": 3703 + }, + { + "epoch": 0.03704, + "grad_norm": 0.9810543408588123, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 3704 + }, + { + "epoch": 0.03705, + "grad_norm": 1.0722834307776175, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 3705 + }, + { + "epoch": 0.03706, + "grad_norm": 0.9721358338604488, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3706 + }, + { + "epoch": 0.03707, + "grad_norm": 0.9790738664900162, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 3707 + }, + { + "epoch": 0.03708, + "grad_norm": 0.9353887990568662, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 3708 + }, + { + "epoch": 0.03709, + "grad_norm": 0.9856206989122267, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 3709 + }, + { + "epoch": 0.0371, + "grad_norm": 1.0451787463045008, + "learning_rate": 0.003, + "loss": 4.138, + "step": 3710 + }, + { + "epoch": 0.03711, + "grad_norm": 1.0753305901256527, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 3711 + }, + { + "epoch": 0.03712, + "grad_norm": 0.9432866794357371, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 3712 + }, + { + "epoch": 0.03713, + "grad_norm": 1.0228561402780971, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 3713 + }, + { + "epoch": 0.03714, + "grad_norm": 1.0155399617885337, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 3714 + }, + { + "epoch": 0.03715, + "grad_norm": 0.9501309057637405, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 3715 + }, + { + "epoch": 0.03716, + "grad_norm": 0.8368341993061407, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 3716 + }, + { + "epoch": 0.03717, + "grad_norm": 0.8200844350513596, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 3717 + }, + { + "epoch": 0.03718, + "grad_norm": 0.8192379479874041, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 3718 + }, + { + "epoch": 0.03719, + "grad_norm": 0.9007315714309677, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 3719 + }, + { + "epoch": 0.0372, + "grad_norm": 0.9179074609676963, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 3720 + }, + { + "epoch": 0.03721, + "grad_norm": 0.848049967304013, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 3721 + }, + { + "epoch": 0.03722, + "grad_norm": 0.7009453544906782, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 3722 + }, + { + "epoch": 0.03723, + "grad_norm": 0.6973322955990157, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 3723 + }, + { + "epoch": 0.03724, + "grad_norm": 0.6858684932844772, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 3724 + }, + { + "epoch": 0.03725, + "grad_norm": 0.5759232117551641, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 3725 + }, + { + "epoch": 0.03726, + "grad_norm": 0.7355358643002436, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 3726 + }, + { + "epoch": 0.03727, + "grad_norm": 0.975453002086878, + "learning_rate": 0.003, + "loss": 4.128, + "step": 3727 + }, + { + "epoch": 0.03728, + "grad_norm": 1.2871266145714035, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 3728 + }, + { + "epoch": 0.03729, + "grad_norm": 0.6800175356883061, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 3729 + }, + { + "epoch": 0.0373, + "grad_norm": 0.8995970728462417, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 3730 + }, + { + "epoch": 0.03731, + "grad_norm": 0.8748991092762279, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 3731 + }, + { + "epoch": 0.03732, + "grad_norm": 0.7674871139345869, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 3732 + }, + { + "epoch": 0.03733, + "grad_norm": 0.7725141224382513, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 3733 + }, + { + "epoch": 0.03734, + "grad_norm": 0.923649436345944, + "learning_rate": 0.003, + "loss": 4.101, + "step": 3734 + }, + { + "epoch": 0.03735, + "grad_norm": 0.85452781445576, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 3735 + }, + { + "epoch": 0.03736, + "grad_norm": 0.8691053620191994, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 3736 + }, + { + "epoch": 0.03737, + "grad_norm": 0.8505997578362463, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 3737 + }, + { + "epoch": 0.03738, + "grad_norm": 0.7742946039499282, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 3738 + }, + { + "epoch": 0.03739, + "grad_norm": 0.7994977680328913, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 3739 + }, + { + "epoch": 0.0374, + "grad_norm": 0.8536215986910768, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 3740 + }, + { + "epoch": 0.03741, + "grad_norm": 0.9409826319598602, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 3741 + }, + { + "epoch": 0.03742, + "grad_norm": 0.8593777917195926, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 3742 + }, + { + "epoch": 0.03743, + "grad_norm": 0.7563917031775911, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 3743 + }, + { + "epoch": 0.03744, + "grad_norm": 0.7103064287871557, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 3744 + }, + { + "epoch": 0.03745, + "grad_norm": 0.7615538447680247, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 3745 + }, + { + "epoch": 0.03746, + "grad_norm": 0.9955801956257352, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 3746 + }, + { + "epoch": 0.03747, + "grad_norm": 1.0171288172949948, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 3747 + }, + { + "epoch": 0.03748, + "grad_norm": 0.7896231689985856, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 3748 + }, + { + "epoch": 0.03749, + "grad_norm": 0.6375567183654871, + "learning_rate": 0.003, + "loss": 4.095, + "step": 3749 + }, + { + "epoch": 0.0375, + "grad_norm": 0.7372117367719312, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3750 + }, + { + "epoch": 0.03751, + "grad_norm": 0.7461639244591386, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 3751 + }, + { + "epoch": 0.03752, + "grad_norm": 0.7827559728941343, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 3752 + }, + { + "epoch": 0.03753, + "grad_norm": 0.7771731000337748, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 3753 + }, + { + "epoch": 0.03754, + "grad_norm": 1.0118496549896825, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3754 + }, + { + "epoch": 0.03755, + "grad_norm": 1.2993270154749912, + "learning_rate": 0.003, + "loss": 4.124, + "step": 3755 + }, + { + "epoch": 0.03756, + "grad_norm": 0.8028019558848566, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 3756 + }, + { + "epoch": 0.03757, + "grad_norm": 0.9071996665258343, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 3757 + }, + { + "epoch": 0.03758, + "grad_norm": 0.8773892948854304, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 3758 + }, + { + "epoch": 0.03759, + "grad_norm": 0.8198221769200625, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 3759 + }, + { + "epoch": 0.0376, + "grad_norm": 0.8540457248460523, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 3760 + }, + { + "epoch": 0.03761, + "grad_norm": 0.7667243415560708, + "learning_rate": 0.003, + "loss": 4.071, + "step": 3761 + }, + { + "epoch": 0.03762, + "grad_norm": 0.8865050328861852, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 3762 + }, + { + "epoch": 0.03763, + "grad_norm": 0.9591731803168125, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 3763 + }, + { + "epoch": 0.03764, + "grad_norm": 0.9480606119145719, + "learning_rate": 0.003, + "loss": 4.125, + "step": 3764 + }, + { + "epoch": 0.03765, + "grad_norm": 0.9684597356680353, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 3765 + }, + { + "epoch": 0.03766, + "grad_norm": 0.8344731198089695, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 3766 + }, + { + "epoch": 0.03767, + "grad_norm": 0.7765792550643911, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 3767 + }, + { + "epoch": 0.03768, + "grad_norm": 0.7642334739357078, + "learning_rate": 0.003, + "loss": 4.098, + "step": 3768 + }, + { + "epoch": 0.03769, + "grad_norm": 0.7211211499394363, + "learning_rate": 0.003, + "loss": 4.089, + "step": 3769 + }, + { + "epoch": 0.0377, + "grad_norm": 0.6106973291196678, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 3770 + }, + { + "epoch": 0.03771, + "grad_norm": 0.5628256340241241, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 3771 + }, + { + "epoch": 0.03772, + "grad_norm": 0.5845993807831852, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 3772 + }, + { + "epoch": 0.03773, + "grad_norm": 0.6697980766862262, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 3773 + }, + { + "epoch": 0.03774, + "grad_norm": 0.8252901528282155, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 3774 + }, + { + "epoch": 0.03775, + "grad_norm": 0.9517527171183254, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 3775 + }, + { + "epoch": 0.03776, + "grad_norm": 1.0897263182368793, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 3776 + }, + { + "epoch": 0.03777, + "grad_norm": 0.8653535869810033, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 3777 + }, + { + "epoch": 0.03778, + "grad_norm": 0.7342155379460369, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 3778 + }, + { + "epoch": 0.03779, + "grad_norm": 0.7679190695396096, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 3779 + }, + { + "epoch": 0.0378, + "grad_norm": 0.8679609457918785, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 3780 + }, + { + "epoch": 0.03781, + "grad_norm": 1.0049631112041457, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 3781 + }, + { + "epoch": 0.03782, + "grad_norm": 1.092459674146767, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 3782 + }, + { + "epoch": 0.03783, + "grad_norm": 0.9622890806365146, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 3783 + }, + { + "epoch": 0.03784, + "grad_norm": 1.0671033317728673, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 3784 + }, + { + "epoch": 0.03785, + "grad_norm": 1.0407807642051592, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 3785 + }, + { + "epoch": 0.03786, + "grad_norm": 1.125690448525855, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 3786 + }, + { + "epoch": 0.03787, + "grad_norm": 1.0253559423721736, + "learning_rate": 0.003, + "loss": 4.108, + "step": 3787 + }, + { + "epoch": 0.03788, + "grad_norm": 1.0136069570600932, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 3788 + }, + { + "epoch": 0.03789, + "grad_norm": 1.1118575565083433, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 3789 + }, + { + "epoch": 0.0379, + "grad_norm": 0.7950650221590647, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 3790 + }, + { + "epoch": 0.03791, + "grad_norm": 0.6642703034618491, + "learning_rate": 0.003, + "loss": 4.102, + "step": 3791 + }, + { + "epoch": 0.03792, + "grad_norm": 0.6118532884669681, + "learning_rate": 0.003, + "loss": 4.131, + "step": 3792 + }, + { + "epoch": 0.03793, + "grad_norm": 0.6538749880963972, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 3793 + }, + { + "epoch": 0.03794, + "grad_norm": 0.7185276449261756, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 3794 + }, + { + "epoch": 0.03795, + "grad_norm": 0.8476725115524525, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 3795 + }, + { + "epoch": 0.03796, + "grad_norm": 0.912186760325603, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 3796 + }, + { + "epoch": 0.03797, + "grad_norm": 0.8535289933328677, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 3797 + }, + { + "epoch": 0.03798, + "grad_norm": 0.8439916131790419, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 3798 + }, + { + "epoch": 0.03799, + "grad_norm": 0.977315385277912, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 3799 + }, + { + "epoch": 0.038, + "grad_norm": 1.2434381910310124, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 3800 + }, + { + "epoch": 0.03801, + "grad_norm": 0.8428497034893558, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 3801 + }, + { + "epoch": 0.03802, + "grad_norm": 0.7525631044925924, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 3802 + }, + { + "epoch": 0.03803, + "grad_norm": 0.8068378132076305, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 3803 + }, + { + "epoch": 0.03804, + "grad_norm": 0.8984500712876651, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 3804 + }, + { + "epoch": 0.03805, + "grad_norm": 0.9405144408554322, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 3805 + }, + { + "epoch": 0.03806, + "grad_norm": 1.019634044722222, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 3806 + }, + { + "epoch": 0.03807, + "grad_norm": 0.8853714667588649, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 3807 + }, + { + "epoch": 0.03808, + "grad_norm": 0.9597650598957446, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 3808 + }, + { + "epoch": 0.03809, + "grad_norm": 1.067683764868459, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 3809 + }, + { + "epoch": 0.0381, + "grad_norm": 0.9853327608719147, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 3810 + }, + { + "epoch": 0.03811, + "grad_norm": 1.1386409326086608, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 3811 + }, + { + "epoch": 0.03812, + "grad_norm": 0.8785836941800041, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 3812 + }, + { + "epoch": 0.03813, + "grad_norm": 0.719604168428238, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 3813 + }, + { + "epoch": 0.03814, + "grad_norm": 0.8437531723051056, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 3814 + }, + { + "epoch": 0.03815, + "grad_norm": 0.9017994260619784, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 3815 + }, + { + "epoch": 0.03816, + "grad_norm": 1.0098474951212937, + "learning_rate": 0.003, + "loss": 4.107, + "step": 3816 + }, + { + "epoch": 0.03817, + "grad_norm": 0.9361762835503497, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 3817 + }, + { + "epoch": 0.03818, + "grad_norm": 1.037066830030048, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 3818 + }, + { + "epoch": 0.03819, + "grad_norm": 1.156007843641428, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 3819 + }, + { + "epoch": 0.0382, + "grad_norm": 0.9923419777257736, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 3820 + }, + { + "epoch": 0.03821, + "grad_norm": 1.1885176826442838, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 3821 + }, + { + "epoch": 0.03822, + "grad_norm": 0.8699977531184279, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 3822 + }, + { + "epoch": 0.03823, + "grad_norm": 0.9310049147972063, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 3823 + }, + { + "epoch": 0.03824, + "grad_norm": 1.0606557350696768, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 3824 + }, + { + "epoch": 0.03825, + "grad_norm": 0.8761874314983482, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3825 + }, + { + "epoch": 0.03826, + "grad_norm": 0.709566654404231, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 3826 + }, + { + "epoch": 0.03827, + "grad_norm": 0.665621012314629, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 3827 + }, + { + "epoch": 0.03828, + "grad_norm": 0.6539927553498923, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 3828 + }, + { + "epoch": 0.03829, + "grad_norm": 0.6659373242823122, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 3829 + }, + { + "epoch": 0.0383, + "grad_norm": 0.719614104181746, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 3830 + }, + { + "epoch": 0.03831, + "grad_norm": 0.7380247193128887, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 3831 + }, + { + "epoch": 0.03832, + "grad_norm": 0.7602873975407207, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 3832 + }, + { + "epoch": 0.03833, + "grad_norm": 0.7801412356291381, + "learning_rate": 0.003, + "loss": 4.091, + "step": 3833 + }, + { + "epoch": 0.03834, + "grad_norm": 1.035527155697683, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 3834 + }, + { + "epoch": 0.03835, + "grad_norm": 1.2017267869538604, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 3835 + }, + { + "epoch": 0.03836, + "grad_norm": 0.7780948979729766, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 3836 + }, + { + "epoch": 0.03837, + "grad_norm": 0.6833191218840556, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 3837 + }, + { + "epoch": 0.03838, + "grad_norm": 0.8599951602639524, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 3838 + }, + { + "epoch": 0.03839, + "grad_norm": 0.9833951212613625, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 3839 + }, + { + "epoch": 0.0384, + "grad_norm": 1.1478082520006916, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 3840 + }, + { + "epoch": 0.03841, + "grad_norm": 0.8829906180137885, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 3841 + }, + { + "epoch": 0.03842, + "grad_norm": 0.7834869010490613, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 3842 + }, + { + "epoch": 0.03843, + "grad_norm": 0.7188713056155882, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 3843 + }, + { + "epoch": 0.03844, + "grad_norm": 0.7540179942394175, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 3844 + }, + { + "epoch": 0.03845, + "grad_norm": 0.7463415722264508, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 3845 + }, + { + "epoch": 0.03846, + "grad_norm": 0.7138861193622222, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 3846 + }, + { + "epoch": 0.03847, + "grad_norm": 0.648731470870398, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 3847 + }, + { + "epoch": 0.03848, + "grad_norm": 0.7547795493913247, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 3848 + }, + { + "epoch": 0.03849, + "grad_norm": 0.799301070310209, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 3849 + }, + { + "epoch": 0.0385, + "grad_norm": 0.9203078148453605, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 3850 + }, + { + "epoch": 0.03851, + "grad_norm": 1.0151353622998731, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 3851 + }, + { + "epoch": 0.03852, + "grad_norm": 1.0144010495517577, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 3852 + }, + { + "epoch": 0.03853, + "grad_norm": 1.0493705006427887, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 3853 + }, + { + "epoch": 0.03854, + "grad_norm": 0.8074337792156867, + "learning_rate": 0.003, + "loss": 4.103, + "step": 3854 + }, + { + "epoch": 0.03855, + "grad_norm": 0.8026707872705597, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 3855 + }, + { + "epoch": 0.03856, + "grad_norm": 0.7804974920617295, + "learning_rate": 0.003, + "loss": 4.091, + "step": 3856 + }, + { + "epoch": 0.03857, + "grad_norm": 0.8054893207885017, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 3857 + }, + { + "epoch": 0.03858, + "grad_norm": 0.8546443108408268, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 3858 + }, + { + "epoch": 0.03859, + "grad_norm": 0.8455331814308655, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 3859 + }, + { + "epoch": 0.0386, + "grad_norm": 0.7390789874576452, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 3860 + }, + { + "epoch": 0.03861, + "grad_norm": 0.7582288889706922, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 3861 + }, + { + "epoch": 0.03862, + "grad_norm": 0.725403944101907, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 3862 + }, + { + "epoch": 0.03863, + "grad_norm": 0.8633932250780971, + "learning_rate": 0.003, + "loss": 4.09, + "step": 3863 + }, + { + "epoch": 0.03864, + "grad_norm": 1.1650833586712557, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3864 + }, + { + "epoch": 0.03865, + "grad_norm": 0.9893693208520536, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3865 + }, + { + "epoch": 0.03866, + "grad_norm": 1.0874614460090002, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 3866 + }, + { + "epoch": 0.03867, + "grad_norm": 0.7886630657768378, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 3867 + }, + { + "epoch": 0.03868, + "grad_norm": 0.885679172374958, + "learning_rate": 0.003, + "loss": 4.108, + "step": 3868 + }, + { + "epoch": 0.03869, + "grad_norm": 1.0141213895298733, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 3869 + }, + { + "epoch": 0.0387, + "grad_norm": 1.0688565646198265, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 3870 + }, + { + "epoch": 0.03871, + "grad_norm": 0.9129625883332004, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 3871 + }, + { + "epoch": 0.03872, + "grad_norm": 0.8736530068321784, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 3872 + }, + { + "epoch": 0.03873, + "grad_norm": 0.9127445482435979, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 3873 + }, + { + "epoch": 0.03874, + "grad_norm": 1.1125648535743125, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 3874 + }, + { + "epoch": 0.03875, + "grad_norm": 1.0031183839809612, + "learning_rate": 0.003, + "loss": 4.1, + "step": 3875 + }, + { + "epoch": 0.03876, + "grad_norm": 0.9060582464048182, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 3876 + }, + { + "epoch": 0.03877, + "grad_norm": 0.8392586887613771, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 3877 + }, + { + "epoch": 0.03878, + "grad_norm": 0.8001456111460843, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 3878 + }, + { + "epoch": 0.03879, + "grad_norm": 0.7457591238961055, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 3879 + }, + { + "epoch": 0.0388, + "grad_norm": 0.6124933426277533, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 3880 + }, + { + "epoch": 0.03881, + "grad_norm": 0.6131747582366786, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 3881 + }, + { + "epoch": 0.03882, + "grad_norm": 0.6113752675559317, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 3882 + }, + { + "epoch": 0.03883, + "grad_norm": 0.6919250963347912, + "learning_rate": 0.003, + "loss": 4.082, + "step": 3883 + }, + { + "epoch": 0.03884, + "grad_norm": 0.7426309370638033, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 3884 + }, + { + "epoch": 0.03885, + "grad_norm": 0.9915545301535461, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 3885 + }, + { + "epoch": 0.03886, + "grad_norm": 1.1987560180799988, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 3886 + }, + { + "epoch": 0.03887, + "grad_norm": 0.8162692852302313, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 3887 + }, + { + "epoch": 0.03888, + "grad_norm": 1.0025366149196155, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 3888 + }, + { + "epoch": 0.03889, + "grad_norm": 1.146222148985757, + "learning_rate": 0.003, + "loss": 4.108, + "step": 3889 + }, + { + "epoch": 0.0389, + "grad_norm": 0.8230126221175265, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 3890 + }, + { + "epoch": 0.03891, + "grad_norm": 0.8705081765994817, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 3891 + }, + { + "epoch": 0.03892, + "grad_norm": 0.7747593450381118, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 3892 + }, + { + "epoch": 0.03893, + "grad_norm": 0.7117098416936728, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 3893 + }, + { + "epoch": 0.03894, + "grad_norm": 0.7380911807082656, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 3894 + }, + { + "epoch": 0.03895, + "grad_norm": 0.9497127876188006, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 3895 + }, + { + "epoch": 0.03896, + "grad_norm": 1.1241836801249476, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 3896 + }, + { + "epoch": 0.03897, + "grad_norm": 0.9331139773236928, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 3897 + }, + { + "epoch": 0.03898, + "grad_norm": 0.8717563333320789, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 3898 + }, + { + "epoch": 0.03899, + "grad_norm": 0.9575796876780424, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 3899 + }, + { + "epoch": 0.039, + "grad_norm": 1.1111071654055427, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 3900 + }, + { + "epoch": 0.03901, + "grad_norm": 0.9101352152485117, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 3901 + }, + { + "epoch": 0.03902, + "grad_norm": 0.9149491392794752, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 3902 + }, + { + "epoch": 0.03903, + "grad_norm": 0.9599549632230119, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 3903 + }, + { + "epoch": 0.03904, + "grad_norm": 0.9977595444316653, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 3904 + }, + { + "epoch": 0.03905, + "grad_norm": 0.9772682724535922, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 3905 + }, + { + "epoch": 0.03906, + "grad_norm": 1.116840612654856, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 3906 + }, + { + "epoch": 0.03907, + "grad_norm": 0.9927827399452427, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 3907 + }, + { + "epoch": 0.03908, + "grad_norm": 1.0513974248161086, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 3908 + }, + { + "epoch": 0.03909, + "grad_norm": 0.951299576586947, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 3909 + }, + { + "epoch": 0.0391, + "grad_norm": 0.9273375029445102, + "learning_rate": 0.003, + "loss": 4.132, + "step": 3910 + }, + { + "epoch": 0.03911, + "grad_norm": 0.960086352721459, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 3911 + }, + { + "epoch": 0.03912, + "grad_norm": 0.9317199647764666, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3912 + }, + { + "epoch": 0.03913, + "grad_norm": 0.8946789663336782, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 3913 + }, + { + "epoch": 0.03914, + "grad_norm": 0.8487878737890102, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 3914 + }, + { + "epoch": 0.03915, + "grad_norm": 0.8668410486160162, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 3915 + }, + { + "epoch": 0.03916, + "grad_norm": 0.8897335359039827, + "learning_rate": 0.003, + "loss": 4.126, + "step": 3916 + }, + { + "epoch": 0.03917, + "grad_norm": 1.0123681906019413, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 3917 + }, + { + "epoch": 0.03918, + "grad_norm": 1.0602217244059164, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 3918 + }, + { + "epoch": 0.03919, + "grad_norm": 1.02799372222634, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 3919 + }, + { + "epoch": 0.0392, + "grad_norm": 0.7728333569294379, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 3920 + }, + { + "epoch": 0.03921, + "grad_norm": 0.7158476135413256, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 3921 + }, + { + "epoch": 0.03922, + "grad_norm": 0.8125567354684526, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3922 + }, + { + "epoch": 0.03923, + "grad_norm": 0.9863034003179841, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 3923 + }, + { + "epoch": 0.03924, + "grad_norm": 1.0668800143383192, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 3924 + }, + { + "epoch": 0.03925, + "grad_norm": 0.9458022068309033, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 3925 + }, + { + "epoch": 0.03926, + "grad_norm": 0.913203888262044, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 3926 + }, + { + "epoch": 0.03927, + "grad_norm": 0.8813175388341771, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 3927 + }, + { + "epoch": 0.03928, + "grad_norm": 0.8794380772976899, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 3928 + }, + { + "epoch": 0.03929, + "grad_norm": 1.0579124240791549, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 3929 + }, + { + "epoch": 0.0393, + "grad_norm": 1.1597780940734346, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 3930 + }, + { + "epoch": 0.03931, + "grad_norm": 0.7081278586289745, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 3931 + }, + { + "epoch": 0.03932, + "grad_norm": 0.7809455628695875, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 3932 + }, + { + "epoch": 0.03933, + "grad_norm": 0.9272931464921295, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 3933 + }, + { + "epoch": 0.03934, + "grad_norm": 0.849076414415576, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 3934 + }, + { + "epoch": 0.03935, + "grad_norm": 0.7972121383758832, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 3935 + }, + { + "epoch": 0.03936, + "grad_norm": 0.736286380065545, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 3936 + }, + { + "epoch": 0.03937, + "grad_norm": 0.7969361690394817, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 3937 + }, + { + "epoch": 0.03938, + "grad_norm": 0.7648853318696774, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 3938 + }, + { + "epoch": 0.03939, + "grad_norm": 0.793082890734816, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 3939 + }, + { + "epoch": 0.0394, + "grad_norm": 0.9421729446698976, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 3940 + }, + { + "epoch": 0.03941, + "grad_norm": 0.9532085204003842, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 3941 + }, + { + "epoch": 0.03942, + "grad_norm": 0.9597084811820548, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 3942 + }, + { + "epoch": 0.03943, + "grad_norm": 0.9484931289816191, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 3943 + }, + { + "epoch": 0.03944, + "grad_norm": 0.9294543546840306, + "learning_rate": 0.003, + "loss": 4.121, + "step": 3944 + }, + { + "epoch": 0.03945, + "grad_norm": 0.8988004519802858, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 3945 + }, + { + "epoch": 0.03946, + "grad_norm": 1.0598680988385802, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 3946 + }, + { + "epoch": 0.03947, + "grad_norm": 0.8720535328674782, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 3947 + }, + { + "epoch": 0.03948, + "grad_norm": 0.9223454341748609, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 3948 + }, + { + "epoch": 0.03949, + "grad_norm": 0.8915715222436728, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 3949 + }, + { + "epoch": 0.0395, + "grad_norm": 0.9448936409245293, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 3950 + }, + { + "epoch": 0.03951, + "grad_norm": 1.0385370830415304, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 3951 + }, + { + "epoch": 0.03952, + "grad_norm": 0.9030178362877832, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 3952 + }, + { + "epoch": 0.03953, + "grad_norm": 0.8709314822989492, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 3953 + }, + { + "epoch": 0.03954, + "grad_norm": 0.6813347239160019, + "learning_rate": 0.003, + "loss": 4.095, + "step": 3954 + }, + { + "epoch": 0.03955, + "grad_norm": 0.5792685005630919, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 3955 + }, + { + "epoch": 0.03956, + "grad_norm": 0.643092840993703, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 3956 + }, + { + "epoch": 0.03957, + "grad_norm": 0.5958384751343148, + "learning_rate": 0.003, + "loss": 4.1, + "step": 3957 + }, + { + "epoch": 0.03958, + "grad_norm": 0.7216735106401475, + "learning_rate": 0.003, + "loss": 4.077, + "step": 3958 + }, + { + "epoch": 0.03959, + "grad_norm": 1.0447079173416385, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 3959 + }, + { + "epoch": 0.0396, + "grad_norm": 1.390251410216987, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 3960 + }, + { + "epoch": 0.03961, + "grad_norm": 0.6008725960965626, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 3961 + }, + { + "epoch": 0.03962, + "grad_norm": 1.007662011162288, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 3962 + }, + { + "epoch": 0.03963, + "grad_norm": 1.1370046987648283, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 3963 + }, + { + "epoch": 0.03964, + "grad_norm": 0.747283204388864, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 3964 + }, + { + "epoch": 0.03965, + "grad_norm": 0.8660291958106119, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3965 + }, + { + "epoch": 0.03966, + "grad_norm": 0.8173353175111122, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 3966 + }, + { + "epoch": 0.03967, + "grad_norm": 0.8293952582170112, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 3967 + }, + { + "epoch": 0.03968, + "grad_norm": 0.7178876491150251, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 3968 + }, + { + "epoch": 0.03969, + "grad_norm": 0.6761246373413493, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 3969 + }, + { + "epoch": 0.0397, + "grad_norm": 0.781718933884442, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 3970 + }, + { + "epoch": 0.03971, + "grad_norm": 0.8449284031201099, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 3971 + }, + { + "epoch": 0.03972, + "grad_norm": 0.7027976770057135, + "learning_rate": 0.003, + "loss": 4.1, + "step": 3972 + }, + { + "epoch": 0.03973, + "grad_norm": 0.7615706188874817, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 3973 + }, + { + "epoch": 0.03974, + "grad_norm": 0.9198349150433933, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 3974 + }, + { + "epoch": 0.03975, + "grad_norm": 0.9827265823761052, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 3975 + }, + { + "epoch": 0.03976, + "grad_norm": 0.9713747801072962, + "learning_rate": 0.003, + "loss": 4.135, + "step": 3976 + }, + { + "epoch": 0.03977, + "grad_norm": 0.9509110934532634, + "learning_rate": 0.003, + "loss": 4.122, + "step": 3977 + }, + { + "epoch": 0.03978, + "grad_norm": 0.8340186563863297, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 3978 + }, + { + "epoch": 0.03979, + "grad_norm": 0.8587622003381145, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 3979 + }, + { + "epoch": 0.0398, + "grad_norm": 0.9900201402082706, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 3980 + }, + { + "epoch": 0.03981, + "grad_norm": 0.8963808643427871, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 3981 + }, + { + "epoch": 0.03982, + "grad_norm": 0.8219983933371988, + "learning_rate": 0.003, + "loss": 4.08, + "step": 3982 + }, + { + "epoch": 0.03983, + "grad_norm": 0.8040656765435304, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 3983 + }, + { + "epoch": 0.03984, + "grad_norm": 0.7325063244064047, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 3984 + }, + { + "epoch": 0.03985, + "grad_norm": 0.883058640101833, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 3985 + }, + { + "epoch": 0.03986, + "grad_norm": 1.0904689835883374, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3986 + }, + { + "epoch": 0.03987, + "grad_norm": 1.0592742752953161, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 3987 + }, + { + "epoch": 0.03988, + "grad_norm": 0.9215756713392262, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 3988 + }, + { + "epoch": 0.03989, + "grad_norm": 0.9308579277666478, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 3989 + }, + { + "epoch": 0.0399, + "grad_norm": 0.6893991687243239, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 3990 + }, + { + "epoch": 0.03991, + "grad_norm": 0.6626762508427944, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 3991 + }, + { + "epoch": 0.03992, + "grad_norm": 0.6277503646624452, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 3992 + }, + { + "epoch": 0.03993, + "grad_norm": 0.5389226959273035, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 3993 + }, + { + "epoch": 0.03994, + "grad_norm": 0.5575784523742429, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 3994 + }, + { + "epoch": 0.03995, + "grad_norm": 0.5740995198289247, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 3995 + }, + { + "epoch": 0.03996, + "grad_norm": 0.6506767895381821, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 3996 + }, + { + "epoch": 0.03997, + "grad_norm": 0.8018658682841067, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 3997 + }, + { + "epoch": 0.03998, + "grad_norm": 1.1165769471951579, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 3998 + }, + { + "epoch": 0.03999, + "grad_norm": 0.9183179386537218, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 3999 + }, + { + "epoch": 0.04, + "grad_norm": 0.7930924507289704, + "learning_rate": 0.003, + "loss": 4.118, + "step": 4000 + }, + { + "epoch": 0.04001, + "grad_norm": 0.7997730049089519, + "learning_rate": 0.003, + "loss": 4.114, + "step": 4001 + }, + { + "epoch": 0.04002, + "grad_norm": 0.9222776505941644, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 4002 + }, + { + "epoch": 0.04003, + "grad_norm": 1.0818631603686821, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 4003 + }, + { + "epoch": 0.04004, + "grad_norm": 0.9791417920974259, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 4004 + }, + { + "epoch": 0.04005, + "grad_norm": 0.9808789569533176, + "learning_rate": 0.003, + "loss": 4.131, + "step": 4005 + }, + { + "epoch": 0.04006, + "grad_norm": 0.8961052673893415, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 4006 + }, + { + "epoch": 0.04007, + "grad_norm": 0.8297969899387416, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 4007 + }, + { + "epoch": 0.04008, + "grad_norm": 0.936733303052199, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 4008 + }, + { + "epoch": 0.04009, + "grad_norm": 0.9644457814855122, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 4009 + }, + { + "epoch": 0.0401, + "grad_norm": 1.013041858960856, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 4010 + }, + { + "epoch": 0.04011, + "grad_norm": 1.0633452177910276, + "learning_rate": 0.003, + "loss": 4.123, + "step": 4011 + }, + { + "epoch": 0.04012, + "grad_norm": 1.0649433557962202, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 4012 + }, + { + "epoch": 0.04013, + "grad_norm": 0.7933050292506824, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 4013 + }, + { + "epoch": 0.04014, + "grad_norm": 0.7879046370570779, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4014 + }, + { + "epoch": 0.04015, + "grad_norm": 0.8241932652214115, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 4015 + }, + { + "epoch": 0.04016, + "grad_norm": 0.8658550465105778, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 4016 + }, + { + "epoch": 0.04017, + "grad_norm": 0.9228756535179865, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 4017 + }, + { + "epoch": 0.04018, + "grad_norm": 0.9531173484679942, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 4018 + }, + { + "epoch": 0.04019, + "grad_norm": 1.0866964724732295, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 4019 + }, + { + "epoch": 0.0402, + "grad_norm": 1.0577908225334638, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 4020 + }, + { + "epoch": 0.04021, + "grad_norm": 1.0725239975062906, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 4021 + }, + { + "epoch": 0.04022, + "grad_norm": 1.0475988978108968, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 4022 + }, + { + "epoch": 0.04023, + "grad_norm": 1.065198125983714, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 4023 + }, + { + "epoch": 0.04024, + "grad_norm": 0.9510719367654007, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 4024 + }, + { + "epoch": 0.04025, + "grad_norm": 0.9965721786284437, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 4025 + }, + { + "epoch": 0.04026, + "grad_norm": 1.1597169306225377, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 4026 + }, + { + "epoch": 0.04027, + "grad_norm": 0.8115276383463242, + "learning_rate": 0.003, + "loss": 4.118, + "step": 4027 + }, + { + "epoch": 0.04028, + "grad_norm": 0.7507067496165764, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 4028 + }, + { + "epoch": 0.04029, + "grad_norm": 0.8043593457637668, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 4029 + }, + { + "epoch": 0.0403, + "grad_norm": 0.8475092967533915, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 4030 + }, + { + "epoch": 0.04031, + "grad_norm": 0.7950265087708841, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 4031 + }, + { + "epoch": 0.04032, + "grad_norm": 0.7929178533450967, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 4032 + }, + { + "epoch": 0.04033, + "grad_norm": 0.9419776254998102, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 4033 + }, + { + "epoch": 0.04034, + "grad_norm": 1.0174202023914847, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 4034 + }, + { + "epoch": 0.04035, + "grad_norm": 1.0008080685810283, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 4035 + }, + { + "epoch": 0.04036, + "grad_norm": 0.9875885643657349, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 4036 + }, + { + "epoch": 0.04037, + "grad_norm": 1.035750371651436, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4037 + }, + { + "epoch": 0.04038, + "grad_norm": 1.089114000312345, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 4038 + }, + { + "epoch": 0.04039, + "grad_norm": 0.7787711745520574, + "learning_rate": 0.003, + "loss": 4.093, + "step": 4039 + }, + { + "epoch": 0.0404, + "grad_norm": 0.8304131011478352, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 4040 + }, + { + "epoch": 0.04041, + "grad_norm": 0.8845677735505116, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 4041 + }, + { + "epoch": 0.04042, + "grad_norm": 1.0056077779577761, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 4042 + }, + { + "epoch": 0.04043, + "grad_norm": 1.303684048760548, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 4043 + }, + { + "epoch": 0.04044, + "grad_norm": 0.6804116358869476, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 4044 + }, + { + "epoch": 0.04045, + "grad_norm": 0.6777286555779822, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 4045 + }, + { + "epoch": 0.04046, + "grad_norm": 0.7168976521438496, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 4046 + }, + { + "epoch": 0.04047, + "grad_norm": 0.8016076938020208, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 4047 + }, + { + "epoch": 0.04048, + "grad_norm": 0.8003360249896011, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 4048 + }, + { + "epoch": 0.04049, + "grad_norm": 0.7905185716828504, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 4049 + }, + { + "epoch": 0.0405, + "grad_norm": 0.7170296620649927, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 4050 + }, + { + "epoch": 0.04051, + "grad_norm": 0.8022124254069631, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4051 + }, + { + "epoch": 0.04052, + "grad_norm": 0.9348016278831669, + "learning_rate": 0.003, + "loss": 4.11, + "step": 4052 + }, + { + "epoch": 0.04053, + "grad_norm": 0.8618520313571836, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 4053 + }, + { + "epoch": 0.04054, + "grad_norm": 0.8261211062660251, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 4054 + }, + { + "epoch": 0.04055, + "grad_norm": 0.6845606042201053, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 4055 + }, + { + "epoch": 0.04056, + "grad_norm": 0.6070419310929447, + "learning_rate": 0.003, + "loss": 4.09, + "step": 4056 + }, + { + "epoch": 0.04057, + "grad_norm": 0.5685943884831188, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 4057 + }, + { + "epoch": 0.04058, + "grad_norm": 0.6709504673340433, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 4058 + }, + { + "epoch": 0.04059, + "grad_norm": 0.8379919064931752, + "learning_rate": 0.003, + "loss": 4.051, + "step": 4059 + }, + { + "epoch": 0.0406, + "grad_norm": 0.8620293391959047, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 4060 + }, + { + "epoch": 0.04061, + "grad_norm": 0.7852271201009998, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 4061 + }, + { + "epoch": 0.04062, + "grad_norm": 0.8559296475729065, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 4062 + }, + { + "epoch": 0.04063, + "grad_norm": 0.8779279835196838, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 4063 + }, + { + "epoch": 0.04064, + "grad_norm": 0.8897248490868346, + "learning_rate": 0.003, + "loss": 4.099, + "step": 4064 + }, + { + "epoch": 0.04065, + "grad_norm": 0.9175138271766847, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 4065 + }, + { + "epoch": 0.04066, + "grad_norm": 0.9362345350031807, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 4066 + }, + { + "epoch": 0.04067, + "grad_norm": 1.1545536887257815, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 4067 + }, + { + "epoch": 0.04068, + "grad_norm": 1.0240578066067638, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 4068 + }, + { + "epoch": 0.04069, + "grad_norm": 0.941652681317373, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 4069 + }, + { + "epoch": 0.0407, + "grad_norm": 0.9346763708267345, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 4070 + }, + { + "epoch": 0.04071, + "grad_norm": 0.9599025044832693, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 4071 + }, + { + "epoch": 0.04072, + "grad_norm": 0.8612213977045744, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 4072 + }, + { + "epoch": 0.04073, + "grad_norm": 0.8763585891787764, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 4073 + }, + { + "epoch": 0.04074, + "grad_norm": 0.8539311608465675, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 4074 + }, + { + "epoch": 0.04075, + "grad_norm": 0.8369021460447633, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 4075 + }, + { + "epoch": 0.04076, + "grad_norm": 0.8578020250558176, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 4076 + }, + { + "epoch": 0.04077, + "grad_norm": 0.85497575989368, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 4077 + }, + { + "epoch": 0.04078, + "grad_norm": 0.9407927776815804, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 4078 + }, + { + "epoch": 0.04079, + "grad_norm": 1.05128762652415, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 4079 + }, + { + "epoch": 0.0408, + "grad_norm": 1.1019406938246195, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 4080 + }, + { + "epoch": 0.04081, + "grad_norm": 0.9496061799293344, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 4081 + }, + { + "epoch": 0.04082, + "grad_norm": 0.8481874389103903, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 4082 + }, + { + "epoch": 0.04083, + "grad_norm": 0.7274191448577992, + "learning_rate": 0.003, + "loss": 4.082, + "step": 4083 + }, + { + "epoch": 0.04084, + "grad_norm": 0.7904195126005293, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 4084 + }, + { + "epoch": 0.04085, + "grad_norm": 1.0202097759202264, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 4085 + }, + { + "epoch": 0.04086, + "grad_norm": 1.201486915510698, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 4086 + }, + { + "epoch": 0.04087, + "grad_norm": 0.868844436475151, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 4087 + }, + { + "epoch": 0.04088, + "grad_norm": 0.9444985463912003, + "learning_rate": 0.003, + "loss": 4.089, + "step": 4088 + }, + { + "epoch": 0.04089, + "grad_norm": 0.9918074019046071, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 4089 + }, + { + "epoch": 0.0409, + "grad_norm": 1.0955168255438645, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 4090 + }, + { + "epoch": 0.04091, + "grad_norm": 1.063937537128444, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 4091 + }, + { + "epoch": 0.04092, + "grad_norm": 0.9343133313137877, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 4092 + }, + { + "epoch": 0.04093, + "grad_norm": 0.8938252574219301, + "learning_rate": 0.003, + "loss": 4.106, + "step": 4093 + }, + { + "epoch": 0.04094, + "grad_norm": 0.8646025553563678, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 4094 + }, + { + "epoch": 0.04095, + "grad_norm": 0.8849098800450139, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 4095 + }, + { + "epoch": 0.04096, + "grad_norm": 0.8639434333725027, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 4096 + }, + { + "epoch": 0.04097, + "grad_norm": 0.8248508344367464, + "learning_rate": 0.003, + "loss": 4.096, + "step": 4097 + }, + { + "epoch": 0.04098, + "grad_norm": 0.7799456648454658, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 4098 + }, + { + "epoch": 0.04099, + "grad_norm": 0.8045819576448443, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 4099 + }, + { + "epoch": 0.041, + "grad_norm": 0.8747001760069023, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 4100 + }, + { + "epoch": 0.04101, + "grad_norm": 0.8938516305008589, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 4101 + }, + { + "epoch": 0.04102, + "grad_norm": 0.8240754586578543, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 4102 + }, + { + "epoch": 0.04103, + "grad_norm": 0.7515676676481744, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 4103 + }, + { + "epoch": 0.04104, + "grad_norm": 0.7588461451530453, + "learning_rate": 0.003, + "loss": 4.099, + "step": 4104 + }, + { + "epoch": 0.04105, + "grad_norm": 0.8416483809920972, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 4105 + }, + { + "epoch": 0.04106, + "grad_norm": 0.9997989043605089, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 4106 + }, + { + "epoch": 0.04107, + "grad_norm": 1.4519560189758212, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 4107 + }, + { + "epoch": 0.04108, + "grad_norm": 0.6881032147408636, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 4108 + }, + { + "epoch": 0.04109, + "grad_norm": 0.7741592484410211, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4109 + }, + { + "epoch": 0.0411, + "grad_norm": 0.8818900570270763, + "learning_rate": 0.003, + "loss": 4.091, + "step": 4110 + }, + { + "epoch": 0.04111, + "grad_norm": 1.0723961097005286, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 4111 + }, + { + "epoch": 0.04112, + "grad_norm": 1.0495234672278315, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 4112 + }, + { + "epoch": 0.04113, + "grad_norm": 1.028878109333595, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 4113 + }, + { + "epoch": 0.04114, + "grad_norm": 0.7814661056488829, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 4114 + }, + { + "epoch": 0.04115, + "grad_norm": 0.7046355162923472, + "learning_rate": 0.003, + "loss": 4.089, + "step": 4115 + }, + { + "epoch": 0.04116, + "grad_norm": 0.7912332755348991, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 4116 + }, + { + "epoch": 0.04117, + "grad_norm": 0.8341278502744938, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 4117 + }, + { + "epoch": 0.04118, + "grad_norm": 0.8875854653843481, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 4118 + }, + { + "epoch": 0.04119, + "grad_norm": 0.8609577714531109, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 4119 + }, + { + "epoch": 0.0412, + "grad_norm": 0.9318473319692638, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 4120 + }, + { + "epoch": 0.04121, + "grad_norm": 0.9493333365772441, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 4121 + }, + { + "epoch": 0.04122, + "grad_norm": 0.9836587921676327, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 4122 + }, + { + "epoch": 0.04123, + "grad_norm": 0.9314136846053894, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 4123 + }, + { + "epoch": 0.04124, + "grad_norm": 0.8615395432982054, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 4124 + }, + { + "epoch": 0.04125, + "grad_norm": 0.8303231719235581, + "learning_rate": 0.003, + "loss": 4.107, + "step": 4125 + }, + { + "epoch": 0.04126, + "grad_norm": 0.9351121203757022, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 4126 + }, + { + "epoch": 0.04127, + "grad_norm": 0.9548721015062305, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 4127 + }, + { + "epoch": 0.04128, + "grad_norm": 0.904134032279689, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4128 + }, + { + "epoch": 0.04129, + "grad_norm": 1.0360809502468598, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 4129 + }, + { + "epoch": 0.0413, + "grad_norm": 1.141478517660115, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 4130 + }, + { + "epoch": 0.04131, + "grad_norm": 1.097730707750305, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 4131 + }, + { + "epoch": 0.04132, + "grad_norm": 0.9015795328637733, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 4132 + }, + { + "epoch": 0.04133, + "grad_norm": 0.9605681859330745, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 4133 + }, + { + "epoch": 0.04134, + "grad_norm": 0.9723288613378216, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 4134 + }, + { + "epoch": 0.04135, + "grad_norm": 0.9769907462898332, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 4135 + }, + { + "epoch": 0.04136, + "grad_norm": 0.8696285037251947, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 4136 + }, + { + "epoch": 0.04137, + "grad_norm": 0.8622412103402255, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 4137 + }, + { + "epoch": 0.04138, + "grad_norm": 0.8940895987636069, + "learning_rate": 0.003, + "loss": 4.114, + "step": 4138 + }, + { + "epoch": 0.04139, + "grad_norm": 0.8558944236718854, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 4139 + }, + { + "epoch": 0.0414, + "grad_norm": 0.7906160709298794, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 4140 + }, + { + "epoch": 0.04141, + "grad_norm": 0.8043674825801352, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 4141 + }, + { + "epoch": 0.04142, + "grad_norm": 0.8668589703050257, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 4142 + }, + { + "epoch": 0.04143, + "grad_norm": 1.0143988611429429, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 4143 + }, + { + "epoch": 0.04144, + "grad_norm": 1.1221724323882003, + "learning_rate": 0.003, + "loss": 4.105, + "step": 4144 + }, + { + "epoch": 0.04145, + "grad_norm": 0.9750842855026217, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 4145 + }, + { + "epoch": 0.04146, + "grad_norm": 0.837063893029182, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 4146 + }, + { + "epoch": 0.04147, + "grad_norm": 0.817279014947959, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 4147 + }, + { + "epoch": 0.04148, + "grad_norm": 0.6652410842078591, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 4148 + }, + { + "epoch": 0.04149, + "grad_norm": 0.601931902805492, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 4149 + }, + { + "epoch": 0.0415, + "grad_norm": 0.5744893147283616, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 4150 + }, + { + "epoch": 0.04151, + "grad_norm": 0.5594756907939531, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 4151 + }, + { + "epoch": 0.04152, + "grad_norm": 0.6696918347600734, + "learning_rate": 0.003, + "loss": 4.097, + "step": 4152 + }, + { + "epoch": 0.04153, + "grad_norm": 0.8337852170452664, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 4153 + }, + { + "epoch": 0.04154, + "grad_norm": 0.8486120723348882, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 4154 + }, + { + "epoch": 0.04155, + "grad_norm": 0.697828069478384, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 4155 + }, + { + "epoch": 0.04156, + "grad_norm": 0.4957422759820407, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 4156 + }, + { + "epoch": 0.04157, + "grad_norm": 0.5563098539794213, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 4157 + }, + { + "epoch": 0.04158, + "grad_norm": 0.6673553499632091, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 4158 + }, + { + "epoch": 0.04159, + "grad_norm": 0.8290986224854208, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 4159 + }, + { + "epoch": 0.0416, + "grad_norm": 0.9164477827441916, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 4160 + }, + { + "epoch": 0.04161, + "grad_norm": 0.8138741601967407, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 4161 + }, + { + "epoch": 0.04162, + "grad_norm": 0.7719431484259072, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 4162 + }, + { + "epoch": 0.04163, + "grad_norm": 0.7815652988597618, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 4163 + }, + { + "epoch": 0.04164, + "grad_norm": 0.9136897661923532, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 4164 + }, + { + "epoch": 0.04165, + "grad_norm": 1.0910864238229012, + "learning_rate": 0.003, + "loss": 4.089, + "step": 4165 + }, + { + "epoch": 0.04166, + "grad_norm": 1.0985022482979871, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 4166 + }, + { + "epoch": 0.04167, + "grad_norm": 0.884758586994079, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 4167 + }, + { + "epoch": 0.04168, + "grad_norm": 0.8375929677822369, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 4168 + }, + { + "epoch": 0.04169, + "grad_norm": 0.8912187343234272, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 4169 + }, + { + "epoch": 0.0417, + "grad_norm": 1.09510323581272, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 4170 + }, + { + "epoch": 0.04171, + "grad_norm": 0.9235321645280937, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4171 + }, + { + "epoch": 0.04172, + "grad_norm": 0.9054535762684148, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 4172 + }, + { + "epoch": 0.04173, + "grad_norm": 0.9533816623318108, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 4173 + }, + { + "epoch": 0.04174, + "grad_norm": 0.9693424734311672, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 4174 + }, + { + "epoch": 0.04175, + "grad_norm": 0.9476037015111809, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 4175 + }, + { + "epoch": 0.04176, + "grad_norm": 0.9584295419287631, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 4176 + }, + { + "epoch": 0.04177, + "grad_norm": 0.9248605600298978, + "learning_rate": 0.003, + "loss": 4.072, + "step": 4177 + }, + { + "epoch": 0.04178, + "grad_norm": 1.0049448032304569, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 4178 + }, + { + "epoch": 0.04179, + "grad_norm": 1.0827168186886107, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 4179 + }, + { + "epoch": 0.0418, + "grad_norm": 0.8883360779651741, + "learning_rate": 0.003, + "loss": 4.098, + "step": 4180 + }, + { + "epoch": 0.04181, + "grad_norm": 1.0344215717913283, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 4181 + }, + { + "epoch": 0.04182, + "grad_norm": 0.9869075813400712, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 4182 + }, + { + "epoch": 0.04183, + "grad_norm": 0.8294921509305445, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 4183 + }, + { + "epoch": 0.04184, + "grad_norm": 0.869514009963607, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 4184 + }, + { + "epoch": 0.04185, + "grad_norm": 1.0166270234192272, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 4185 + }, + { + "epoch": 0.04186, + "grad_norm": 0.9675533699418017, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 4186 + }, + { + "epoch": 0.04187, + "grad_norm": 1.118617978668176, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 4187 + }, + { + "epoch": 0.04188, + "grad_norm": 0.9597380404807206, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 4188 + }, + { + "epoch": 0.04189, + "grad_norm": 0.8821463731911939, + "learning_rate": 0.003, + "loss": 4.1, + "step": 4189 + }, + { + "epoch": 0.0419, + "grad_norm": 0.8335574771570168, + "learning_rate": 0.003, + "loss": 4.097, + "step": 4190 + }, + { + "epoch": 0.04191, + "grad_norm": 0.960590386364999, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 4191 + }, + { + "epoch": 0.04192, + "grad_norm": 1.1082087641443477, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 4192 + }, + { + "epoch": 0.04193, + "grad_norm": 1.1014870968077684, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 4193 + }, + { + "epoch": 0.04194, + "grad_norm": 1.004137077094656, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 4194 + }, + { + "epoch": 0.04195, + "grad_norm": 1.0060456499295114, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 4195 + }, + { + "epoch": 0.04196, + "grad_norm": 0.9781412326562324, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 4196 + }, + { + "epoch": 0.04197, + "grad_norm": 0.966982025696319, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 4197 + }, + { + "epoch": 0.04198, + "grad_norm": 1.0321323425834754, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 4198 + }, + { + "epoch": 0.04199, + "grad_norm": 0.9725763984900901, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 4199 + }, + { + "epoch": 0.042, + "grad_norm": 0.9044032352315242, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 4200 + }, + { + "epoch": 0.04201, + "grad_norm": 0.876685657514267, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4201 + }, + { + "epoch": 0.04202, + "grad_norm": 1.0379948329277224, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 4202 + }, + { + "epoch": 0.04203, + "grad_norm": 0.9407666140032972, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 4203 + }, + { + "epoch": 0.04204, + "grad_norm": 0.9627655702680004, + "learning_rate": 0.003, + "loss": 4.085, + "step": 4204 + }, + { + "epoch": 0.04205, + "grad_norm": 0.879727149930489, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 4205 + }, + { + "epoch": 0.04206, + "grad_norm": 0.7201424170354667, + "learning_rate": 0.003, + "loss": 4.098, + "step": 4206 + }, + { + "epoch": 0.04207, + "grad_norm": 0.8722237657873413, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 4207 + }, + { + "epoch": 0.04208, + "grad_norm": 1.057153458391563, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 4208 + }, + { + "epoch": 0.04209, + "grad_norm": 1.0856375627812824, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 4209 + }, + { + "epoch": 0.0421, + "grad_norm": 0.8686311196866661, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 4210 + }, + { + "epoch": 0.04211, + "grad_norm": 0.772423867231878, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 4211 + }, + { + "epoch": 0.04212, + "grad_norm": 0.9206795604812895, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 4212 + }, + { + "epoch": 0.04213, + "grad_norm": 1.024206681316176, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 4213 + }, + { + "epoch": 0.04214, + "grad_norm": 0.9670322877487537, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 4214 + }, + { + "epoch": 0.04215, + "grad_norm": 0.7494961355008928, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 4215 + }, + { + "epoch": 0.04216, + "grad_norm": 0.765486635210703, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 4216 + }, + { + "epoch": 0.04217, + "grad_norm": 0.7160617513742286, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 4217 + }, + { + "epoch": 0.04218, + "grad_norm": 0.8033027325871259, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 4218 + }, + { + "epoch": 0.04219, + "grad_norm": 0.8144707096133446, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 4219 + }, + { + "epoch": 0.0422, + "grad_norm": 0.8870053013997257, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 4220 + }, + { + "epoch": 0.04221, + "grad_norm": 1.0572558853767193, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 4221 + }, + { + "epoch": 0.04222, + "grad_norm": 0.8601884630133385, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 4222 + }, + { + "epoch": 0.04223, + "grad_norm": 0.7192916764502633, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 4223 + }, + { + "epoch": 0.04224, + "grad_norm": 0.7121512553768459, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 4224 + }, + { + "epoch": 0.04225, + "grad_norm": 0.7021515831824932, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 4225 + }, + { + "epoch": 0.04226, + "grad_norm": 0.8314637642734086, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 4226 + }, + { + "epoch": 0.04227, + "grad_norm": 0.9670578496344213, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 4227 + }, + { + "epoch": 0.04228, + "grad_norm": 1.049287879558258, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 4228 + }, + { + "epoch": 0.04229, + "grad_norm": 1.0992300949270042, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 4229 + }, + { + "epoch": 0.0423, + "grad_norm": 0.900897411198522, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 4230 + }, + { + "epoch": 0.04231, + "grad_norm": 0.9451671959275939, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 4231 + }, + { + "epoch": 0.04232, + "grad_norm": 0.7956005679951864, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 4232 + }, + { + "epoch": 0.04233, + "grad_norm": 0.8292582126895408, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 4233 + }, + { + "epoch": 0.04234, + "grad_norm": 0.8871903907197981, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 4234 + }, + { + "epoch": 0.04235, + "grad_norm": 0.9300140410326806, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 4235 + }, + { + "epoch": 0.04236, + "grad_norm": 1.0741902913167412, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 4236 + }, + { + "epoch": 0.04237, + "grad_norm": 0.9580888994428801, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 4237 + }, + { + "epoch": 0.04238, + "grad_norm": 0.9482815531910007, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 4238 + }, + { + "epoch": 0.04239, + "grad_norm": 0.9555806662772239, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 4239 + }, + { + "epoch": 0.0424, + "grad_norm": 0.8664629875137623, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 4240 + }, + { + "epoch": 0.04241, + "grad_norm": 0.9630766502725931, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 4241 + }, + { + "epoch": 0.04242, + "grad_norm": 1.0804387798244068, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 4242 + }, + { + "epoch": 0.04243, + "grad_norm": 0.9274350779718051, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 4243 + }, + { + "epoch": 0.04244, + "grad_norm": 0.913816427911704, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 4244 + }, + { + "epoch": 0.04245, + "grad_norm": 1.106413525319224, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 4245 + }, + { + "epoch": 0.04246, + "grad_norm": 0.9030634078700811, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 4246 + }, + { + "epoch": 0.04247, + "grad_norm": 0.8416832028937715, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 4247 + }, + { + "epoch": 0.04248, + "grad_norm": 0.8169991638726913, + "learning_rate": 0.003, + "loss": 4.081, + "step": 4248 + }, + { + "epoch": 0.04249, + "grad_norm": 0.8044824810175584, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 4249 + }, + { + "epoch": 0.0425, + "grad_norm": 0.8896696091741753, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 4250 + }, + { + "epoch": 0.04251, + "grad_norm": 0.9502948772765745, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 4251 + }, + { + "epoch": 0.04252, + "grad_norm": 1.0326556881542461, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 4252 + }, + { + "epoch": 0.04253, + "grad_norm": 1.0173408114295495, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 4253 + }, + { + "epoch": 0.04254, + "grad_norm": 1.0646057446624475, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 4254 + }, + { + "epoch": 0.04255, + "grad_norm": 0.8420645317602736, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 4255 + }, + { + "epoch": 0.04256, + "grad_norm": 0.7201557468187585, + "learning_rate": 0.003, + "loss": 4.078, + "step": 4256 + }, + { + "epoch": 0.04257, + "grad_norm": 0.7996495415748023, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 4257 + }, + { + "epoch": 0.04258, + "grad_norm": 0.8858738615733014, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 4258 + }, + { + "epoch": 0.04259, + "grad_norm": 0.9662802527822355, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 4259 + }, + { + "epoch": 0.0426, + "grad_norm": 0.9169898773347952, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 4260 + }, + { + "epoch": 0.04261, + "grad_norm": 0.9208708917372306, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 4261 + }, + { + "epoch": 0.04262, + "grad_norm": 0.8866710041215375, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 4262 + }, + { + "epoch": 0.04263, + "grad_norm": 0.8426381040253105, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 4263 + }, + { + "epoch": 0.04264, + "grad_norm": 0.743799590275064, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 4264 + }, + { + "epoch": 0.04265, + "grad_norm": 0.8815249453115196, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 4265 + }, + { + "epoch": 0.04266, + "grad_norm": 0.8452841678134956, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 4266 + }, + { + "epoch": 0.04267, + "grad_norm": 0.7585523353787604, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 4267 + }, + { + "epoch": 0.04268, + "grad_norm": 0.7042766430430472, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 4268 + }, + { + "epoch": 0.04269, + "grad_norm": 0.6767603788338118, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 4269 + }, + { + "epoch": 0.0427, + "grad_norm": 0.6619260992554636, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 4270 + }, + { + "epoch": 0.04271, + "grad_norm": 0.716784572802792, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 4271 + }, + { + "epoch": 0.04272, + "grad_norm": 0.9018222570354946, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 4272 + }, + { + "epoch": 0.04273, + "grad_norm": 1.0946540429372817, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 4273 + }, + { + "epoch": 0.04274, + "grad_norm": 0.8458329250448817, + "learning_rate": 0.003, + "loss": 4.086, + "step": 4274 + }, + { + "epoch": 0.04275, + "grad_norm": 0.6435172962120423, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 4275 + }, + { + "epoch": 0.04276, + "grad_norm": 0.6643590249615142, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 4276 + }, + { + "epoch": 0.04277, + "grad_norm": 0.79792656427568, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 4277 + }, + { + "epoch": 0.04278, + "grad_norm": 0.740840553173491, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 4278 + }, + { + "epoch": 0.04279, + "grad_norm": 0.7711154804154717, + "learning_rate": 0.003, + "loss": 4.071, + "step": 4279 + }, + { + "epoch": 0.0428, + "grad_norm": 0.7250008586500746, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 4280 + }, + { + "epoch": 0.04281, + "grad_norm": 0.7499387037166975, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 4281 + }, + { + "epoch": 0.04282, + "grad_norm": 0.7545960253753359, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 4282 + }, + { + "epoch": 0.04283, + "grad_norm": 0.839588483108746, + "learning_rate": 0.003, + "loss": 4.12, + "step": 4283 + }, + { + "epoch": 0.04284, + "grad_norm": 0.86181150410575, + "learning_rate": 0.003, + "loss": 4.074, + "step": 4284 + }, + { + "epoch": 0.04285, + "grad_norm": 0.8662575012952002, + "learning_rate": 0.003, + "loss": 4.09, + "step": 4285 + }, + { + "epoch": 0.04286, + "grad_norm": 1.0398949684975605, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 4286 + }, + { + "epoch": 0.04287, + "grad_norm": 1.1167633903119278, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 4287 + }, + { + "epoch": 0.04288, + "grad_norm": 0.9953157675421085, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 4288 + }, + { + "epoch": 0.04289, + "grad_norm": 0.9963288991872585, + "learning_rate": 0.003, + "loss": 4.087, + "step": 4289 + }, + { + "epoch": 0.0429, + "grad_norm": 1.0134088286440928, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 4290 + }, + { + "epoch": 0.04291, + "grad_norm": 1.0632191551515977, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4291 + }, + { + "epoch": 0.04292, + "grad_norm": 1.1744325304960588, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 4292 + }, + { + "epoch": 0.04293, + "grad_norm": 0.8830847280464275, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 4293 + }, + { + "epoch": 0.04294, + "grad_norm": 0.8997282546954833, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 4294 + }, + { + "epoch": 0.04295, + "grad_norm": 1.0130781104561735, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 4295 + }, + { + "epoch": 0.04296, + "grad_norm": 0.7688499577075354, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 4296 + }, + { + "epoch": 0.04297, + "grad_norm": 0.8069754114362383, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 4297 + }, + { + "epoch": 0.04298, + "grad_norm": 0.7696772164144003, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 4298 + }, + { + "epoch": 0.04299, + "grad_norm": 0.731529824570407, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 4299 + }, + { + "epoch": 0.043, + "grad_norm": 0.7813911597495976, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 4300 + }, + { + "epoch": 0.04301, + "grad_norm": 0.9515070484621972, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 4301 + }, + { + "epoch": 0.04302, + "grad_norm": 1.1524985858220247, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 4302 + }, + { + "epoch": 0.04303, + "grad_norm": 0.9661803127063716, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 4303 + }, + { + "epoch": 0.04304, + "grad_norm": 0.8958980804239224, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 4304 + }, + { + "epoch": 0.04305, + "grad_norm": 0.9457226757083944, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 4305 + }, + { + "epoch": 0.04306, + "grad_norm": 0.866320067732626, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 4306 + }, + { + "epoch": 0.04307, + "grad_norm": 0.9815566634548711, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 4307 + }, + { + "epoch": 0.04308, + "grad_norm": 1.1045714176495842, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 4308 + }, + { + "epoch": 0.04309, + "grad_norm": 0.9997081776371058, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 4309 + }, + { + "epoch": 0.0431, + "grad_norm": 1.014433219426494, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 4310 + }, + { + "epoch": 0.04311, + "grad_norm": 1.0841431057954531, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 4311 + }, + { + "epoch": 0.04312, + "grad_norm": 0.8706786032806877, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4312 + }, + { + "epoch": 0.04313, + "grad_norm": 1.0443926449585144, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 4313 + }, + { + "epoch": 0.04314, + "grad_norm": 1.0606632584516302, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 4314 + }, + { + "epoch": 0.04315, + "grad_norm": 0.8710054154104404, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 4315 + }, + { + "epoch": 0.04316, + "grad_norm": 0.7770801944004984, + "learning_rate": 0.003, + "loss": 4.101, + "step": 4316 + }, + { + "epoch": 0.04317, + "grad_norm": 0.808693446756647, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 4317 + }, + { + "epoch": 0.04318, + "grad_norm": 0.783023904953245, + "learning_rate": 0.003, + "loss": 4.059, + "step": 4318 + }, + { + "epoch": 0.04319, + "grad_norm": 0.7045592780420906, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 4319 + }, + { + "epoch": 0.0432, + "grad_norm": 0.6211212627554941, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 4320 + }, + { + "epoch": 0.04321, + "grad_norm": 0.7218417386943828, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 4321 + }, + { + "epoch": 0.04322, + "grad_norm": 0.7864622459064595, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 4322 + }, + { + "epoch": 0.04323, + "grad_norm": 1.0759947667136016, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 4323 + }, + { + "epoch": 0.04324, + "grad_norm": 1.3618558609167744, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 4324 + }, + { + "epoch": 0.04325, + "grad_norm": 0.8537661961267661, + "learning_rate": 0.003, + "loss": 4.108, + "step": 4325 + }, + { + "epoch": 0.04326, + "grad_norm": 0.8673429528992446, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 4326 + }, + { + "epoch": 0.04327, + "grad_norm": 0.8784142198291053, + "learning_rate": 0.003, + "loss": 4.071, + "step": 4327 + }, + { + "epoch": 0.04328, + "grad_norm": 0.964766867252245, + "learning_rate": 0.003, + "loss": 4.088, + "step": 4328 + }, + { + "epoch": 0.04329, + "grad_norm": 1.159486755106838, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 4329 + }, + { + "epoch": 0.0433, + "grad_norm": 0.9798503929358533, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 4330 + }, + { + "epoch": 0.04331, + "grad_norm": 0.8036875374792932, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 4331 + }, + { + "epoch": 0.04332, + "grad_norm": 0.8974020360029874, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 4332 + }, + { + "epoch": 0.04333, + "grad_norm": 0.952759627899482, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 4333 + }, + { + "epoch": 0.04334, + "grad_norm": 0.9486168347147088, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 4334 + }, + { + "epoch": 0.04335, + "grad_norm": 0.9411584678122592, + "learning_rate": 0.003, + "loss": 4.073, + "step": 4335 + }, + { + "epoch": 0.04336, + "grad_norm": 0.8380130707060495, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 4336 + }, + { + "epoch": 0.04337, + "grad_norm": 0.8435104385458285, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 4337 + }, + { + "epoch": 0.04338, + "grad_norm": 0.8283613873296221, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 4338 + }, + { + "epoch": 0.04339, + "grad_norm": 0.7575807206093103, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 4339 + }, + { + "epoch": 0.0434, + "grad_norm": 0.7052060232541129, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 4340 + }, + { + "epoch": 0.04341, + "grad_norm": 0.7393447733605004, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 4341 + }, + { + "epoch": 0.04342, + "grad_norm": 0.6868972479892425, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 4342 + }, + { + "epoch": 0.04343, + "grad_norm": 0.642426794922686, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 4343 + }, + { + "epoch": 0.04344, + "grad_norm": 0.6793615281465072, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 4344 + }, + { + "epoch": 0.04345, + "grad_norm": 0.6547984850581937, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 4345 + }, + { + "epoch": 0.04346, + "grad_norm": 0.7129305907572272, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 4346 + }, + { + "epoch": 0.04347, + "grad_norm": 0.7392919981263668, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 4347 + }, + { + "epoch": 0.04348, + "grad_norm": 0.7595005587543163, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 4348 + }, + { + "epoch": 0.04349, + "grad_norm": 0.8185974571279835, + "learning_rate": 0.003, + "loss": 4.087, + "step": 4349 + }, + { + "epoch": 0.0435, + "grad_norm": 0.9514565408114718, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 4350 + }, + { + "epoch": 0.04351, + "grad_norm": 1.2109347120622453, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 4351 + }, + { + "epoch": 0.04352, + "grad_norm": 0.9471998868314153, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 4352 + }, + { + "epoch": 0.04353, + "grad_norm": 0.9889455977002414, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 4353 + }, + { + "epoch": 0.04354, + "grad_norm": 1.0731295960755523, + "learning_rate": 0.003, + "loss": 4.088, + "step": 4354 + }, + { + "epoch": 0.04355, + "grad_norm": 0.9572032046556027, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 4355 + }, + { + "epoch": 0.04356, + "grad_norm": 0.9581656158071228, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 4356 + }, + { + "epoch": 0.04357, + "grad_norm": 0.9195125428358254, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 4357 + }, + { + "epoch": 0.04358, + "grad_norm": 0.9636453763618577, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 4358 + }, + { + "epoch": 0.04359, + "grad_norm": 1.0660232103277558, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 4359 + }, + { + "epoch": 0.0436, + "grad_norm": 0.8061306785792844, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 4360 + }, + { + "epoch": 0.04361, + "grad_norm": 0.8093145893752948, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 4361 + }, + { + "epoch": 0.04362, + "grad_norm": 0.8171945284904929, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 4362 + }, + { + "epoch": 0.04363, + "grad_norm": 0.9868190592217813, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 4363 + }, + { + "epoch": 0.04364, + "grad_norm": 1.0426300462439566, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 4364 + }, + { + "epoch": 0.04365, + "grad_norm": 0.8588093044292321, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 4365 + }, + { + "epoch": 0.04366, + "grad_norm": 0.8909026739864028, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 4366 + }, + { + "epoch": 0.04367, + "grad_norm": 0.873925255651534, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 4367 + }, + { + "epoch": 0.04368, + "grad_norm": 0.779588637997916, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 4368 + }, + { + "epoch": 0.04369, + "grad_norm": 0.7966324124265414, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 4369 + }, + { + "epoch": 0.0437, + "grad_norm": 0.669780598735826, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 4370 + }, + { + "epoch": 0.04371, + "grad_norm": 0.7760483978802495, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 4371 + }, + { + "epoch": 0.04372, + "grad_norm": 0.9716643560234968, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 4372 + }, + { + "epoch": 0.04373, + "grad_norm": 1.2693018325935452, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 4373 + }, + { + "epoch": 0.04374, + "grad_norm": 1.0224726158531943, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 4374 + }, + { + "epoch": 0.04375, + "grad_norm": 0.8263535068880948, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 4375 + }, + { + "epoch": 0.04376, + "grad_norm": 0.7522103171040029, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 4376 + }, + { + "epoch": 0.04377, + "grad_norm": 0.8208201264915056, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 4377 + }, + { + "epoch": 0.04378, + "grad_norm": 1.005702386332168, + "learning_rate": 0.003, + "loss": 4.109, + "step": 4378 + }, + { + "epoch": 0.04379, + "grad_norm": 1.1786773838771147, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 4379 + }, + { + "epoch": 0.0438, + "grad_norm": 0.7871810906501763, + "learning_rate": 0.003, + "loss": 4.092, + "step": 4380 + }, + { + "epoch": 0.04381, + "grad_norm": 0.7577866085120181, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 4381 + }, + { + "epoch": 0.04382, + "grad_norm": 0.830788547447954, + "learning_rate": 0.003, + "loss": 4.096, + "step": 4382 + }, + { + "epoch": 0.04383, + "grad_norm": 0.9439340501432005, + "learning_rate": 0.003, + "loss": 4.076, + "step": 4383 + }, + { + "epoch": 0.04384, + "grad_norm": 0.9569811292204884, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 4384 + }, + { + "epoch": 0.04385, + "grad_norm": 0.9621273589739346, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 4385 + }, + { + "epoch": 0.04386, + "grad_norm": 1.0027765639035047, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 4386 + }, + { + "epoch": 0.04387, + "grad_norm": 0.8877446565419372, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 4387 + }, + { + "epoch": 0.04388, + "grad_norm": 1.004365074328248, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 4388 + }, + { + "epoch": 0.04389, + "grad_norm": 1.167012365586639, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 4389 + }, + { + "epoch": 0.0439, + "grad_norm": 1.126759575747849, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 4390 + }, + { + "epoch": 0.04391, + "grad_norm": 0.8766190371623821, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 4391 + }, + { + "epoch": 0.04392, + "grad_norm": 0.9422353250001032, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 4392 + }, + { + "epoch": 0.04393, + "grad_norm": 0.9388161319056185, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 4393 + }, + { + "epoch": 0.04394, + "grad_norm": 0.9030510084400544, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 4394 + }, + { + "epoch": 0.04395, + "grad_norm": 0.8932781141162982, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4395 + }, + { + "epoch": 0.04396, + "grad_norm": 0.8796817611377403, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 4396 + }, + { + "epoch": 0.04397, + "grad_norm": 0.844265233977852, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 4397 + }, + { + "epoch": 0.04398, + "grad_norm": 0.9152454694688225, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 4398 + }, + { + "epoch": 0.04399, + "grad_norm": 0.9294951448787108, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 4399 + }, + { + "epoch": 0.044, + "grad_norm": 1.0106829137077165, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 4400 + }, + { + "epoch": 0.04401, + "grad_norm": 0.9222214369606679, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 4401 + }, + { + "epoch": 0.04402, + "grad_norm": 0.8861574452686165, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 4402 + }, + { + "epoch": 0.04403, + "grad_norm": 0.9244720061111934, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 4403 + }, + { + "epoch": 0.04404, + "grad_norm": 1.0163229703427172, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 4404 + }, + { + "epoch": 0.04405, + "grad_norm": 1.112663606677766, + "learning_rate": 0.003, + "loss": 4.103, + "step": 4405 + }, + { + "epoch": 0.04406, + "grad_norm": 0.7836665936004462, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 4406 + }, + { + "epoch": 0.04407, + "grad_norm": 0.6834208693181686, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 4407 + }, + { + "epoch": 0.04408, + "grad_norm": 0.7215319828238491, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 4408 + }, + { + "epoch": 0.04409, + "grad_norm": 0.7547734730978946, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 4409 + }, + { + "epoch": 0.0441, + "grad_norm": 0.9855505388299907, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 4410 + }, + { + "epoch": 0.04411, + "grad_norm": 1.3745879066451838, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 4411 + }, + { + "epoch": 0.04412, + "grad_norm": 0.6697260575539041, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 4412 + }, + { + "epoch": 0.04413, + "grad_norm": 0.7289649196117617, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 4413 + }, + { + "epoch": 0.04414, + "grad_norm": 0.8710542961593792, + "learning_rate": 0.003, + "loss": 4.121, + "step": 4414 + }, + { + "epoch": 0.04415, + "grad_norm": 1.006707768649903, + "learning_rate": 0.003, + "loss": 4.089, + "step": 4415 + }, + { + "epoch": 0.04416, + "grad_norm": 0.9528175763218202, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 4416 + }, + { + "epoch": 0.04417, + "grad_norm": 0.8424334779705969, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 4417 + }, + { + "epoch": 0.04418, + "grad_norm": 0.97351224925848, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 4418 + }, + { + "epoch": 0.04419, + "grad_norm": 1.0808697355237324, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 4419 + }, + { + "epoch": 0.0442, + "grad_norm": 1.167054123021739, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 4420 + }, + { + "epoch": 0.04421, + "grad_norm": 0.8176085170190497, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 4421 + }, + { + "epoch": 0.04422, + "grad_norm": 0.7560953731068145, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 4422 + }, + { + "epoch": 0.04423, + "grad_norm": 0.8430900886347851, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4423 + }, + { + "epoch": 0.04424, + "grad_norm": 0.7698126783686144, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 4424 + }, + { + "epoch": 0.04425, + "grad_norm": 0.7635205731836768, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 4425 + }, + { + "epoch": 0.04426, + "grad_norm": 0.7393576832666733, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 4426 + }, + { + "epoch": 0.04427, + "grad_norm": 0.7066164659139067, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 4427 + }, + { + "epoch": 0.04428, + "grad_norm": 0.584092377564762, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 4428 + }, + { + "epoch": 0.04429, + "grad_norm": 0.5955745561562374, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 4429 + }, + { + "epoch": 0.0443, + "grad_norm": 0.7642349609553867, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 4430 + }, + { + "epoch": 0.04431, + "grad_norm": 0.9897440622578424, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 4431 + }, + { + "epoch": 0.04432, + "grad_norm": 1.3796673035154565, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 4432 + }, + { + "epoch": 0.04433, + "grad_norm": 0.7338783584401088, + "learning_rate": 0.003, + "loss": 4.079, + "step": 4433 + }, + { + "epoch": 0.04434, + "grad_norm": 0.9478779466631931, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 4434 + }, + { + "epoch": 0.04435, + "grad_norm": 1.0373098684540567, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 4435 + }, + { + "epoch": 0.04436, + "grad_norm": 0.9002494417955366, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 4436 + }, + { + "epoch": 0.04437, + "grad_norm": 0.9423712815921459, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 4437 + }, + { + "epoch": 0.04438, + "grad_norm": 0.9881181207836567, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 4438 + }, + { + "epoch": 0.04439, + "grad_norm": 1.140055930440485, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 4439 + }, + { + "epoch": 0.0444, + "grad_norm": 0.8857720718409576, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 4440 + }, + { + "epoch": 0.04441, + "grad_norm": 0.9085661359962524, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 4441 + }, + { + "epoch": 0.04442, + "grad_norm": 0.8204868786177558, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 4442 + }, + { + "epoch": 0.04443, + "grad_norm": 0.6457268453463154, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 4443 + }, + { + "epoch": 0.04444, + "grad_norm": 0.6492396831004539, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 4444 + }, + { + "epoch": 0.04445, + "grad_norm": 0.6928510619497573, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 4445 + }, + { + "epoch": 0.04446, + "grad_norm": 0.851264848376963, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 4446 + }, + { + "epoch": 0.04447, + "grad_norm": 0.9403414167914219, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 4447 + }, + { + "epoch": 0.04448, + "grad_norm": 1.0475208155252718, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 4448 + }, + { + "epoch": 0.04449, + "grad_norm": 0.8839010679595062, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 4449 + }, + { + "epoch": 0.0445, + "grad_norm": 0.9663307778578972, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 4450 + }, + { + "epoch": 0.04451, + "grad_norm": 1.0371788520078717, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 4451 + }, + { + "epoch": 0.04452, + "grad_norm": 1.2078981755286253, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 4452 + }, + { + "epoch": 0.04453, + "grad_norm": 0.8296222855961709, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 4453 + }, + { + "epoch": 0.04454, + "grad_norm": 0.8721163575738796, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 4454 + }, + { + "epoch": 0.04455, + "grad_norm": 0.9017669849588411, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 4455 + }, + { + "epoch": 0.04456, + "grad_norm": 1.0543326711835521, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 4456 + }, + { + "epoch": 0.04457, + "grad_norm": 1.1758149593074942, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 4457 + }, + { + "epoch": 0.04458, + "grad_norm": 0.795675945576436, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 4458 + }, + { + "epoch": 0.04459, + "grad_norm": 0.8944676165743426, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 4459 + }, + { + "epoch": 0.0446, + "grad_norm": 1.0309911114547456, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 4460 + }, + { + "epoch": 0.04461, + "grad_norm": 1.1633164535503087, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 4461 + }, + { + "epoch": 0.04462, + "grad_norm": 1.167208135838096, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 4462 + }, + { + "epoch": 0.04463, + "grad_norm": 1.183108803762208, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 4463 + }, + { + "epoch": 0.04464, + "grad_norm": 0.7854170437651848, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 4464 + }, + { + "epoch": 0.04465, + "grad_norm": 0.8018127380897019, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 4465 + }, + { + "epoch": 0.04466, + "grad_norm": 0.7724293795857595, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 4466 + }, + { + "epoch": 0.04467, + "grad_norm": 0.8927290074378922, + "learning_rate": 0.003, + "loss": 4.089, + "step": 4467 + }, + { + "epoch": 0.04468, + "grad_norm": 0.9529441187417813, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 4468 + }, + { + "epoch": 0.04469, + "grad_norm": 0.7635512052136587, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 4469 + }, + { + "epoch": 0.0447, + "grad_norm": 0.7111288881650668, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 4470 + }, + { + "epoch": 0.04471, + "grad_norm": 0.6842975948143225, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 4471 + }, + { + "epoch": 0.04472, + "grad_norm": 0.6721731379887363, + "learning_rate": 0.003, + "loss": 4.064, + "step": 4472 + }, + { + "epoch": 0.04473, + "grad_norm": 0.8269116687874477, + "learning_rate": 0.003, + "loss": 4.069, + "step": 4473 + }, + { + "epoch": 0.04474, + "grad_norm": 0.9372410627748204, + "learning_rate": 0.003, + "loss": 4.07, + "step": 4474 + }, + { + "epoch": 0.04475, + "grad_norm": 1.1298342557384584, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 4475 + }, + { + "epoch": 0.04476, + "grad_norm": 0.8988046555647741, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 4476 + }, + { + "epoch": 0.04477, + "grad_norm": 0.795000264212953, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 4477 + }, + { + "epoch": 0.04478, + "grad_norm": 0.8406418984097118, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 4478 + }, + { + "epoch": 0.04479, + "grad_norm": 0.9513259886842036, + "learning_rate": 0.003, + "loss": 4.067, + "step": 4479 + }, + { + "epoch": 0.0448, + "grad_norm": 1.0486642112843845, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 4480 + }, + { + "epoch": 0.04481, + "grad_norm": 1.0910923440009397, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 4481 + }, + { + "epoch": 0.04482, + "grad_norm": 0.9923003720594974, + "learning_rate": 0.003, + "loss": 4.102, + "step": 4482 + }, + { + "epoch": 0.04483, + "grad_norm": 0.9838181891048627, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 4483 + }, + { + "epoch": 0.04484, + "grad_norm": 0.8441164141274783, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 4484 + }, + { + "epoch": 0.04485, + "grad_norm": 0.8006899110895409, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 4485 + }, + { + "epoch": 0.04486, + "grad_norm": 0.8707281199253866, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 4486 + }, + { + "epoch": 0.04487, + "grad_norm": 0.8680647327916373, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 4487 + }, + { + "epoch": 0.04488, + "grad_norm": 0.8736811662681273, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 4488 + }, + { + "epoch": 0.04489, + "grad_norm": 0.8803676696260273, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 4489 + }, + { + "epoch": 0.0449, + "grad_norm": 0.9871186527446516, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 4490 + }, + { + "epoch": 0.04491, + "grad_norm": 1.0838956165086664, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 4491 + }, + { + "epoch": 0.04492, + "grad_norm": 1.0246550076014433, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 4492 + }, + { + "epoch": 0.04493, + "grad_norm": 1.119573546351604, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 4493 + }, + { + "epoch": 0.04494, + "grad_norm": 0.9017080906502098, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 4494 + }, + { + "epoch": 0.04495, + "grad_norm": 0.8048737469068201, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 4495 + }, + { + "epoch": 0.04496, + "grad_norm": 0.8207659865646967, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 4496 + }, + { + "epoch": 0.04497, + "grad_norm": 1.1245269296552654, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 4497 + }, + { + "epoch": 0.04498, + "grad_norm": 1.0002119651690387, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 4498 + }, + { + "epoch": 0.04499, + "grad_norm": 0.9808448366317947, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 4499 + }, + { + "epoch": 0.045, + "grad_norm": 1.132093441653165, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 4500 + }, + { + "epoch": 0.04501, + "grad_norm": 0.8718373099791662, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 4501 + }, + { + "epoch": 0.04502, + "grad_norm": 0.9228626903708667, + "learning_rate": 0.003, + "loss": 4.116, + "step": 4502 + }, + { + "epoch": 0.04503, + "grad_norm": 0.8263232740111328, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 4503 + }, + { + "epoch": 0.04504, + "grad_norm": 0.7734875104374885, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 4504 + }, + { + "epoch": 0.04505, + "grad_norm": 0.7885987258017754, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 4505 + }, + { + "epoch": 0.04506, + "grad_norm": 0.9407227338991642, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 4506 + }, + { + "epoch": 0.04507, + "grad_norm": 1.0832398695950372, + "learning_rate": 0.003, + "loss": 4.067, + "step": 4507 + }, + { + "epoch": 0.04508, + "grad_norm": 0.9418456435744189, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 4508 + }, + { + "epoch": 0.04509, + "grad_norm": 0.9175750111895549, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 4509 + }, + { + "epoch": 0.0451, + "grad_norm": 0.9398070401461067, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 4510 + }, + { + "epoch": 0.04511, + "grad_norm": 0.9217534503191119, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 4511 + }, + { + "epoch": 0.04512, + "grad_norm": 1.0427993141187701, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 4512 + }, + { + "epoch": 0.04513, + "grad_norm": 0.8428325423508638, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 4513 + }, + { + "epoch": 0.04514, + "grad_norm": 1.0089526014399501, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 4514 + }, + { + "epoch": 0.04515, + "grad_norm": 0.9789233122430068, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 4515 + }, + { + "epoch": 0.04516, + "grad_norm": 0.9091489043065707, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 4516 + }, + { + "epoch": 0.04517, + "grad_norm": 0.9751493943869993, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 4517 + }, + { + "epoch": 0.04518, + "grad_norm": 0.8917249628104398, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 4518 + }, + { + "epoch": 0.04519, + "grad_norm": 1.004402772465955, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 4519 + }, + { + "epoch": 0.0452, + "grad_norm": 1.2586466593520984, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 4520 + }, + { + "epoch": 0.04521, + "grad_norm": 0.9330729244910515, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 4521 + }, + { + "epoch": 0.04522, + "grad_norm": 0.8641019927023269, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 4522 + }, + { + "epoch": 0.04523, + "grad_norm": 0.9140037813457487, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 4523 + }, + { + "epoch": 0.04524, + "grad_norm": 0.8919093436216252, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 4524 + }, + { + "epoch": 0.04525, + "grad_norm": 1.0121384333442254, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 4525 + }, + { + "epoch": 0.04526, + "grad_norm": 1.1210191689178208, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 4526 + }, + { + "epoch": 0.04527, + "grad_norm": 0.682641155436752, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 4527 + }, + { + "epoch": 0.04528, + "grad_norm": 0.6585540834263701, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 4528 + }, + { + "epoch": 0.04529, + "grad_norm": 0.7678794616279656, + "learning_rate": 0.003, + "loss": 4.061, + "step": 4529 + }, + { + "epoch": 0.0453, + "grad_norm": 0.8419648665385003, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 4530 + }, + { + "epoch": 0.04531, + "grad_norm": 0.9205297479236283, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 4531 + }, + { + "epoch": 0.04532, + "grad_norm": 0.8853526961425466, + "learning_rate": 0.003, + "loss": 4.088, + "step": 4532 + }, + { + "epoch": 0.04533, + "grad_norm": 0.8992375264745746, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 4533 + }, + { + "epoch": 0.04534, + "grad_norm": 0.9811346625863182, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 4534 + }, + { + "epoch": 0.04535, + "grad_norm": 1.037436568279496, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 4535 + }, + { + "epoch": 0.04536, + "grad_norm": 0.8288199794219244, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 4536 + }, + { + "epoch": 0.04537, + "grad_norm": 0.8088616928239108, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 4537 + }, + { + "epoch": 0.04538, + "grad_norm": 0.7318781047913877, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 4538 + }, + { + "epoch": 0.04539, + "grad_norm": 0.7179435097962413, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 4539 + }, + { + "epoch": 0.0454, + "grad_norm": 0.8058129172445117, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 4540 + }, + { + "epoch": 0.04541, + "grad_norm": 0.9730325188123338, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 4541 + }, + { + "epoch": 0.04542, + "grad_norm": 1.1629088900069842, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 4542 + }, + { + "epoch": 0.04543, + "grad_norm": 1.0223952095742048, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 4543 + }, + { + "epoch": 0.04544, + "grad_norm": 0.9007892884805351, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 4544 + }, + { + "epoch": 0.04545, + "grad_norm": 0.8810595549441049, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 4545 + }, + { + "epoch": 0.04546, + "grad_norm": 1.0706790887149014, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 4546 + }, + { + "epoch": 0.04547, + "grad_norm": 1.0783793172148153, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 4547 + }, + { + "epoch": 0.04548, + "grad_norm": 0.8754970851766884, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 4548 + }, + { + "epoch": 0.04549, + "grad_norm": 1.006047833218977, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 4549 + }, + { + "epoch": 0.0455, + "grad_norm": 1.0770476162131184, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 4550 + }, + { + "epoch": 0.04551, + "grad_norm": 0.9601215888010274, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 4551 + }, + { + "epoch": 0.04552, + "grad_norm": 0.9774413017455548, + "learning_rate": 0.003, + "loss": 4.111, + "step": 4552 + }, + { + "epoch": 0.04553, + "grad_norm": 1.0213364834049192, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 4553 + }, + { + "epoch": 0.04554, + "grad_norm": 0.9568143904158197, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 4554 + }, + { + "epoch": 0.04555, + "grad_norm": 0.7973060747821642, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 4555 + }, + { + "epoch": 0.04556, + "grad_norm": 0.9299995488489511, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 4556 + }, + { + "epoch": 0.04557, + "grad_norm": 1.1293974930486246, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 4557 + }, + { + "epoch": 0.04558, + "grad_norm": 1.026675541394777, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 4558 + }, + { + "epoch": 0.04559, + "grad_norm": 1.0345981426986643, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 4559 + }, + { + "epoch": 0.0456, + "grad_norm": 1.1305484499753016, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 4560 + }, + { + "epoch": 0.04561, + "grad_norm": 0.925988269592052, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 4561 + }, + { + "epoch": 0.04562, + "grad_norm": 0.8510922703082149, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 4562 + }, + { + "epoch": 0.04563, + "grad_norm": 0.9226408925609122, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 4563 + }, + { + "epoch": 0.04564, + "grad_norm": 1.0130794105331695, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 4564 + }, + { + "epoch": 0.04565, + "grad_norm": 1.1969435548126093, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 4565 + }, + { + "epoch": 0.04566, + "grad_norm": 0.9275288015749137, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 4566 + }, + { + "epoch": 0.04567, + "grad_norm": 0.8651130584077846, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 4567 + }, + { + "epoch": 0.04568, + "grad_norm": 0.8986618248546338, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 4568 + }, + { + "epoch": 0.04569, + "grad_norm": 0.6837244864462902, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4569 + }, + { + "epoch": 0.0457, + "grad_norm": 0.6768551414670718, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 4570 + }, + { + "epoch": 0.04571, + "grad_norm": 0.7428397016306537, + "learning_rate": 0.003, + "loss": 4.107, + "step": 4571 + }, + { + "epoch": 0.04572, + "grad_norm": 0.8995755995048413, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 4572 + }, + { + "epoch": 0.04573, + "grad_norm": 0.9009642352606402, + "learning_rate": 0.003, + "loss": 4.058, + "step": 4573 + }, + { + "epoch": 0.04574, + "grad_norm": 0.7114211134538346, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 4574 + }, + { + "epoch": 0.04575, + "grad_norm": 0.6599729401731194, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 4575 + }, + { + "epoch": 0.04576, + "grad_norm": 0.7597125240743299, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 4576 + }, + { + "epoch": 0.04577, + "grad_norm": 0.8421667189567135, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 4577 + }, + { + "epoch": 0.04578, + "grad_norm": 0.8104388984795634, + "learning_rate": 0.003, + "loss": 4.066, + "step": 4578 + }, + { + "epoch": 0.04579, + "grad_norm": 0.8976119076785662, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 4579 + }, + { + "epoch": 0.0458, + "grad_norm": 0.9887238152154557, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4580 + }, + { + "epoch": 0.04581, + "grad_norm": 1.0944325064747162, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 4581 + }, + { + "epoch": 0.04582, + "grad_norm": 1.1228836648844638, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 4582 + }, + { + "epoch": 0.04583, + "grad_norm": 1.037982851445284, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 4583 + }, + { + "epoch": 0.04584, + "grad_norm": 1.1309942180537078, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 4584 + }, + { + "epoch": 0.04585, + "grad_norm": 0.8719784933921241, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 4585 + }, + { + "epoch": 0.04586, + "grad_norm": 0.9988395861483548, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 4586 + }, + { + "epoch": 0.04587, + "grad_norm": 1.1757119262154812, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 4587 + }, + { + "epoch": 0.04588, + "grad_norm": 0.9935517321787154, + "learning_rate": 0.003, + "loss": 4.106, + "step": 4588 + }, + { + "epoch": 0.04589, + "grad_norm": 0.9699416826899608, + "learning_rate": 0.003, + "loss": 4.073, + "step": 4589 + }, + { + "epoch": 0.0459, + "grad_norm": 0.8608811328558719, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 4590 + }, + { + "epoch": 0.04591, + "grad_norm": 0.8189922265739307, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 4591 + }, + { + "epoch": 0.04592, + "grad_norm": 0.7122732264854349, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 4592 + }, + { + "epoch": 0.04593, + "grad_norm": 0.6916533373498136, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 4593 + }, + { + "epoch": 0.04594, + "grad_norm": 0.7141399590985198, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 4594 + }, + { + "epoch": 0.04595, + "grad_norm": 0.5874937539781525, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 4595 + }, + { + "epoch": 0.04596, + "grad_norm": 0.5901031684112441, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4596 + }, + { + "epoch": 0.04597, + "grad_norm": 0.6145935292761844, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 4597 + }, + { + "epoch": 0.04598, + "grad_norm": 0.6900732502328394, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 4598 + }, + { + "epoch": 0.04599, + "grad_norm": 0.7532465802179442, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 4599 + }, + { + "epoch": 0.046, + "grad_norm": 0.7204851942827531, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 4600 + }, + { + "epoch": 0.04601, + "grad_norm": 0.8000251382960614, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 4601 + }, + { + "epoch": 0.04602, + "grad_norm": 0.9181816267947781, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 4602 + }, + { + "epoch": 0.04603, + "grad_norm": 1.1329954986924105, + "learning_rate": 0.003, + "loss": 4.1, + "step": 4603 + }, + { + "epoch": 0.04604, + "grad_norm": 0.9229244630258843, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 4604 + }, + { + "epoch": 0.04605, + "grad_norm": 0.9397750901932713, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 4605 + }, + { + "epoch": 0.04606, + "grad_norm": 0.9807853977592482, + "learning_rate": 0.003, + "loss": 4.071, + "step": 4606 + }, + { + "epoch": 0.04607, + "grad_norm": 1.0004066586342284, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 4607 + }, + { + "epoch": 0.04608, + "grad_norm": 0.8365719223227974, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 4608 + }, + { + "epoch": 0.04609, + "grad_norm": 0.8004776907383365, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 4609 + }, + { + "epoch": 0.0461, + "grad_norm": 1.0497099318779148, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 4610 + }, + { + "epoch": 0.04611, + "grad_norm": 1.2600264680618676, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 4611 + }, + { + "epoch": 0.04612, + "grad_norm": 0.7894125524130393, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 4612 + }, + { + "epoch": 0.04613, + "grad_norm": 0.8137383483366754, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 4613 + }, + { + "epoch": 0.04614, + "grad_norm": 1.0780550037519088, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 4614 + }, + { + "epoch": 0.04615, + "grad_norm": 1.146445498524777, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 4615 + }, + { + "epoch": 0.04616, + "grad_norm": 1.0658805778199005, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 4616 + }, + { + "epoch": 0.04617, + "grad_norm": 1.0441620047388833, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 4617 + }, + { + "epoch": 0.04618, + "grad_norm": 0.8556036446407603, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 4618 + }, + { + "epoch": 0.04619, + "grad_norm": 0.8711936990908583, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 4619 + }, + { + "epoch": 0.0462, + "grad_norm": 0.9909063580253566, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 4620 + }, + { + "epoch": 0.04621, + "grad_norm": 1.0674313514316285, + "learning_rate": 0.003, + "loss": 4.118, + "step": 4621 + }, + { + "epoch": 0.04622, + "grad_norm": 0.9135451914376392, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 4622 + }, + { + "epoch": 0.04623, + "grad_norm": 0.9086448162432442, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4623 + }, + { + "epoch": 0.04624, + "grad_norm": 0.929118113001063, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4624 + }, + { + "epoch": 0.04625, + "grad_norm": 0.8923285451003576, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 4625 + }, + { + "epoch": 0.04626, + "grad_norm": 0.8550528854318953, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 4626 + }, + { + "epoch": 0.04627, + "grad_norm": 0.8738286232684426, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 4627 + }, + { + "epoch": 0.04628, + "grad_norm": 1.040219821170605, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 4628 + }, + { + "epoch": 0.04629, + "grad_norm": 1.0457108285440184, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 4629 + }, + { + "epoch": 0.0463, + "grad_norm": 1.0237213701996026, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 4630 + }, + { + "epoch": 0.04631, + "grad_norm": 0.903960943241525, + "learning_rate": 0.003, + "loss": 4.084, + "step": 4631 + }, + { + "epoch": 0.04632, + "grad_norm": 0.848296335406008, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 4632 + }, + { + "epoch": 0.04633, + "grad_norm": 0.9208369707573915, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 4633 + }, + { + "epoch": 0.04634, + "grad_norm": 0.9324467970591866, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 4634 + }, + { + "epoch": 0.04635, + "grad_norm": 1.1007094021276311, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 4635 + }, + { + "epoch": 0.04636, + "grad_norm": 1.1761724983902668, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 4636 + }, + { + "epoch": 0.04637, + "grad_norm": 0.8928241988529365, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 4637 + }, + { + "epoch": 0.04638, + "grad_norm": 0.7909525696684022, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 4638 + }, + { + "epoch": 0.04639, + "grad_norm": 0.8316115510623557, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 4639 + }, + { + "epoch": 0.0464, + "grad_norm": 0.8890413282477022, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 4640 + }, + { + "epoch": 0.04641, + "grad_norm": 0.9442456572250497, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 4641 + }, + { + "epoch": 0.04642, + "grad_norm": 1.1131605186896782, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 4642 + }, + { + "epoch": 0.04643, + "grad_norm": 1.151049967466462, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 4643 + }, + { + "epoch": 0.04644, + "grad_norm": 1.0077381059472987, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 4644 + }, + { + "epoch": 0.04645, + "grad_norm": 0.8760793972978849, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 4645 + }, + { + "epoch": 0.04646, + "grad_norm": 0.8512107506066473, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 4646 + }, + { + "epoch": 0.04647, + "grad_norm": 0.8732029277276215, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 4647 + }, + { + "epoch": 0.04648, + "grad_norm": 0.9302078959957535, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 4648 + }, + { + "epoch": 0.04649, + "grad_norm": 0.9169568810760613, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 4649 + }, + { + "epoch": 0.0465, + "grad_norm": 0.9303804967387479, + "learning_rate": 0.003, + "loss": 4.053, + "step": 4650 + }, + { + "epoch": 0.04651, + "grad_norm": 0.872676445058754, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 4651 + }, + { + "epoch": 0.04652, + "grad_norm": 0.8018180143670698, + "learning_rate": 0.003, + "loss": 4.09, + "step": 4652 + }, + { + "epoch": 0.04653, + "grad_norm": 0.8064464814963113, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 4653 + }, + { + "epoch": 0.04654, + "grad_norm": 0.8944906023744553, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 4654 + }, + { + "epoch": 0.04655, + "grad_norm": 1.1219617714586472, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 4655 + }, + { + "epoch": 0.04656, + "grad_norm": 1.0550396369369868, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 4656 + }, + { + "epoch": 0.04657, + "grad_norm": 1.0337811454324184, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 4657 + }, + { + "epoch": 0.04658, + "grad_norm": 0.9390842279569726, + "learning_rate": 0.003, + "loss": 4.083, + "step": 4658 + }, + { + "epoch": 0.04659, + "grad_norm": 0.7870888779187648, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 4659 + }, + { + "epoch": 0.0466, + "grad_norm": 0.7813047740300597, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 4660 + }, + { + "epoch": 0.04661, + "grad_norm": 0.8185250491523086, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 4661 + }, + { + "epoch": 0.04662, + "grad_norm": 0.7657744046442502, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 4662 + }, + { + "epoch": 0.04663, + "grad_norm": 0.9612044009806899, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 4663 + }, + { + "epoch": 0.04664, + "grad_norm": 1.091240235491955, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 4664 + }, + { + "epoch": 0.04665, + "grad_norm": 0.9496659607676349, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 4665 + }, + { + "epoch": 0.04666, + "grad_norm": 0.8885966480506947, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 4666 + }, + { + "epoch": 0.04667, + "grad_norm": 0.8146510839872884, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 4667 + }, + { + "epoch": 0.04668, + "grad_norm": 0.7003726363991867, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 4668 + }, + { + "epoch": 0.04669, + "grad_norm": 0.6213740348621104, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 4669 + }, + { + "epoch": 0.0467, + "grad_norm": 0.7033914078971897, + "learning_rate": 0.003, + "loss": 4.04, + "step": 4670 + }, + { + "epoch": 0.04671, + "grad_norm": 0.6956570433878502, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 4671 + }, + { + "epoch": 0.04672, + "grad_norm": 0.7790564265361141, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 4672 + }, + { + "epoch": 0.04673, + "grad_norm": 0.8582519469717382, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 4673 + }, + { + "epoch": 0.04674, + "grad_norm": 1.0257732379563222, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 4674 + }, + { + "epoch": 0.04675, + "grad_norm": 1.0228753347895319, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 4675 + }, + { + "epoch": 0.04676, + "grad_norm": 1.091796013099298, + "learning_rate": 0.003, + "loss": 4.096, + "step": 4676 + }, + { + "epoch": 0.04677, + "grad_norm": 0.9874564467619078, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 4677 + }, + { + "epoch": 0.04678, + "grad_norm": 0.9545956541830211, + "learning_rate": 0.003, + "loss": 4.055, + "step": 4678 + }, + { + "epoch": 0.04679, + "grad_norm": 0.9188392135996144, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 4679 + }, + { + "epoch": 0.0468, + "grad_norm": 0.9976774725576849, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 4680 + }, + { + "epoch": 0.04681, + "grad_norm": 1.2926044572885032, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 4681 + }, + { + "epoch": 0.04682, + "grad_norm": 0.8936432704892819, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4682 + }, + { + "epoch": 0.04683, + "grad_norm": 0.9387765394338359, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 4683 + }, + { + "epoch": 0.04684, + "grad_norm": 1.0010607126157993, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 4684 + }, + { + "epoch": 0.04685, + "grad_norm": 1.1642588881453941, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 4685 + }, + { + "epoch": 0.04686, + "grad_norm": 0.9177452996715223, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4686 + }, + { + "epoch": 0.04687, + "grad_norm": 1.0845507224363737, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 4687 + }, + { + "epoch": 0.04688, + "grad_norm": 1.0851979161146954, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 4688 + }, + { + "epoch": 0.04689, + "grad_norm": 1.0323716074650593, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 4689 + }, + { + "epoch": 0.0469, + "grad_norm": 0.8881158922609561, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 4690 + }, + { + "epoch": 0.04691, + "grad_norm": 0.7530974336635942, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 4691 + }, + { + "epoch": 0.04692, + "grad_norm": 0.8251300897824291, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 4692 + }, + { + "epoch": 0.04693, + "grad_norm": 1.1080787687675981, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 4693 + }, + { + "epoch": 0.04694, + "grad_norm": 1.2008844146719992, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 4694 + }, + { + "epoch": 0.04695, + "grad_norm": 0.8692807098125874, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 4695 + }, + { + "epoch": 0.04696, + "grad_norm": 0.7273592465913997, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 4696 + }, + { + "epoch": 0.04697, + "grad_norm": 0.791219049234071, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 4697 + }, + { + "epoch": 0.04698, + "grad_norm": 0.8259619113948167, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 4698 + }, + { + "epoch": 0.04699, + "grad_norm": 0.986873640288411, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 4699 + }, + { + "epoch": 0.047, + "grad_norm": 1.124829270421113, + "learning_rate": 0.003, + "loss": 4.103, + "step": 4700 + }, + { + "epoch": 0.04701, + "grad_norm": 0.8041528371793156, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 4701 + }, + { + "epoch": 0.04702, + "grad_norm": 0.9002565040000517, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 4702 + }, + { + "epoch": 0.04703, + "grad_norm": 0.928884750922695, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 4703 + }, + { + "epoch": 0.04704, + "grad_norm": 0.9517910705080795, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 4704 + }, + { + "epoch": 0.04705, + "grad_norm": 1.0793630137006167, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 4705 + }, + { + "epoch": 0.04706, + "grad_norm": 0.8932410896556339, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 4706 + }, + { + "epoch": 0.04707, + "grad_norm": 0.8273663973299694, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 4707 + }, + { + "epoch": 0.04708, + "grad_norm": 0.7707848347191034, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 4708 + }, + { + "epoch": 0.04709, + "grad_norm": 0.6881441521153875, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 4709 + }, + { + "epoch": 0.0471, + "grad_norm": 0.714094436830912, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 4710 + }, + { + "epoch": 0.04711, + "grad_norm": 0.8507678879908461, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 4711 + }, + { + "epoch": 0.04712, + "grad_norm": 0.8931516416205477, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 4712 + }, + { + "epoch": 0.04713, + "grad_norm": 0.9279444773212077, + "learning_rate": 0.003, + "loss": 4.07, + "step": 4713 + }, + { + "epoch": 0.04714, + "grad_norm": 0.7844617012393187, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 4714 + }, + { + "epoch": 0.04715, + "grad_norm": 0.7834150113825615, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 4715 + }, + { + "epoch": 0.04716, + "grad_norm": 0.7743590099580365, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 4716 + }, + { + "epoch": 0.04717, + "grad_norm": 0.7709401253853543, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 4717 + }, + { + "epoch": 0.04718, + "grad_norm": 0.8235615104288091, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 4718 + }, + { + "epoch": 0.04719, + "grad_norm": 0.8708271313882437, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 4719 + }, + { + "epoch": 0.0472, + "grad_norm": 0.8040720392649453, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 4720 + }, + { + "epoch": 0.04721, + "grad_norm": 0.7637937594189287, + "learning_rate": 0.003, + "loss": 4.054, + "step": 4721 + }, + { + "epoch": 0.04722, + "grad_norm": 0.7154347513877589, + "learning_rate": 0.003, + "loss": 4.069, + "step": 4722 + }, + { + "epoch": 0.04723, + "grad_norm": 0.8612068842939984, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 4723 + }, + { + "epoch": 0.04724, + "grad_norm": 0.9614536713142897, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 4724 + }, + { + "epoch": 0.04725, + "grad_norm": 1.0189566497638871, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 4725 + }, + { + "epoch": 0.04726, + "grad_norm": 1.0839491352411363, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 4726 + }, + { + "epoch": 0.04727, + "grad_norm": 1.0649373566906404, + "learning_rate": 0.003, + "loss": 4.064, + "step": 4727 + }, + { + "epoch": 0.04728, + "grad_norm": 0.867382918842643, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 4728 + }, + { + "epoch": 0.04729, + "grad_norm": 0.8648790898500417, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 4729 + }, + { + "epoch": 0.0473, + "grad_norm": 0.8269978275529186, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 4730 + }, + { + "epoch": 0.04731, + "grad_norm": 0.8941716511884031, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 4731 + }, + { + "epoch": 0.04732, + "grad_norm": 0.9280984555504275, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 4732 + }, + { + "epoch": 0.04733, + "grad_norm": 1.147397771954512, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 4733 + }, + { + "epoch": 0.04734, + "grad_norm": 0.9874050482010743, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 4734 + }, + { + "epoch": 0.04735, + "grad_norm": 1.2239662812431078, + "learning_rate": 0.003, + "loss": 4.114, + "step": 4735 + }, + { + "epoch": 0.04736, + "grad_norm": 0.9512556578156451, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 4736 + }, + { + "epoch": 0.04737, + "grad_norm": 1.1383559521224216, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 4737 + }, + { + "epoch": 0.04738, + "grad_norm": 0.9981707221550351, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 4738 + }, + { + "epoch": 0.04739, + "grad_norm": 0.9631032735800266, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 4739 + }, + { + "epoch": 0.0474, + "grad_norm": 0.8943309500832445, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 4740 + }, + { + "epoch": 0.04741, + "grad_norm": 0.9142912447416025, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 4741 + }, + { + "epoch": 0.04742, + "grad_norm": 0.9095935688180709, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 4742 + }, + { + "epoch": 0.04743, + "grad_norm": 0.9621787693657161, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 4743 + }, + { + "epoch": 0.04744, + "grad_norm": 0.983730502100782, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 4744 + }, + { + "epoch": 0.04745, + "grad_norm": 1.0519698186730737, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 4745 + }, + { + "epoch": 0.04746, + "grad_norm": 1.0692770481585765, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 4746 + }, + { + "epoch": 0.04747, + "grad_norm": 0.9800763430595273, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 4747 + }, + { + "epoch": 0.04748, + "grad_norm": 1.0447027226711574, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 4748 + }, + { + "epoch": 0.04749, + "grad_norm": 0.9780745361519866, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 4749 + }, + { + "epoch": 0.0475, + "grad_norm": 1.1823864036419054, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 4750 + }, + { + "epoch": 0.04751, + "grad_norm": 1.0317082296257587, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 4751 + }, + { + "epoch": 0.04752, + "grad_norm": 1.0191266508391366, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 4752 + }, + { + "epoch": 0.04753, + "grad_norm": 0.9650995183204055, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 4753 + }, + { + "epoch": 0.04754, + "grad_norm": 0.9626529774574477, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 4754 + }, + { + "epoch": 0.04755, + "grad_norm": 1.011836432397673, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 4755 + }, + { + "epoch": 0.04756, + "grad_norm": 0.9061088923081047, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4756 + }, + { + "epoch": 0.04757, + "grad_norm": 0.8653623623227834, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 4757 + }, + { + "epoch": 0.04758, + "grad_norm": 0.8875207606926206, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 4758 + }, + { + "epoch": 0.04759, + "grad_norm": 0.8817482562664115, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 4759 + }, + { + "epoch": 0.0476, + "grad_norm": 0.7350149992314764, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 4760 + }, + { + "epoch": 0.04761, + "grad_norm": 0.6886778817907339, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 4761 + }, + { + "epoch": 0.04762, + "grad_norm": 0.6706585457203895, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 4762 + }, + { + "epoch": 0.04763, + "grad_norm": 0.7887867887553318, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 4763 + }, + { + "epoch": 0.04764, + "grad_norm": 0.97596832551055, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 4764 + }, + { + "epoch": 0.04765, + "grad_norm": 1.2783386527166918, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 4765 + }, + { + "epoch": 0.04766, + "grad_norm": 0.87011816316084, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 4766 + }, + { + "epoch": 0.04767, + "grad_norm": 0.995725590420577, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 4767 + }, + { + "epoch": 0.04768, + "grad_norm": 0.9157007548076511, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 4768 + }, + { + "epoch": 0.04769, + "grad_norm": 0.7763424443251495, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 4769 + }, + { + "epoch": 0.0477, + "grad_norm": 0.9132161182733682, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 4770 + }, + { + "epoch": 0.04771, + "grad_norm": 1.0690816936012133, + "learning_rate": 0.003, + "loss": 4.081, + "step": 4771 + }, + { + "epoch": 0.04772, + "grad_norm": 1.0183011941792959, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 4772 + }, + { + "epoch": 0.04773, + "grad_norm": 0.9142174256967822, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 4773 + }, + { + "epoch": 0.04774, + "grad_norm": 0.7603001306934604, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 4774 + }, + { + "epoch": 0.04775, + "grad_norm": 0.8346482465320779, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 4775 + }, + { + "epoch": 0.04776, + "grad_norm": 0.9773364899564101, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 4776 + }, + { + "epoch": 0.04777, + "grad_norm": 1.0784655700916257, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 4777 + }, + { + "epoch": 0.04778, + "grad_norm": 0.9412592454712865, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 4778 + }, + { + "epoch": 0.04779, + "grad_norm": 1.0145035098418551, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 4779 + }, + { + "epoch": 0.0478, + "grad_norm": 1.0762920782151484, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 4780 + }, + { + "epoch": 0.04781, + "grad_norm": 0.8651372390591274, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 4781 + }, + { + "epoch": 0.04782, + "grad_norm": 0.8769680695281393, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 4782 + }, + { + "epoch": 0.04783, + "grad_norm": 0.9787359970897872, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 4783 + }, + { + "epoch": 0.04784, + "grad_norm": 0.9721858465952792, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 4784 + }, + { + "epoch": 0.04785, + "grad_norm": 1.0635527582181876, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4785 + }, + { + "epoch": 0.04786, + "grad_norm": 0.9396709467834784, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 4786 + }, + { + "epoch": 0.04787, + "grad_norm": 1.00148310702931, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 4787 + }, + { + "epoch": 0.04788, + "grad_norm": 0.8743917541049051, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 4788 + }, + { + "epoch": 0.04789, + "grad_norm": 0.7884230141024988, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 4789 + }, + { + "epoch": 0.0479, + "grad_norm": 0.8498046311242973, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 4790 + }, + { + "epoch": 0.04791, + "grad_norm": 0.956319683384673, + "learning_rate": 0.003, + "loss": 4.095, + "step": 4791 + }, + { + "epoch": 0.04792, + "grad_norm": 1.228255623374456, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 4792 + }, + { + "epoch": 0.04793, + "grad_norm": 0.9294780478699636, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 4793 + }, + { + "epoch": 0.04794, + "grad_norm": 0.9672358732016223, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 4794 + }, + { + "epoch": 0.04795, + "grad_norm": 1.0525050179084454, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 4795 + }, + { + "epoch": 0.04796, + "grad_norm": 1.1544771283976443, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 4796 + }, + { + "epoch": 0.04797, + "grad_norm": 0.915501476651925, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 4797 + }, + { + "epoch": 0.04798, + "grad_norm": 0.8655114100315004, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 4798 + }, + { + "epoch": 0.04799, + "grad_norm": 0.9840921488311046, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 4799 + }, + { + "epoch": 0.048, + "grad_norm": 0.9567911011323519, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 4800 + }, + { + "epoch": 0.04801, + "grad_norm": 1.0260566498791253, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 4801 + }, + { + "epoch": 0.04802, + "grad_norm": 0.9765671371087857, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 4802 + }, + { + "epoch": 0.04803, + "grad_norm": 0.8954590032392303, + "learning_rate": 0.003, + "loss": 4.067, + "step": 4803 + }, + { + "epoch": 0.04804, + "grad_norm": 0.9397394535107887, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 4804 + }, + { + "epoch": 0.04805, + "grad_norm": 0.8330259017268677, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 4805 + }, + { + "epoch": 0.04806, + "grad_norm": 0.8897845335382429, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 4806 + }, + { + "epoch": 0.04807, + "grad_norm": 0.8951815464379674, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 4807 + }, + { + "epoch": 0.04808, + "grad_norm": 0.8091055618521581, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 4808 + }, + { + "epoch": 0.04809, + "grad_norm": 0.90919679772142, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 4809 + }, + { + "epoch": 0.0481, + "grad_norm": 1.0315510358211395, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 4810 + }, + { + "epoch": 0.04811, + "grad_norm": 1.0721233459919157, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 4811 + }, + { + "epoch": 0.04812, + "grad_norm": 1.0618859760248345, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 4812 + }, + { + "epoch": 0.04813, + "grad_norm": 0.9313083894483718, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4813 + }, + { + "epoch": 0.04814, + "grad_norm": 0.8584027305263373, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 4814 + }, + { + "epoch": 0.04815, + "grad_norm": 0.8633671828950363, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 4815 + }, + { + "epoch": 0.04816, + "grad_norm": 1.0445614739609703, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 4816 + }, + { + "epoch": 0.04817, + "grad_norm": 1.156557836130519, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 4817 + }, + { + "epoch": 0.04818, + "grad_norm": 0.9056636167430384, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4818 + }, + { + "epoch": 0.04819, + "grad_norm": 0.9644804670241944, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 4819 + }, + { + "epoch": 0.0482, + "grad_norm": 0.8579108360881226, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 4820 + }, + { + "epoch": 0.04821, + "grad_norm": 0.738347738618665, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 4821 + }, + { + "epoch": 0.04822, + "grad_norm": 0.77578338268237, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 4822 + }, + { + "epoch": 0.04823, + "grad_norm": 0.7660224855432475, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 4823 + }, + { + "epoch": 0.04824, + "grad_norm": 0.7699061557992386, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 4824 + }, + { + "epoch": 0.04825, + "grad_norm": 0.9058101965573263, + "learning_rate": 0.003, + "loss": 4.072, + "step": 4825 + }, + { + "epoch": 0.04826, + "grad_norm": 1.1063436218286011, + "learning_rate": 0.003, + "loss": 4.083, + "step": 4826 + }, + { + "epoch": 0.04827, + "grad_norm": 0.8962605854230049, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 4827 + }, + { + "epoch": 0.04828, + "grad_norm": 0.9401425903734211, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 4828 + }, + { + "epoch": 0.04829, + "grad_norm": 1.008827045223568, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 4829 + }, + { + "epoch": 0.0483, + "grad_norm": 1.1759354719434816, + "learning_rate": 0.003, + "loss": 4.094, + "step": 4830 + }, + { + "epoch": 0.04831, + "grad_norm": 0.9149475371771189, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 4831 + }, + { + "epoch": 0.04832, + "grad_norm": 0.9417806047622538, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 4832 + }, + { + "epoch": 0.04833, + "grad_norm": 0.8568983398054711, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 4833 + }, + { + "epoch": 0.04834, + "grad_norm": 0.9118080863225161, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 4834 + }, + { + "epoch": 0.04835, + "grad_norm": 0.8727113022874836, + "learning_rate": 0.003, + "loss": 4.091, + "step": 4835 + }, + { + "epoch": 0.04836, + "grad_norm": 0.7282176510987858, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 4836 + }, + { + "epoch": 0.04837, + "grad_norm": 0.8743286431817264, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 4837 + }, + { + "epoch": 0.04838, + "grad_norm": 1.0886280382022706, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 4838 + }, + { + "epoch": 0.04839, + "grad_norm": 0.9289173758415538, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 4839 + }, + { + "epoch": 0.0484, + "grad_norm": 0.9689477229941422, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 4840 + }, + { + "epoch": 0.04841, + "grad_norm": 0.8968968888481653, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 4841 + }, + { + "epoch": 0.04842, + "grad_norm": 0.7785478768547701, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 4842 + }, + { + "epoch": 0.04843, + "grad_norm": 0.7603045256063694, + "learning_rate": 0.003, + "loss": 4.048, + "step": 4843 + }, + { + "epoch": 0.04844, + "grad_norm": 0.8780025270605276, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 4844 + }, + { + "epoch": 0.04845, + "grad_norm": 0.9254852484657723, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 4845 + }, + { + "epoch": 0.04846, + "grad_norm": 0.8987509251388802, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 4846 + }, + { + "epoch": 0.04847, + "grad_norm": 0.8893955297373247, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 4847 + }, + { + "epoch": 0.04848, + "grad_norm": 0.8912302394312529, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 4848 + }, + { + "epoch": 0.04849, + "grad_norm": 0.8726877737099056, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 4849 + }, + { + "epoch": 0.0485, + "grad_norm": 0.8314222830339668, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 4850 + }, + { + "epoch": 0.04851, + "grad_norm": 0.7822732329238694, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 4851 + }, + { + "epoch": 0.04852, + "grad_norm": 0.9092871934087126, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 4852 + }, + { + "epoch": 0.04853, + "grad_norm": 1.2348482197912367, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 4853 + }, + { + "epoch": 0.04854, + "grad_norm": 0.8979894181749922, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 4854 + }, + { + "epoch": 0.04855, + "grad_norm": 0.8934339928496606, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 4855 + }, + { + "epoch": 0.04856, + "grad_norm": 1.0136848244881436, + "learning_rate": 0.003, + "loss": 4.095, + "step": 4856 + }, + { + "epoch": 0.04857, + "grad_norm": 1.2509395241639578, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 4857 + }, + { + "epoch": 0.04858, + "grad_norm": 0.9372409373437549, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 4858 + }, + { + "epoch": 0.04859, + "grad_norm": 0.9546537840113466, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 4859 + }, + { + "epoch": 0.0486, + "grad_norm": 1.082853408289471, + "learning_rate": 0.003, + "loss": 4.093, + "step": 4860 + }, + { + "epoch": 0.04861, + "grad_norm": 1.01675025574686, + "learning_rate": 0.003, + "loss": 4.086, + "step": 4861 + }, + { + "epoch": 0.04862, + "grad_norm": 0.9176376091053398, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 4862 + }, + { + "epoch": 0.04863, + "grad_norm": 0.8568974528178754, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 4863 + }, + { + "epoch": 0.04864, + "grad_norm": 0.8166420481645057, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 4864 + }, + { + "epoch": 0.04865, + "grad_norm": 0.9234428400906274, + "learning_rate": 0.003, + "loss": 4.098, + "step": 4865 + }, + { + "epoch": 0.04866, + "grad_norm": 0.8738302868938767, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4866 + }, + { + "epoch": 0.04867, + "grad_norm": 0.85532860997535, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 4867 + }, + { + "epoch": 0.04868, + "grad_norm": 0.7330290515083526, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 4868 + }, + { + "epoch": 0.04869, + "grad_norm": 0.7237705668329806, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 4869 + }, + { + "epoch": 0.0487, + "grad_norm": 0.7073551331896761, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 4870 + }, + { + "epoch": 0.04871, + "grad_norm": 0.7482228718063552, + "learning_rate": 0.003, + "loss": 4.094, + "step": 4871 + }, + { + "epoch": 0.04872, + "grad_norm": 0.7934599742401708, + "learning_rate": 0.003, + "loss": 4.06, + "step": 4872 + }, + { + "epoch": 0.04873, + "grad_norm": 0.8520994037973396, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 4873 + }, + { + "epoch": 0.04874, + "grad_norm": 0.9436390490861155, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 4874 + }, + { + "epoch": 0.04875, + "grad_norm": 1.3353781337660462, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 4875 + }, + { + "epoch": 0.04876, + "grad_norm": 0.8304614123526641, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 4876 + }, + { + "epoch": 0.04877, + "grad_norm": 0.7707719582887104, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 4877 + }, + { + "epoch": 0.04878, + "grad_norm": 0.9308445091065936, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 4878 + }, + { + "epoch": 0.04879, + "grad_norm": 0.9815137055251556, + "learning_rate": 0.003, + "loss": 4.057, + "step": 4879 + }, + { + "epoch": 0.0488, + "grad_norm": 1.3666093708712401, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4880 + }, + { + "epoch": 0.04881, + "grad_norm": 0.8702473732106857, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 4881 + }, + { + "epoch": 0.04882, + "grad_norm": 0.879457851478898, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4882 + }, + { + "epoch": 0.04883, + "grad_norm": 0.8303309342477992, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 4883 + }, + { + "epoch": 0.04884, + "grad_norm": 0.8052075077532388, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 4884 + }, + { + "epoch": 0.04885, + "grad_norm": 0.8973915391486328, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 4885 + }, + { + "epoch": 0.04886, + "grad_norm": 1.0019930120717506, + "learning_rate": 0.003, + "loss": 4.078, + "step": 4886 + }, + { + "epoch": 0.04887, + "grad_norm": 1.347620612956241, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 4887 + }, + { + "epoch": 0.04888, + "grad_norm": 0.8778759238009314, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 4888 + }, + { + "epoch": 0.04889, + "grad_norm": 1.0906727199530957, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 4889 + }, + { + "epoch": 0.0489, + "grad_norm": 0.9502197503744106, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 4890 + }, + { + "epoch": 0.04891, + "grad_norm": 0.9462880674223806, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 4891 + }, + { + "epoch": 0.04892, + "grad_norm": 0.9960546118887255, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 4892 + }, + { + "epoch": 0.04893, + "grad_norm": 0.9133067753769846, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 4893 + }, + { + "epoch": 0.04894, + "grad_norm": 0.910365622657426, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 4894 + }, + { + "epoch": 0.04895, + "grad_norm": 1.0152626137013607, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 4895 + }, + { + "epoch": 0.04896, + "grad_norm": 0.9625608021661926, + "learning_rate": 0.003, + "loss": 4.096, + "step": 4896 + }, + { + "epoch": 0.04897, + "grad_norm": 0.9603064678490546, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 4897 + }, + { + "epoch": 0.04898, + "grad_norm": 1.0125115002322624, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 4898 + }, + { + "epoch": 0.04899, + "grad_norm": 1.0091142549096461, + "learning_rate": 0.003, + "loss": 4.079, + "step": 4899 + }, + { + "epoch": 0.049, + "grad_norm": 1.1619884083457441, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 4900 + }, + { + "epoch": 0.04901, + "grad_norm": 0.8775379815040303, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 4901 + }, + { + "epoch": 0.04902, + "grad_norm": 0.793220630533508, + "learning_rate": 0.003, + "loss": 4.095, + "step": 4902 + }, + { + "epoch": 0.04903, + "grad_norm": 0.8017930562442611, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4903 + }, + { + "epoch": 0.04904, + "grad_norm": 0.8507583142575733, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 4904 + }, + { + "epoch": 0.04905, + "grad_norm": 0.9366175720443591, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 4905 + }, + { + "epoch": 0.04906, + "grad_norm": 1.0894519512603444, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 4906 + }, + { + "epoch": 0.04907, + "grad_norm": 0.9566147882681758, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 4907 + }, + { + "epoch": 0.04908, + "grad_norm": 0.8813724661117732, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 4908 + }, + { + "epoch": 0.04909, + "grad_norm": 0.7697604768117718, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 4909 + }, + { + "epoch": 0.0491, + "grad_norm": 0.9551094520562223, + "learning_rate": 0.003, + "loss": 4.071, + "step": 4910 + }, + { + "epoch": 0.04911, + "grad_norm": 1.5155833386194069, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 4911 + }, + { + "epoch": 0.04912, + "grad_norm": 0.9025182185743093, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 4912 + }, + { + "epoch": 0.04913, + "grad_norm": 0.7963791253987615, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 4913 + }, + { + "epoch": 0.04914, + "grad_norm": 0.7975682531864906, + "learning_rate": 0.003, + "loss": 4.074, + "step": 4914 + }, + { + "epoch": 0.04915, + "grad_norm": 0.7680989535713021, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 4915 + }, + { + "epoch": 0.04916, + "grad_norm": 0.8541651719195734, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 4916 + }, + { + "epoch": 0.04917, + "grad_norm": 0.9306809194401052, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 4917 + }, + { + "epoch": 0.04918, + "grad_norm": 1.0201811580852054, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 4918 + }, + { + "epoch": 0.04919, + "grad_norm": 0.9164268845754078, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 4919 + }, + { + "epoch": 0.0492, + "grad_norm": 0.8305372570501433, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 4920 + }, + { + "epoch": 0.04921, + "grad_norm": 0.7707198235784101, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 4921 + }, + { + "epoch": 0.04922, + "grad_norm": 0.7928692496541254, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 4922 + }, + { + "epoch": 0.04923, + "grad_norm": 0.7521164440028039, + "learning_rate": 0.003, + "loss": 4.045, + "step": 4923 + }, + { + "epoch": 0.04924, + "grad_norm": 0.8188114735033863, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 4924 + }, + { + "epoch": 0.04925, + "grad_norm": 0.8564277609742044, + "learning_rate": 0.003, + "loss": 4.051, + "step": 4925 + }, + { + "epoch": 0.04926, + "grad_norm": 0.8486097281410112, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 4926 + }, + { + "epoch": 0.04927, + "grad_norm": 0.889369882812886, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 4927 + }, + { + "epoch": 0.04928, + "grad_norm": 1.040367513666876, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 4928 + }, + { + "epoch": 0.04929, + "grad_norm": 1.1046422380445349, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 4929 + }, + { + "epoch": 0.0493, + "grad_norm": 0.8664686841477672, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 4930 + }, + { + "epoch": 0.04931, + "grad_norm": 0.9761798460720659, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 4931 + }, + { + "epoch": 0.04932, + "grad_norm": 1.0969681843858505, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 4932 + }, + { + "epoch": 0.04933, + "grad_norm": 0.8634620800542036, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 4933 + }, + { + "epoch": 0.04934, + "grad_norm": 1.0167878960199042, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 4934 + }, + { + "epoch": 0.04935, + "grad_norm": 1.2105132259455145, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 4935 + }, + { + "epoch": 0.04936, + "grad_norm": 0.8823692534511746, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 4936 + }, + { + "epoch": 0.04937, + "grad_norm": 1.0035117633460957, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 4937 + }, + { + "epoch": 0.04938, + "grad_norm": 1.21119716146334, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 4938 + }, + { + "epoch": 0.04939, + "grad_norm": 0.9542017309667814, + "learning_rate": 0.003, + "loss": 4.093, + "step": 4939 + }, + { + "epoch": 0.0494, + "grad_norm": 0.9693778311510872, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 4940 + }, + { + "epoch": 0.04941, + "grad_norm": 1.1934773830149854, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 4941 + }, + { + "epoch": 0.04942, + "grad_norm": 0.9941479800765747, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 4942 + }, + { + "epoch": 0.04943, + "grad_norm": 1.2980963908573873, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 4943 + }, + { + "epoch": 0.04944, + "grad_norm": 0.7685905125841351, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 4944 + }, + { + "epoch": 0.04945, + "grad_norm": 0.77621337036946, + "learning_rate": 0.003, + "loss": 4.094, + "step": 4945 + }, + { + "epoch": 0.04946, + "grad_norm": 0.7935358182030233, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 4946 + }, + { + "epoch": 0.04947, + "grad_norm": 0.95777398203718, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 4947 + }, + { + "epoch": 0.04948, + "grad_norm": 1.0505854626251887, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 4948 + }, + { + "epoch": 0.04949, + "grad_norm": 0.9843835143059065, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 4949 + }, + { + "epoch": 0.0495, + "grad_norm": 0.9820238950682217, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 4950 + }, + { + "epoch": 0.04951, + "grad_norm": 1.0176777494089932, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 4951 + }, + { + "epoch": 0.04952, + "grad_norm": 0.8631784047281243, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 4952 + }, + { + "epoch": 0.04953, + "grad_norm": 0.8078946550955691, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 4953 + }, + { + "epoch": 0.04954, + "grad_norm": 0.8107838454898298, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 4954 + }, + { + "epoch": 0.04955, + "grad_norm": 0.6872102056615723, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 4955 + }, + { + "epoch": 0.04956, + "grad_norm": 0.7000353354694138, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 4956 + }, + { + "epoch": 0.04957, + "grad_norm": 0.768512912596234, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 4957 + }, + { + "epoch": 0.04958, + "grad_norm": 0.773414971171959, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 4958 + }, + { + "epoch": 0.04959, + "grad_norm": 0.6897656034365274, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 4959 + }, + { + "epoch": 0.0496, + "grad_norm": 0.5963565715617187, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 4960 + }, + { + "epoch": 0.04961, + "grad_norm": 0.6834698784405567, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 4961 + }, + { + "epoch": 0.04962, + "grad_norm": 0.7972061518569966, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 4962 + }, + { + "epoch": 0.04963, + "grad_norm": 0.8471174104945988, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 4963 + }, + { + "epoch": 0.04964, + "grad_norm": 0.8890963190037607, + "learning_rate": 0.003, + "loss": 4.079, + "step": 4964 + }, + { + "epoch": 0.04965, + "grad_norm": 0.9621708392835049, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 4965 + }, + { + "epoch": 0.04966, + "grad_norm": 1.2746050729963243, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 4966 + }, + { + "epoch": 0.04967, + "grad_norm": 0.9487241879206226, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 4967 + }, + { + "epoch": 0.04968, + "grad_norm": 1.0115561803434678, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 4968 + }, + { + "epoch": 0.04969, + "grad_norm": 1.0743944507868857, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 4969 + }, + { + "epoch": 0.0497, + "grad_norm": 0.9957497513808655, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 4970 + }, + { + "epoch": 0.04971, + "grad_norm": 0.9301130535679807, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 4971 + }, + { + "epoch": 0.04972, + "grad_norm": 1.0252366398011912, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 4972 + }, + { + "epoch": 0.04973, + "grad_norm": 1.0321786535910968, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 4973 + }, + { + "epoch": 0.04974, + "grad_norm": 1.3675268731180446, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 4974 + }, + { + "epoch": 0.04975, + "grad_norm": 0.9689102407574199, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 4975 + }, + { + "epoch": 0.04976, + "grad_norm": 1.034378936349661, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 4976 + }, + { + "epoch": 0.04977, + "grad_norm": 1.0659592144925003, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 4977 + }, + { + "epoch": 0.04978, + "grad_norm": 0.9463690418294757, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 4978 + }, + { + "epoch": 0.04979, + "grad_norm": 1.0778624556709335, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 4979 + }, + { + "epoch": 0.0498, + "grad_norm": 1.1912796103263936, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 4980 + }, + { + "epoch": 0.04981, + "grad_norm": 0.9428519138134729, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 4981 + }, + { + "epoch": 0.04982, + "grad_norm": 0.9114023013651604, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 4982 + }, + { + "epoch": 0.04983, + "grad_norm": 0.843362002569619, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 4983 + }, + { + "epoch": 0.04984, + "grad_norm": 0.7329533030111096, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 4984 + }, + { + "epoch": 0.04985, + "grad_norm": 0.8771158538637082, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 4985 + }, + { + "epoch": 0.04986, + "grad_norm": 0.9492314223910526, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 4986 + }, + { + "epoch": 0.04987, + "grad_norm": 0.9063125309075447, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 4987 + }, + { + "epoch": 0.04988, + "grad_norm": 0.8211391661395855, + "learning_rate": 0.003, + "loss": 4.09, + "step": 4988 + }, + { + "epoch": 0.04989, + "grad_norm": 0.87309817323828, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 4989 + }, + { + "epoch": 0.0499, + "grad_norm": 0.8139798258933807, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 4990 + }, + { + "epoch": 0.04991, + "grad_norm": 0.8390355952116492, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 4991 + }, + { + "epoch": 0.04992, + "grad_norm": 0.9935950349965598, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 4992 + }, + { + "epoch": 0.04993, + "grad_norm": 1.0610402805979944, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 4993 + }, + { + "epoch": 0.04994, + "grad_norm": 0.94882186728126, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 4994 + }, + { + "epoch": 0.04995, + "grad_norm": 0.867388049906827, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 4995 + }, + { + "epoch": 0.04996, + "grad_norm": 0.9650777431351639, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 4996 + }, + { + "epoch": 0.04997, + "grad_norm": 0.9684613017855864, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 4997 + }, + { + "epoch": 0.04998, + "grad_norm": 0.9878003417576896, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 4998 + }, + { + "epoch": 0.04999, + "grad_norm": 0.953046742937088, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 4999 + }, + { + "epoch": 0.05, + "grad_norm": 0.844866115668994, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 5000 + }, + { + "epoch": 0.05001, + "grad_norm": 0.8155144329490257, + "learning_rate": 0.003, + "loss": 4.098, + "step": 5001 + }, + { + "epoch": 0.05002, + "grad_norm": 0.9071734968733114, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 5002 + }, + { + "epoch": 0.05003, + "grad_norm": 1.0523308479446838, + "learning_rate": 0.003, + "loss": 4.092, + "step": 5003 + }, + { + "epoch": 0.05004, + "grad_norm": 1.0673372138869082, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 5004 + }, + { + "epoch": 0.05005, + "grad_norm": 1.0190354983748227, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 5005 + }, + { + "epoch": 0.05006, + "grad_norm": 1.0502501422500765, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 5006 + }, + { + "epoch": 0.05007, + "grad_norm": 1.0050657203422344, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 5007 + }, + { + "epoch": 0.05008, + "grad_norm": 1.0632875671321045, + "learning_rate": 0.003, + "loss": 4.063, + "step": 5008 + }, + { + "epoch": 0.05009, + "grad_norm": 0.7483220321269478, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 5009 + }, + { + "epoch": 0.0501, + "grad_norm": 0.6604154006789746, + "learning_rate": 0.003, + "loss": 4.076, + "step": 5010 + }, + { + "epoch": 0.05011, + "grad_norm": 0.7127635769855144, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 5011 + }, + { + "epoch": 0.05012, + "grad_norm": 0.8658367408694297, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 5012 + }, + { + "epoch": 0.05013, + "grad_norm": 1.0526587712962345, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 5013 + }, + { + "epoch": 0.05014, + "grad_norm": 1.2523503854735638, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 5014 + }, + { + "epoch": 0.05015, + "grad_norm": 0.8154569818223713, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 5015 + }, + { + "epoch": 0.05016, + "grad_norm": 0.8907021793572791, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 5016 + }, + { + "epoch": 0.05017, + "grad_norm": 0.9052920115396309, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5017 + }, + { + "epoch": 0.05018, + "grad_norm": 0.8754711897047392, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 5018 + }, + { + "epoch": 0.05019, + "grad_norm": 0.8774784593403933, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 5019 + }, + { + "epoch": 0.0502, + "grad_norm": 0.9389866910530438, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 5020 + }, + { + "epoch": 0.05021, + "grad_norm": 0.9591253397524557, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 5021 + }, + { + "epoch": 0.05022, + "grad_norm": 1.1105484396556573, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5022 + }, + { + "epoch": 0.05023, + "grad_norm": 1.3065590449165418, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 5023 + }, + { + "epoch": 0.05024, + "grad_norm": 0.9093611617258934, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 5024 + }, + { + "epoch": 0.05025, + "grad_norm": 0.8149464138741495, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 5025 + }, + { + "epoch": 0.05026, + "grad_norm": 0.8063387216832264, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 5026 + }, + { + "epoch": 0.05027, + "grad_norm": 0.9928443146318517, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 5027 + }, + { + "epoch": 0.05028, + "grad_norm": 1.1022004638491405, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 5028 + }, + { + "epoch": 0.05029, + "grad_norm": 1.0254112827493802, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 5029 + }, + { + "epoch": 0.0503, + "grad_norm": 0.9739559101423857, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 5030 + }, + { + "epoch": 0.05031, + "grad_norm": 1.195596806455749, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 5031 + }, + { + "epoch": 0.05032, + "grad_norm": 0.8132570593113878, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 5032 + }, + { + "epoch": 0.05033, + "grad_norm": 0.7335990765850869, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 5033 + }, + { + "epoch": 0.05034, + "grad_norm": 0.6554999770414173, + "learning_rate": 0.003, + "loss": 4.055, + "step": 5034 + }, + { + "epoch": 0.05035, + "grad_norm": 0.7532736427609922, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 5035 + }, + { + "epoch": 0.05036, + "grad_norm": 0.8283547847263479, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 5036 + }, + { + "epoch": 0.05037, + "grad_norm": 0.9606336416914282, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 5037 + }, + { + "epoch": 0.05038, + "grad_norm": 1.1756105988578247, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 5038 + }, + { + "epoch": 0.05039, + "grad_norm": 0.7260800281879718, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 5039 + }, + { + "epoch": 0.0504, + "grad_norm": 0.6895678700118688, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 5040 + }, + { + "epoch": 0.05041, + "grad_norm": 0.9054512238934349, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 5041 + }, + { + "epoch": 0.05042, + "grad_norm": 1.0681246931889021, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 5042 + }, + { + "epoch": 0.05043, + "grad_norm": 1.0081790953059804, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 5043 + }, + { + "epoch": 0.05044, + "grad_norm": 1.0083278627521677, + "learning_rate": 0.003, + "loss": 4.095, + "step": 5044 + }, + { + "epoch": 0.05045, + "grad_norm": 1.1323410375959377, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 5045 + }, + { + "epoch": 0.05046, + "grad_norm": 1.038046747857265, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 5046 + }, + { + "epoch": 0.05047, + "grad_norm": 0.8846272273108629, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 5047 + }, + { + "epoch": 0.05048, + "grad_norm": 1.039375733492065, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 5048 + }, + { + "epoch": 0.05049, + "grad_norm": 1.0198096987245278, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 5049 + }, + { + "epoch": 0.0505, + "grad_norm": 0.9914651248213096, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 5050 + }, + { + "epoch": 0.05051, + "grad_norm": 0.9522133302767952, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 5051 + }, + { + "epoch": 0.05052, + "grad_norm": 1.020178995716012, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 5052 + }, + { + "epoch": 0.05053, + "grad_norm": 0.9136513702774933, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 5053 + }, + { + "epoch": 0.05054, + "grad_norm": 0.8516625497650816, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 5054 + }, + { + "epoch": 0.05055, + "grad_norm": 0.8444636192801542, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 5055 + }, + { + "epoch": 0.05056, + "grad_norm": 0.8646124548092219, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 5056 + }, + { + "epoch": 0.05057, + "grad_norm": 0.954021030827355, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 5057 + }, + { + "epoch": 0.05058, + "grad_norm": 1.1354179546437237, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 5058 + }, + { + "epoch": 0.05059, + "grad_norm": 0.8412247009608202, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 5059 + }, + { + "epoch": 0.0506, + "grad_norm": 0.8258030312088399, + "learning_rate": 0.003, + "loss": 4.089, + "step": 5060 + }, + { + "epoch": 0.05061, + "grad_norm": 0.8796678252375678, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 5061 + }, + { + "epoch": 0.05062, + "grad_norm": 0.9882911088154981, + "learning_rate": 0.003, + "loss": 4.055, + "step": 5062 + }, + { + "epoch": 0.05063, + "grad_norm": 0.9432186009346578, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 5063 + }, + { + "epoch": 0.05064, + "grad_norm": 0.9485814708559243, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 5064 + }, + { + "epoch": 0.05065, + "grad_norm": 0.9973027340770267, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 5065 + }, + { + "epoch": 0.05066, + "grad_norm": 1.055670001942788, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 5066 + }, + { + "epoch": 0.05067, + "grad_norm": 0.9244772620512077, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 5067 + }, + { + "epoch": 0.05068, + "grad_norm": 0.958855347779479, + "learning_rate": 0.003, + "loss": 4.093, + "step": 5068 + }, + { + "epoch": 0.05069, + "grad_norm": 0.9074306933330328, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5069 + }, + { + "epoch": 0.0507, + "grad_norm": 0.7465736760883044, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 5070 + }, + { + "epoch": 0.05071, + "grad_norm": 0.7907220925527945, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 5071 + }, + { + "epoch": 0.05072, + "grad_norm": 0.8791832214881413, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 5072 + }, + { + "epoch": 0.05073, + "grad_norm": 1.1193497551932123, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 5073 + }, + { + "epoch": 0.05074, + "grad_norm": 1.007997199633059, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 5074 + }, + { + "epoch": 0.05075, + "grad_norm": 0.9210544932505602, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 5075 + }, + { + "epoch": 0.05076, + "grad_norm": 0.8971244157459003, + "learning_rate": 0.003, + "loss": 4.09, + "step": 5076 + }, + { + "epoch": 0.05077, + "grad_norm": 0.8845891734232929, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 5077 + }, + { + "epoch": 0.05078, + "grad_norm": 0.9113721083676786, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 5078 + }, + { + "epoch": 0.05079, + "grad_norm": 0.9084231790144179, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 5079 + }, + { + "epoch": 0.0508, + "grad_norm": 1.0477669246554797, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 5080 + }, + { + "epoch": 0.05081, + "grad_norm": 1.0274679159676723, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 5081 + }, + { + "epoch": 0.05082, + "grad_norm": 0.9811435796751755, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 5082 + }, + { + "epoch": 0.05083, + "grad_norm": 0.9115900027432353, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 5083 + }, + { + "epoch": 0.05084, + "grad_norm": 0.9665650959530513, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 5084 + }, + { + "epoch": 0.05085, + "grad_norm": 1.094785200544881, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 5085 + }, + { + "epoch": 0.05086, + "grad_norm": 1.0395235168756072, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 5086 + }, + { + "epoch": 0.05087, + "grad_norm": 1.1914983032220043, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 5087 + }, + { + "epoch": 0.05088, + "grad_norm": 0.9164887728560919, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 5088 + }, + { + "epoch": 0.05089, + "grad_norm": 0.9984418259866299, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 5089 + }, + { + "epoch": 0.0509, + "grad_norm": 0.9643661674080093, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 5090 + }, + { + "epoch": 0.05091, + "grad_norm": 0.9837794177673876, + "learning_rate": 0.003, + "loss": 4.067, + "step": 5091 + }, + { + "epoch": 0.05092, + "grad_norm": 0.8924921438341328, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 5092 + }, + { + "epoch": 0.05093, + "grad_norm": 0.8256552974904879, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 5093 + }, + { + "epoch": 0.05094, + "grad_norm": 0.9039163713516233, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 5094 + }, + { + "epoch": 0.05095, + "grad_norm": 1.2383238891133954, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 5095 + }, + { + "epoch": 0.05096, + "grad_norm": 0.9650892595888488, + "learning_rate": 0.003, + "loss": 4.074, + "step": 5096 + }, + { + "epoch": 0.05097, + "grad_norm": 0.9342147612808765, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 5097 + }, + { + "epoch": 0.05098, + "grad_norm": 0.9150581045088569, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 5098 + }, + { + "epoch": 0.05099, + "grad_norm": 1.0196257914961222, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 5099 + }, + { + "epoch": 0.051, + "grad_norm": 1.0100480797466045, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 5100 + }, + { + "epoch": 0.05101, + "grad_norm": 1.076613085076168, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 5101 + }, + { + "epoch": 0.05102, + "grad_norm": 0.9177055767593738, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 5102 + }, + { + "epoch": 0.05103, + "grad_norm": 0.7867724366196349, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 5103 + }, + { + "epoch": 0.05104, + "grad_norm": 0.8457986934357559, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 5104 + }, + { + "epoch": 0.05105, + "grad_norm": 0.8367744837015725, + "learning_rate": 0.003, + "loss": 4.055, + "step": 5105 + }, + { + "epoch": 0.05106, + "grad_norm": 0.8022124167980758, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 5106 + }, + { + "epoch": 0.05107, + "grad_norm": 0.6952811544982499, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 5107 + }, + { + "epoch": 0.05108, + "grad_norm": 0.7444855516396648, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 5108 + }, + { + "epoch": 0.05109, + "grad_norm": 0.878589896084288, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 5109 + }, + { + "epoch": 0.0511, + "grad_norm": 0.8745594675885457, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 5110 + }, + { + "epoch": 0.05111, + "grad_norm": 1.0139622357551754, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 5111 + }, + { + "epoch": 0.05112, + "grad_norm": 1.248129365292357, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 5112 + }, + { + "epoch": 0.05113, + "grad_norm": 0.9885210087032184, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 5113 + }, + { + "epoch": 0.05114, + "grad_norm": 0.9571971369820982, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 5114 + }, + { + "epoch": 0.05115, + "grad_norm": 0.9106967184247701, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 5115 + }, + { + "epoch": 0.05116, + "grad_norm": 1.053212984649482, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 5116 + }, + { + "epoch": 0.05117, + "grad_norm": 0.9762026130378781, + "learning_rate": 0.003, + "loss": 4.066, + "step": 5117 + }, + { + "epoch": 0.05118, + "grad_norm": 0.9813459473673732, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 5118 + }, + { + "epoch": 0.05119, + "grad_norm": 1.0553578473596843, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 5119 + }, + { + "epoch": 0.0512, + "grad_norm": 1.070915563107028, + "learning_rate": 0.003, + "loss": 4.045, + "step": 5120 + }, + { + "epoch": 0.05121, + "grad_norm": 0.8748371385007153, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 5121 + }, + { + "epoch": 0.05122, + "grad_norm": 0.960144764905716, + "learning_rate": 0.003, + "loss": 4.052, + "step": 5122 + }, + { + "epoch": 0.05123, + "grad_norm": 1.210628825057938, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5123 + }, + { + "epoch": 0.05124, + "grad_norm": 0.9135871799109425, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 5124 + }, + { + "epoch": 0.05125, + "grad_norm": 0.9676901109206436, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 5125 + }, + { + "epoch": 0.05126, + "grad_norm": 0.8403551330424058, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 5126 + }, + { + "epoch": 0.05127, + "grad_norm": 0.7466734661308947, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 5127 + }, + { + "epoch": 0.05128, + "grad_norm": 0.728346628542258, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 5128 + }, + { + "epoch": 0.05129, + "grad_norm": 0.7017928229749225, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 5129 + }, + { + "epoch": 0.0513, + "grad_norm": 0.7221803684673601, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 5130 + }, + { + "epoch": 0.05131, + "grad_norm": 0.7808750308328987, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 5131 + }, + { + "epoch": 0.05132, + "grad_norm": 0.8913684947688636, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 5132 + }, + { + "epoch": 0.05133, + "grad_norm": 1.1454285750946989, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 5133 + }, + { + "epoch": 0.05134, + "grad_norm": 0.9684078269337082, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 5134 + }, + { + "epoch": 0.05135, + "grad_norm": 1.1615129798672017, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5135 + }, + { + "epoch": 0.05136, + "grad_norm": 0.8816470981524503, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 5136 + }, + { + "epoch": 0.05137, + "grad_norm": 0.8808116644656396, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 5137 + }, + { + "epoch": 0.05138, + "grad_norm": 1.030677698421762, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 5138 + }, + { + "epoch": 0.05139, + "grad_norm": 1.1279946491267732, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 5139 + }, + { + "epoch": 0.0514, + "grad_norm": 0.9370222106532408, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 5140 + }, + { + "epoch": 0.05141, + "grad_norm": 0.8884938807254016, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 5141 + }, + { + "epoch": 0.05142, + "grad_norm": 0.840850706102145, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 5142 + }, + { + "epoch": 0.05143, + "grad_norm": 0.880595971401337, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 5143 + }, + { + "epoch": 0.05144, + "grad_norm": 0.8704477269765287, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 5144 + }, + { + "epoch": 0.05145, + "grad_norm": 0.8916851706228778, + "learning_rate": 0.003, + "loss": 4.055, + "step": 5145 + }, + { + "epoch": 0.05146, + "grad_norm": 0.9663208313115196, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 5146 + }, + { + "epoch": 0.05147, + "grad_norm": 1.1070241756865764, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 5147 + }, + { + "epoch": 0.05148, + "grad_norm": 0.9885155736285973, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 5148 + }, + { + "epoch": 0.05149, + "grad_norm": 0.9670853678419679, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 5149 + }, + { + "epoch": 0.0515, + "grad_norm": 0.9350123569517975, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 5150 + }, + { + "epoch": 0.05151, + "grad_norm": 0.9079989062981128, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 5151 + }, + { + "epoch": 0.05152, + "grad_norm": 0.9776926173785698, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 5152 + }, + { + "epoch": 0.05153, + "grad_norm": 0.9166249719052514, + "learning_rate": 0.003, + "loss": 4.07, + "step": 5153 + }, + { + "epoch": 0.05154, + "grad_norm": 0.8423591519799284, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 5154 + }, + { + "epoch": 0.05155, + "grad_norm": 0.9495784465505834, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 5155 + }, + { + "epoch": 0.05156, + "grad_norm": 0.8711852658243976, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 5156 + }, + { + "epoch": 0.05157, + "grad_norm": 0.9761077463552009, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 5157 + }, + { + "epoch": 0.05158, + "grad_norm": 0.9931352807114164, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 5158 + }, + { + "epoch": 0.05159, + "grad_norm": 1.1193695002600963, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 5159 + }, + { + "epoch": 0.0516, + "grad_norm": 0.9267022082064607, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 5160 + }, + { + "epoch": 0.05161, + "grad_norm": 0.9665280786599717, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 5161 + }, + { + "epoch": 0.05162, + "grad_norm": 1.0869042755370812, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 5162 + }, + { + "epoch": 0.05163, + "grad_norm": 1.0568255095175747, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 5163 + }, + { + "epoch": 0.05164, + "grad_norm": 1.1739265193923354, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 5164 + }, + { + "epoch": 0.05165, + "grad_norm": 0.9577726183728671, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 5165 + }, + { + "epoch": 0.05166, + "grad_norm": 1.1352998589608443, + "learning_rate": 0.003, + "loss": 4.084, + "step": 5166 + }, + { + "epoch": 0.05167, + "grad_norm": 0.8717433391514583, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 5167 + }, + { + "epoch": 0.05168, + "grad_norm": 0.870068697924481, + "learning_rate": 0.003, + "loss": 4.115, + "step": 5168 + }, + { + "epoch": 0.05169, + "grad_norm": 1.0023358246624934, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 5169 + }, + { + "epoch": 0.0517, + "grad_norm": 1.2431734452049883, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 5170 + }, + { + "epoch": 0.05171, + "grad_norm": 0.9116624433354724, + "learning_rate": 0.003, + "loss": 4.068, + "step": 5171 + }, + { + "epoch": 0.05172, + "grad_norm": 0.9323927967551716, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 5172 + }, + { + "epoch": 0.05173, + "grad_norm": 1.054805595580774, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 5173 + }, + { + "epoch": 0.05174, + "grad_norm": 1.122856751648051, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 5174 + }, + { + "epoch": 0.05175, + "grad_norm": 0.9937471915910451, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 5175 + }, + { + "epoch": 0.05176, + "grad_norm": 0.9840832744172103, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 5176 + }, + { + "epoch": 0.05177, + "grad_norm": 0.7800980213457716, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5177 + }, + { + "epoch": 0.05178, + "grad_norm": 0.8673592583630511, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 5178 + }, + { + "epoch": 0.05179, + "grad_norm": 0.8980602578393677, + "learning_rate": 0.003, + "loss": 4.07, + "step": 5179 + }, + { + "epoch": 0.0518, + "grad_norm": 1.0782172804196768, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 5180 + }, + { + "epoch": 0.05181, + "grad_norm": 1.3393343782861993, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 5181 + }, + { + "epoch": 0.05182, + "grad_norm": 0.6874879870433082, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 5182 + }, + { + "epoch": 0.05183, + "grad_norm": 0.8933939115278953, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 5183 + }, + { + "epoch": 0.05184, + "grad_norm": 1.0375617190613013, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 5184 + }, + { + "epoch": 0.05185, + "grad_norm": 0.978363863806633, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 5185 + }, + { + "epoch": 0.05186, + "grad_norm": 0.9787815400755827, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 5186 + }, + { + "epoch": 0.05187, + "grad_norm": 1.026722875042982, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 5187 + }, + { + "epoch": 0.05188, + "grad_norm": 1.0387976995995625, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 5188 + }, + { + "epoch": 0.05189, + "grad_norm": 0.9737693143733961, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 5189 + }, + { + "epoch": 0.0519, + "grad_norm": 1.1745663438513876, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 5190 + }, + { + "epoch": 0.05191, + "grad_norm": 0.9594735851770623, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 5191 + }, + { + "epoch": 0.05192, + "grad_norm": 0.9706881201053492, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 5192 + }, + { + "epoch": 0.05193, + "grad_norm": 0.927921231735236, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5193 + }, + { + "epoch": 0.05194, + "grad_norm": 0.8511889522904593, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 5194 + }, + { + "epoch": 0.05195, + "grad_norm": 0.8688574305845654, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 5195 + }, + { + "epoch": 0.05196, + "grad_norm": 1.0602526179506908, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 5196 + }, + { + "epoch": 0.05197, + "grad_norm": 1.004033518921005, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 5197 + }, + { + "epoch": 0.05198, + "grad_norm": 0.9382654504233616, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 5198 + }, + { + "epoch": 0.05199, + "grad_norm": 0.8240832698733139, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 5199 + }, + { + "epoch": 0.052, + "grad_norm": 0.8010190091673817, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 5200 + }, + { + "epoch": 0.05201, + "grad_norm": 0.9345369763160736, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5201 + }, + { + "epoch": 0.05202, + "grad_norm": 1.1219968562809288, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 5202 + }, + { + "epoch": 0.05203, + "grad_norm": 1.165320649672993, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 5203 + }, + { + "epoch": 0.05204, + "grad_norm": 1.0069940016760848, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 5204 + }, + { + "epoch": 0.05205, + "grad_norm": 0.9073269196171544, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 5205 + }, + { + "epoch": 0.05206, + "grad_norm": 0.9676783509672855, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5206 + }, + { + "epoch": 0.05207, + "grad_norm": 0.9614251201683252, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5207 + }, + { + "epoch": 0.05208, + "grad_norm": 0.8233264549032968, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 5208 + }, + { + "epoch": 0.05209, + "grad_norm": 0.795950781260021, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 5209 + }, + { + "epoch": 0.0521, + "grad_norm": 0.850100905132024, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 5210 + }, + { + "epoch": 0.05211, + "grad_norm": 1.0102855739696595, + "learning_rate": 0.003, + "loss": 4.064, + "step": 5211 + }, + { + "epoch": 0.05212, + "grad_norm": 1.3483238868956038, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 5212 + }, + { + "epoch": 0.05213, + "grad_norm": 0.771838626249321, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 5213 + }, + { + "epoch": 0.05214, + "grad_norm": 0.6993971591998677, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 5214 + }, + { + "epoch": 0.05215, + "grad_norm": 0.7581728248150723, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 5215 + }, + { + "epoch": 0.05216, + "grad_norm": 0.8519198915236995, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 5216 + }, + { + "epoch": 0.05217, + "grad_norm": 1.0745388491695431, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 5217 + }, + { + "epoch": 0.05218, + "grad_norm": 0.9935688948883226, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 5218 + }, + { + "epoch": 0.05219, + "grad_norm": 0.9084561291587792, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 5219 + }, + { + "epoch": 0.0522, + "grad_norm": 1.0337978058374222, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 5220 + }, + { + "epoch": 0.05221, + "grad_norm": 1.1422321793715118, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 5221 + }, + { + "epoch": 0.05222, + "grad_norm": 0.7554460463065314, + "learning_rate": 0.003, + "loss": 4.069, + "step": 5222 + }, + { + "epoch": 0.05223, + "grad_norm": 0.8000343955599867, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 5223 + }, + { + "epoch": 0.05224, + "grad_norm": 0.9379433719212857, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 5224 + }, + { + "epoch": 0.05225, + "grad_norm": 1.2417850745548258, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 5225 + }, + { + "epoch": 0.05226, + "grad_norm": 0.9382790639135026, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 5226 + }, + { + "epoch": 0.05227, + "grad_norm": 0.8974481342791866, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 5227 + }, + { + "epoch": 0.05228, + "grad_norm": 0.8629619700937163, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 5228 + }, + { + "epoch": 0.05229, + "grad_norm": 1.019751882064597, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 5229 + }, + { + "epoch": 0.0523, + "grad_norm": 1.1377445229403527, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 5230 + }, + { + "epoch": 0.05231, + "grad_norm": 0.9424656899524195, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 5231 + }, + { + "epoch": 0.05232, + "grad_norm": 0.9575611751186932, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 5232 + }, + { + "epoch": 0.05233, + "grad_norm": 0.9293995877028672, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 5233 + }, + { + "epoch": 0.05234, + "grad_norm": 0.9615259872002241, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 5234 + }, + { + "epoch": 0.05235, + "grad_norm": 0.983028393165795, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 5235 + }, + { + "epoch": 0.05236, + "grad_norm": 1.0016310432668707, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 5236 + }, + { + "epoch": 0.05237, + "grad_norm": 1.015432602850747, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 5237 + }, + { + "epoch": 0.05238, + "grad_norm": 1.078845385798474, + "learning_rate": 0.003, + "loss": 4.058, + "step": 5238 + }, + { + "epoch": 0.05239, + "grad_norm": 0.9307396208378222, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 5239 + }, + { + "epoch": 0.0524, + "grad_norm": 0.9571027247324106, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 5240 + }, + { + "epoch": 0.05241, + "grad_norm": 0.9792332364709184, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 5241 + }, + { + "epoch": 0.05242, + "grad_norm": 0.9182971281795369, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 5242 + }, + { + "epoch": 0.05243, + "grad_norm": 0.8942676834653476, + "learning_rate": 0.003, + "loss": 4.082, + "step": 5243 + }, + { + "epoch": 0.05244, + "grad_norm": 0.89594845212589, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 5244 + }, + { + "epoch": 0.05245, + "grad_norm": 1.0089126308943768, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 5245 + }, + { + "epoch": 0.05246, + "grad_norm": 1.192205008646422, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 5246 + }, + { + "epoch": 0.05247, + "grad_norm": 1.0822271048921281, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 5247 + }, + { + "epoch": 0.05248, + "grad_norm": 0.9132052877500991, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 5248 + }, + { + "epoch": 0.05249, + "grad_norm": 1.0439345558274598, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 5249 + }, + { + "epoch": 0.0525, + "grad_norm": 1.1063976766968928, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 5250 + }, + { + "epoch": 0.05251, + "grad_norm": 0.8427299394254306, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 5251 + }, + { + "epoch": 0.05252, + "grad_norm": 0.8460576745435954, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 5252 + }, + { + "epoch": 0.05253, + "grad_norm": 0.7631623954154223, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 5253 + }, + { + "epoch": 0.05254, + "grad_norm": 0.7582085848780449, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 5254 + }, + { + "epoch": 0.05255, + "grad_norm": 0.871269617254244, + "learning_rate": 0.003, + "loss": 4.05, + "step": 5255 + }, + { + "epoch": 0.05256, + "grad_norm": 1.1274865245345602, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5256 + }, + { + "epoch": 0.05257, + "grad_norm": 0.9482986096135555, + "learning_rate": 0.003, + "loss": 4.07, + "step": 5257 + }, + { + "epoch": 0.05258, + "grad_norm": 0.8120437150184056, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 5258 + }, + { + "epoch": 0.05259, + "grad_norm": 0.7370127557005074, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 5259 + }, + { + "epoch": 0.0526, + "grad_norm": 0.8055148797864925, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 5260 + }, + { + "epoch": 0.05261, + "grad_norm": 0.7414676121854172, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 5261 + }, + { + "epoch": 0.05262, + "grad_norm": 0.8431409042373192, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 5262 + }, + { + "epoch": 0.05263, + "grad_norm": 0.9936473719599972, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 5263 + }, + { + "epoch": 0.05264, + "grad_norm": 1.2454811247723998, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 5264 + }, + { + "epoch": 0.05265, + "grad_norm": 0.7140907414977667, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 5265 + }, + { + "epoch": 0.05266, + "grad_norm": 0.6844293037667735, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 5266 + }, + { + "epoch": 0.05267, + "grad_norm": 0.7232891737003544, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 5267 + }, + { + "epoch": 0.05268, + "grad_norm": 0.8650861484246375, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 5268 + }, + { + "epoch": 0.05269, + "grad_norm": 1.078417956716715, + "learning_rate": 0.003, + "loss": 4.04, + "step": 5269 + }, + { + "epoch": 0.0527, + "grad_norm": 1.081957308191504, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 5270 + }, + { + "epoch": 0.05271, + "grad_norm": 1.0713979709889552, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 5271 + }, + { + "epoch": 0.05272, + "grad_norm": 0.9799954386145029, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 5272 + }, + { + "epoch": 0.05273, + "grad_norm": 1.1944522490266327, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 5273 + }, + { + "epoch": 0.05274, + "grad_norm": 0.9595893728165487, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 5274 + }, + { + "epoch": 0.05275, + "grad_norm": 1.0373746989136285, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 5275 + }, + { + "epoch": 0.05276, + "grad_norm": 0.9783693469288199, + "learning_rate": 0.003, + "loss": 4.055, + "step": 5276 + }, + { + "epoch": 0.05277, + "grad_norm": 0.9308459462602469, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 5277 + }, + { + "epoch": 0.05278, + "grad_norm": 0.9127924883155014, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 5278 + }, + { + "epoch": 0.05279, + "grad_norm": 0.8905127840769295, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 5279 + }, + { + "epoch": 0.0528, + "grad_norm": 0.9160560894552265, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 5280 + }, + { + "epoch": 0.05281, + "grad_norm": 0.9898967226775388, + "learning_rate": 0.003, + "loss": 4.073, + "step": 5281 + }, + { + "epoch": 0.05282, + "grad_norm": 1.0607654232659915, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 5282 + }, + { + "epoch": 0.05283, + "grad_norm": 0.7891736411399345, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 5283 + }, + { + "epoch": 0.05284, + "grad_norm": 0.7117985343229531, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 5284 + }, + { + "epoch": 0.05285, + "grad_norm": 0.7116143826691809, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 5285 + }, + { + "epoch": 0.05286, + "grad_norm": 0.7444977551621328, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 5286 + }, + { + "epoch": 0.05287, + "grad_norm": 0.7927132129772895, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 5287 + }, + { + "epoch": 0.05288, + "grad_norm": 0.8599837583361685, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 5288 + }, + { + "epoch": 0.05289, + "grad_norm": 1.030013828539183, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 5289 + }, + { + "epoch": 0.0529, + "grad_norm": 1.320070502048142, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 5290 + }, + { + "epoch": 0.05291, + "grad_norm": 0.708003930048573, + "learning_rate": 0.003, + "loss": 4.083, + "step": 5291 + }, + { + "epoch": 0.05292, + "grad_norm": 0.7099716170045386, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5292 + }, + { + "epoch": 0.05293, + "grad_norm": 1.0571312622562001, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 5293 + }, + { + "epoch": 0.05294, + "grad_norm": 1.0749670114406467, + "learning_rate": 0.003, + "loss": 4.063, + "step": 5294 + }, + { + "epoch": 0.05295, + "grad_norm": 0.9466743696328925, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 5295 + }, + { + "epoch": 0.05296, + "grad_norm": 0.9334033172520125, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 5296 + }, + { + "epoch": 0.05297, + "grad_norm": 0.9085832862861088, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 5297 + }, + { + "epoch": 0.05298, + "grad_norm": 0.8733996457415404, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 5298 + }, + { + "epoch": 0.05299, + "grad_norm": 0.9180130832605389, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 5299 + }, + { + "epoch": 0.053, + "grad_norm": 0.9730509978164584, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 5300 + }, + { + "epoch": 0.05301, + "grad_norm": 1.0895476694519508, + "learning_rate": 0.003, + "loss": 4.09, + "step": 5301 + }, + { + "epoch": 0.05302, + "grad_norm": 1.0558202746074077, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 5302 + }, + { + "epoch": 0.05303, + "grad_norm": 0.9797012960446656, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 5303 + }, + { + "epoch": 0.05304, + "grad_norm": 1.0039073817988884, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 5304 + }, + { + "epoch": 0.05305, + "grad_norm": 1.1408892390131204, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 5305 + }, + { + "epoch": 0.05306, + "grad_norm": 0.93995449289434, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 5306 + }, + { + "epoch": 0.05307, + "grad_norm": 0.9726004341078571, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 5307 + }, + { + "epoch": 0.05308, + "grad_norm": 0.9846823344904868, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 5308 + }, + { + "epoch": 0.05309, + "grad_norm": 1.1645641513184255, + "learning_rate": 0.003, + "loss": 4.061, + "step": 5309 + }, + { + "epoch": 0.0531, + "grad_norm": 0.9255234669205171, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 5310 + }, + { + "epoch": 0.05311, + "grad_norm": 0.9954755426824138, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 5311 + }, + { + "epoch": 0.05312, + "grad_norm": 1.140912604498098, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 5312 + }, + { + "epoch": 0.05313, + "grad_norm": 1.0342859706036291, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 5313 + }, + { + "epoch": 0.05314, + "grad_norm": 1.0050903165394378, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 5314 + }, + { + "epoch": 0.05315, + "grad_norm": 1.082132110001949, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 5315 + }, + { + "epoch": 0.05316, + "grad_norm": 0.8775743349709987, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 5316 + }, + { + "epoch": 0.05317, + "grad_norm": 0.7524051363352736, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 5317 + }, + { + "epoch": 0.05318, + "grad_norm": 0.6968397634473481, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 5318 + }, + { + "epoch": 0.05319, + "grad_norm": 0.7303279353638819, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 5319 + }, + { + "epoch": 0.0532, + "grad_norm": 0.8312560764557734, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 5320 + }, + { + "epoch": 0.05321, + "grad_norm": 0.9440547584257499, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 5321 + }, + { + "epoch": 0.05322, + "grad_norm": 1.0145241749604252, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 5322 + }, + { + "epoch": 0.05323, + "grad_norm": 1.0026467673881212, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 5323 + }, + { + "epoch": 0.05324, + "grad_norm": 0.9550475870668266, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 5324 + }, + { + "epoch": 0.05325, + "grad_norm": 1.0200704518446446, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5325 + }, + { + "epoch": 0.05326, + "grad_norm": 0.9264156483614246, + "learning_rate": 0.003, + "loss": 4.095, + "step": 5326 + }, + { + "epoch": 0.05327, + "grad_norm": 0.9074863872575771, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 5327 + }, + { + "epoch": 0.05328, + "grad_norm": 0.9484736745601532, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 5328 + }, + { + "epoch": 0.05329, + "grad_norm": 1.1056353318349814, + "learning_rate": 0.003, + "loss": 4.076, + "step": 5329 + }, + { + "epoch": 0.0533, + "grad_norm": 0.8601194407643497, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 5330 + }, + { + "epoch": 0.05331, + "grad_norm": 0.8648893165616912, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 5331 + }, + { + "epoch": 0.05332, + "grad_norm": 1.0628193250784792, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 5332 + }, + { + "epoch": 0.05333, + "grad_norm": 1.1066111469551312, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 5333 + }, + { + "epoch": 0.05334, + "grad_norm": 0.953883540436792, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 5334 + }, + { + "epoch": 0.05335, + "grad_norm": 0.9008830143608041, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 5335 + }, + { + "epoch": 0.05336, + "grad_norm": 0.8996222966966763, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 5336 + }, + { + "epoch": 0.05337, + "grad_norm": 0.9470909772908184, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 5337 + }, + { + "epoch": 0.05338, + "grad_norm": 0.9368087966120492, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 5338 + }, + { + "epoch": 0.05339, + "grad_norm": 0.9839774636806952, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 5339 + }, + { + "epoch": 0.0534, + "grad_norm": 1.10664494283966, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 5340 + }, + { + "epoch": 0.05341, + "grad_norm": 1.0229490590816641, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 5341 + }, + { + "epoch": 0.05342, + "grad_norm": 1.0465584419606189, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 5342 + }, + { + "epoch": 0.05343, + "grad_norm": 1.1640384929974874, + "learning_rate": 0.003, + "loss": 4.086, + "step": 5343 + }, + { + "epoch": 0.05344, + "grad_norm": 0.884957738973249, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 5344 + }, + { + "epoch": 0.05345, + "grad_norm": 0.8433240280430108, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 5345 + }, + { + "epoch": 0.05346, + "grad_norm": 0.9566468834365702, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 5346 + }, + { + "epoch": 0.05347, + "grad_norm": 0.9368336590324607, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 5347 + }, + { + "epoch": 0.05348, + "grad_norm": 1.0141732765839058, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 5348 + }, + { + "epoch": 0.05349, + "grad_norm": 1.154078618191757, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 5349 + }, + { + "epoch": 0.0535, + "grad_norm": 1.0175917042790932, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 5350 + }, + { + "epoch": 0.05351, + "grad_norm": 0.8496580619444323, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 5351 + }, + { + "epoch": 0.05352, + "grad_norm": 0.8895367323216903, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 5352 + }, + { + "epoch": 0.05353, + "grad_norm": 0.9383414607042285, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 5353 + }, + { + "epoch": 0.05354, + "grad_norm": 0.9932999421264617, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 5354 + }, + { + "epoch": 0.05355, + "grad_norm": 1.1090959785058532, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 5355 + }, + { + "epoch": 0.05356, + "grad_norm": 0.9113253650658273, + "learning_rate": 0.003, + "loss": 4.089, + "step": 5356 + }, + { + "epoch": 0.05357, + "grad_norm": 0.9713969124835692, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 5357 + }, + { + "epoch": 0.05358, + "grad_norm": 1.1111797677781718, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 5358 + }, + { + "epoch": 0.05359, + "grad_norm": 1.12964921180821, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 5359 + }, + { + "epoch": 0.0536, + "grad_norm": 0.7734821023817453, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 5360 + }, + { + "epoch": 0.05361, + "grad_norm": 0.8519551241940354, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 5361 + }, + { + "epoch": 0.05362, + "grad_norm": 0.9322670717841903, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 5362 + }, + { + "epoch": 0.05363, + "grad_norm": 1.0074134211588726, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 5363 + }, + { + "epoch": 0.05364, + "grad_norm": 0.9873618582440303, + "learning_rate": 0.003, + "loss": 4.09, + "step": 5364 + }, + { + "epoch": 0.05365, + "grad_norm": 0.9112488220616313, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 5365 + }, + { + "epoch": 0.05366, + "grad_norm": 0.9087748964782536, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 5366 + }, + { + "epoch": 0.05367, + "grad_norm": 1.1130213121604795, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 5367 + }, + { + "epoch": 0.05368, + "grad_norm": 1.051640889768411, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 5368 + }, + { + "epoch": 0.05369, + "grad_norm": 0.9572517998403266, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 5369 + }, + { + "epoch": 0.0537, + "grad_norm": 1.132811606167689, + "learning_rate": 0.003, + "loss": 4.077, + "step": 5370 + }, + { + "epoch": 0.05371, + "grad_norm": 0.9187614369569057, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 5371 + }, + { + "epoch": 0.05372, + "grad_norm": 1.0470001010321264, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 5372 + }, + { + "epoch": 0.05373, + "grad_norm": 0.9617588492397875, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 5373 + }, + { + "epoch": 0.05374, + "grad_norm": 0.9433193815843771, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 5374 + }, + { + "epoch": 0.05375, + "grad_norm": 0.9043465599508047, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 5375 + }, + { + "epoch": 0.05376, + "grad_norm": 0.8650017321204917, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 5376 + }, + { + "epoch": 0.05377, + "grad_norm": 0.8366118978178542, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 5377 + }, + { + "epoch": 0.05378, + "grad_norm": 0.9549435064058331, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 5378 + }, + { + "epoch": 0.05379, + "grad_norm": 1.1106796516282058, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 5379 + }, + { + "epoch": 0.0538, + "grad_norm": 0.831546358079104, + "learning_rate": 0.003, + "loss": 4.054, + "step": 5380 + }, + { + "epoch": 0.05381, + "grad_norm": 0.7873840620558364, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 5381 + }, + { + "epoch": 0.05382, + "grad_norm": 0.7745681783586985, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 5382 + }, + { + "epoch": 0.05383, + "grad_norm": 0.7612145463814121, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 5383 + }, + { + "epoch": 0.05384, + "grad_norm": 0.8198652694354317, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 5384 + }, + { + "epoch": 0.05385, + "grad_norm": 0.9417575755533557, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 5385 + }, + { + "epoch": 0.05386, + "grad_norm": 0.9333515318457634, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 5386 + }, + { + "epoch": 0.05387, + "grad_norm": 0.9470842553395012, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 5387 + }, + { + "epoch": 0.05388, + "grad_norm": 1.0202541931803695, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 5388 + }, + { + "epoch": 0.05389, + "grad_norm": 1.0837366526435621, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 5389 + }, + { + "epoch": 0.0539, + "grad_norm": 0.8073770997786047, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 5390 + }, + { + "epoch": 0.05391, + "grad_norm": 0.7682773134551675, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 5391 + }, + { + "epoch": 0.05392, + "grad_norm": 0.7018203354535223, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 5392 + }, + { + "epoch": 0.05393, + "grad_norm": 0.8253528505413662, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 5393 + }, + { + "epoch": 0.05394, + "grad_norm": 0.8821082824869386, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 5394 + }, + { + "epoch": 0.05395, + "grad_norm": 1.0233479675976396, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 5395 + }, + { + "epoch": 0.05396, + "grad_norm": 0.9868685749565336, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 5396 + }, + { + "epoch": 0.05397, + "grad_norm": 0.9032364856754863, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 5397 + }, + { + "epoch": 0.05398, + "grad_norm": 0.9305986337257256, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 5398 + }, + { + "epoch": 0.05399, + "grad_norm": 0.943643567039578, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 5399 + }, + { + "epoch": 0.054, + "grad_norm": 0.9727836672752513, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 5400 + }, + { + "epoch": 0.05401, + "grad_norm": 1.0950746328718843, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 5401 + }, + { + "epoch": 0.05402, + "grad_norm": 0.9394359475802941, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 5402 + }, + { + "epoch": 0.05403, + "grad_norm": 1.100262618217403, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 5403 + }, + { + "epoch": 0.05404, + "grad_norm": 0.8289796141686676, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 5404 + }, + { + "epoch": 0.05405, + "grad_norm": 0.8162230398030023, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 5405 + }, + { + "epoch": 0.05406, + "grad_norm": 0.7883046496886217, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 5406 + }, + { + "epoch": 0.05407, + "grad_norm": 0.7936421744252384, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 5407 + }, + { + "epoch": 0.05408, + "grad_norm": 0.9053334375282556, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 5408 + }, + { + "epoch": 0.05409, + "grad_norm": 0.9451799555181426, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 5409 + }, + { + "epoch": 0.0541, + "grad_norm": 1.0666670568701657, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 5410 + }, + { + "epoch": 0.05411, + "grad_norm": 0.9181326660070915, + "learning_rate": 0.003, + "loss": 4.07, + "step": 5411 + }, + { + "epoch": 0.05412, + "grad_norm": 1.021682961374821, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 5412 + }, + { + "epoch": 0.05413, + "grad_norm": 1.1798050564607474, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 5413 + }, + { + "epoch": 0.05414, + "grad_norm": 0.9232912835820161, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 5414 + }, + { + "epoch": 0.05415, + "grad_norm": 0.8725384280761546, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 5415 + }, + { + "epoch": 0.05416, + "grad_norm": 0.8010341905607388, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 5416 + }, + { + "epoch": 0.05417, + "grad_norm": 0.9197583841797057, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 5417 + }, + { + "epoch": 0.05418, + "grad_norm": 0.9755108515112911, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 5418 + }, + { + "epoch": 0.05419, + "grad_norm": 0.9843859573873037, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 5419 + }, + { + "epoch": 0.0542, + "grad_norm": 1.1783693707948757, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 5420 + }, + { + "epoch": 0.05421, + "grad_norm": 1.0734083109765062, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 5421 + }, + { + "epoch": 0.05422, + "grad_norm": 0.980524671161552, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 5422 + }, + { + "epoch": 0.05423, + "grad_norm": 1.134071438565773, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5423 + }, + { + "epoch": 0.05424, + "grad_norm": 1.2330179209043135, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 5424 + }, + { + "epoch": 0.05425, + "grad_norm": 1.013866607135996, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 5425 + }, + { + "epoch": 0.05426, + "grad_norm": 1.1436320829532476, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 5426 + }, + { + "epoch": 0.05427, + "grad_norm": 0.8459842849855579, + "learning_rate": 0.003, + "loss": 4.057, + "step": 5427 + }, + { + "epoch": 0.05428, + "grad_norm": 0.9549700978254552, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 5428 + }, + { + "epoch": 0.05429, + "grad_norm": 0.9172026670565415, + "learning_rate": 0.003, + "loss": 4.077, + "step": 5429 + }, + { + "epoch": 0.0543, + "grad_norm": 1.120952695970183, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 5430 + }, + { + "epoch": 0.05431, + "grad_norm": 1.1974245320079564, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 5431 + }, + { + "epoch": 0.05432, + "grad_norm": 0.8852096173514169, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 5432 + }, + { + "epoch": 0.05433, + "grad_norm": 0.9401011460255838, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 5433 + }, + { + "epoch": 0.05434, + "grad_norm": 0.9678587871600712, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 5434 + }, + { + "epoch": 0.05435, + "grad_norm": 1.0899343219435063, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 5435 + }, + { + "epoch": 0.05436, + "grad_norm": 1.0563921820873317, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 5436 + }, + { + "epoch": 0.05437, + "grad_norm": 1.0856872480417759, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 5437 + }, + { + "epoch": 0.05438, + "grad_norm": 0.7984063525797823, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 5438 + }, + { + "epoch": 0.05439, + "grad_norm": 0.7454173407684945, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 5439 + }, + { + "epoch": 0.0544, + "grad_norm": 0.7921737605396232, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 5440 + }, + { + "epoch": 0.05441, + "grad_norm": 1.0225413249269952, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 5441 + }, + { + "epoch": 0.05442, + "grad_norm": 1.2118058028339453, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 5442 + }, + { + "epoch": 0.05443, + "grad_norm": 0.6321819000655003, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 5443 + }, + { + "epoch": 0.05444, + "grad_norm": 0.8027542268287786, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 5444 + }, + { + "epoch": 0.05445, + "grad_norm": 0.8735412506958715, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 5445 + }, + { + "epoch": 0.05446, + "grad_norm": 0.7936222595707908, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 5446 + }, + { + "epoch": 0.05447, + "grad_norm": 0.6924205266934994, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 5447 + }, + { + "epoch": 0.05448, + "grad_norm": 0.6812700924878988, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 5448 + }, + { + "epoch": 0.05449, + "grad_norm": 0.7401926125886807, + "learning_rate": 0.003, + "loss": 4.073, + "step": 5449 + }, + { + "epoch": 0.0545, + "grad_norm": 0.8398967204050066, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5450 + }, + { + "epoch": 0.05451, + "grad_norm": 0.9232648502496631, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 5451 + }, + { + "epoch": 0.05452, + "grad_norm": 1.018982529104012, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 5452 + }, + { + "epoch": 0.05453, + "grad_norm": 1.09144709823669, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 5453 + }, + { + "epoch": 0.05454, + "grad_norm": 0.9389422208676496, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 5454 + }, + { + "epoch": 0.05455, + "grad_norm": 1.011860130767919, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 5455 + }, + { + "epoch": 0.05456, + "grad_norm": 1.1005201477477284, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 5456 + }, + { + "epoch": 0.05457, + "grad_norm": 0.8394502161551594, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 5457 + }, + { + "epoch": 0.05458, + "grad_norm": 0.8522093521223234, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 5458 + }, + { + "epoch": 0.05459, + "grad_norm": 0.8521617398500968, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 5459 + }, + { + "epoch": 0.0546, + "grad_norm": 0.9898185774528117, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 5460 + }, + { + "epoch": 0.05461, + "grad_norm": 1.037914610295231, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 5461 + }, + { + "epoch": 0.05462, + "grad_norm": 1.0314633015920627, + "learning_rate": 0.003, + "loss": 4.099, + "step": 5462 + }, + { + "epoch": 0.05463, + "grad_norm": 1.0714702174113098, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 5463 + }, + { + "epoch": 0.05464, + "grad_norm": 1.0656649423089701, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 5464 + }, + { + "epoch": 0.05465, + "grad_norm": 1.0070370867844776, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 5465 + }, + { + "epoch": 0.05466, + "grad_norm": 0.971949088054457, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 5466 + }, + { + "epoch": 0.05467, + "grad_norm": 1.1027271603054372, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 5467 + }, + { + "epoch": 0.05468, + "grad_norm": 0.9722631661342838, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 5468 + }, + { + "epoch": 0.05469, + "grad_norm": 1.535529064177968, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 5469 + }, + { + "epoch": 0.0547, + "grad_norm": 0.8945992610917329, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 5470 + }, + { + "epoch": 0.05471, + "grad_norm": 0.8971438224187216, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 5471 + }, + { + "epoch": 0.05472, + "grad_norm": 0.785537906391861, + "learning_rate": 0.003, + "loss": 4.061, + "step": 5472 + }, + { + "epoch": 0.05473, + "grad_norm": 0.7645096419487161, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5473 + }, + { + "epoch": 0.05474, + "grad_norm": 0.9043493315770855, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 5474 + }, + { + "epoch": 0.05475, + "grad_norm": 0.9826087228695892, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 5475 + }, + { + "epoch": 0.05476, + "grad_norm": 1.0455880982626005, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 5476 + }, + { + "epoch": 0.05477, + "grad_norm": 1.0460718667427993, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 5477 + }, + { + "epoch": 0.05478, + "grad_norm": 0.9784914027077458, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 5478 + }, + { + "epoch": 0.05479, + "grad_norm": 1.06468120790793, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 5479 + }, + { + "epoch": 0.0548, + "grad_norm": 0.9724427087606514, + "learning_rate": 0.003, + "loss": 4.08, + "step": 5480 + }, + { + "epoch": 0.05481, + "grad_norm": 1.139450823817563, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 5481 + }, + { + "epoch": 0.05482, + "grad_norm": 1.0103917095506671, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 5482 + }, + { + "epoch": 0.05483, + "grad_norm": 0.9941838375298019, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 5483 + }, + { + "epoch": 0.05484, + "grad_norm": 1.0065630407668353, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 5484 + }, + { + "epoch": 0.05485, + "grad_norm": 1.1859794607003078, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 5485 + }, + { + "epoch": 0.05486, + "grad_norm": 0.9117774295489004, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 5486 + }, + { + "epoch": 0.05487, + "grad_norm": 1.0354321077724413, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5487 + }, + { + "epoch": 0.05488, + "grad_norm": 0.9520742156522823, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 5488 + }, + { + "epoch": 0.05489, + "grad_norm": 0.9583850608961483, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 5489 + }, + { + "epoch": 0.0549, + "grad_norm": 0.9323245193356091, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5490 + }, + { + "epoch": 0.05491, + "grad_norm": 0.9691425932975921, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 5491 + }, + { + "epoch": 0.05492, + "grad_norm": 1.0404125399983395, + "learning_rate": 0.003, + "loss": 4.092, + "step": 5492 + }, + { + "epoch": 0.05493, + "grad_norm": 0.8797071139515741, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 5493 + }, + { + "epoch": 0.05494, + "grad_norm": 0.9257700859047452, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 5494 + }, + { + "epoch": 0.05495, + "grad_norm": 0.837046815044842, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 5495 + }, + { + "epoch": 0.05496, + "grad_norm": 0.868795138162349, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 5496 + }, + { + "epoch": 0.05497, + "grad_norm": 1.1666053988158387, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 5497 + }, + { + "epoch": 0.05498, + "grad_norm": 0.9182958212992614, + "learning_rate": 0.003, + "loss": 4.066, + "step": 5498 + }, + { + "epoch": 0.05499, + "grad_norm": 0.8243177544407708, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 5499 + }, + { + "epoch": 0.055, + "grad_norm": 0.930633426581144, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 5500 + }, + { + "epoch": 0.05501, + "grad_norm": 0.9853106164483082, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 5501 + }, + { + "epoch": 0.05502, + "grad_norm": 1.002532887849376, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 5502 + }, + { + "epoch": 0.05503, + "grad_norm": 1.134728987679396, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 5503 + }, + { + "epoch": 0.05504, + "grad_norm": 0.8688543177042416, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 5504 + }, + { + "epoch": 0.05505, + "grad_norm": 0.8885019557224482, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 5505 + }, + { + "epoch": 0.05506, + "grad_norm": 1.0191696763985227, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 5506 + }, + { + "epoch": 0.05507, + "grad_norm": 0.9581989274432549, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 5507 + }, + { + "epoch": 0.05508, + "grad_norm": 1.2059958374753366, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 5508 + }, + { + "epoch": 0.05509, + "grad_norm": 1.0509079701116766, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 5509 + }, + { + "epoch": 0.0551, + "grad_norm": 0.9629459111896204, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 5510 + }, + { + "epoch": 0.05511, + "grad_norm": 0.9385152557998871, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 5511 + }, + { + "epoch": 0.05512, + "grad_norm": 0.9992228341288117, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5512 + }, + { + "epoch": 0.05513, + "grad_norm": 0.9076413869206522, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 5513 + }, + { + "epoch": 0.05514, + "grad_norm": 0.9454063730057051, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 5514 + }, + { + "epoch": 0.05515, + "grad_norm": 0.9333899274103106, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 5515 + }, + { + "epoch": 0.05516, + "grad_norm": 0.9128223147263614, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 5516 + }, + { + "epoch": 0.05517, + "grad_norm": 0.9501252498361874, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 5517 + }, + { + "epoch": 0.05518, + "grad_norm": 0.8979581497014404, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 5518 + }, + { + "epoch": 0.05519, + "grad_norm": 0.8443581212066782, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 5519 + }, + { + "epoch": 0.0552, + "grad_norm": 0.8902006864056781, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 5520 + }, + { + "epoch": 0.05521, + "grad_norm": 1.087402109927747, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 5521 + }, + { + "epoch": 0.05522, + "grad_norm": 1.0502403491974475, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 5522 + }, + { + "epoch": 0.05523, + "grad_norm": 1.0452410374132677, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 5523 + }, + { + "epoch": 0.05524, + "grad_norm": 0.9514484278835924, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 5524 + }, + { + "epoch": 0.05525, + "grad_norm": 1.0272241045462294, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 5525 + }, + { + "epoch": 0.05526, + "grad_norm": 0.8733718754810375, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 5526 + }, + { + "epoch": 0.05527, + "grad_norm": 0.8731879871732975, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 5527 + }, + { + "epoch": 0.05528, + "grad_norm": 1.1037182994286576, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 5528 + }, + { + "epoch": 0.05529, + "grad_norm": 1.0027208430036605, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 5529 + }, + { + "epoch": 0.0553, + "grad_norm": 1.035023109428945, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 5530 + }, + { + "epoch": 0.05531, + "grad_norm": 0.9462736114003235, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 5531 + }, + { + "epoch": 0.05532, + "grad_norm": 1.1792902449799267, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 5532 + }, + { + "epoch": 0.05533, + "grad_norm": 1.0162972400663883, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 5533 + }, + { + "epoch": 0.05534, + "grad_norm": 1.1486180781871198, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 5534 + }, + { + "epoch": 0.05535, + "grad_norm": 1.0752582119191083, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 5535 + }, + { + "epoch": 0.05536, + "grad_norm": 0.9581234101005884, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 5536 + }, + { + "epoch": 0.05537, + "grad_norm": 0.8208048349609377, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 5537 + }, + { + "epoch": 0.05538, + "grad_norm": 0.9473770252339098, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 5538 + }, + { + "epoch": 0.05539, + "grad_norm": 1.2702509072007524, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 5539 + }, + { + "epoch": 0.0554, + "grad_norm": 0.9017522335362036, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 5540 + }, + { + "epoch": 0.05541, + "grad_norm": 0.8359582648956668, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 5541 + }, + { + "epoch": 0.05542, + "grad_norm": 0.9218480107385333, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 5542 + }, + { + "epoch": 0.05543, + "grad_norm": 0.8696347559650821, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5543 + }, + { + "epoch": 0.05544, + "grad_norm": 1.055625206756629, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 5544 + }, + { + "epoch": 0.05545, + "grad_norm": 1.1728733370559363, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 5545 + }, + { + "epoch": 0.05546, + "grad_norm": 0.7788999498236301, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 5546 + }, + { + "epoch": 0.05547, + "grad_norm": 0.8311570515952018, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 5547 + }, + { + "epoch": 0.05548, + "grad_norm": 0.9335591947026343, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 5548 + }, + { + "epoch": 0.05549, + "grad_norm": 1.0573495296822653, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 5549 + }, + { + "epoch": 0.0555, + "grad_norm": 1.0502011204824901, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5550 + }, + { + "epoch": 0.05551, + "grad_norm": 0.9831951258747678, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 5551 + }, + { + "epoch": 0.05552, + "grad_norm": 1.1313502639381925, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 5552 + }, + { + "epoch": 0.05553, + "grad_norm": 0.8713935309218466, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 5553 + }, + { + "epoch": 0.05554, + "grad_norm": 0.9135700599486382, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 5554 + }, + { + "epoch": 0.05555, + "grad_norm": 0.8386399502596644, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 5555 + }, + { + "epoch": 0.05556, + "grad_norm": 0.8427182198646174, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 5556 + }, + { + "epoch": 0.05557, + "grad_norm": 0.8853744024878168, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 5557 + }, + { + "epoch": 0.05558, + "grad_norm": 0.8682691991197287, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 5558 + }, + { + "epoch": 0.05559, + "grad_norm": 0.9168284316303011, + "learning_rate": 0.003, + "loss": 4.065, + "step": 5559 + }, + { + "epoch": 0.0556, + "grad_norm": 0.8495910068211255, + "learning_rate": 0.003, + "loss": 4.088, + "step": 5560 + }, + { + "epoch": 0.05561, + "grad_norm": 0.8580860935781918, + "learning_rate": 0.003, + "loss": 4.045, + "step": 5561 + }, + { + "epoch": 0.05562, + "grad_norm": 0.8995255593141265, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 5562 + }, + { + "epoch": 0.05563, + "grad_norm": 0.9397411181545186, + "learning_rate": 0.003, + "loss": 4.067, + "step": 5563 + }, + { + "epoch": 0.05564, + "grad_norm": 0.9961124788426614, + "learning_rate": 0.003, + "loss": 4.052, + "step": 5564 + }, + { + "epoch": 0.05565, + "grad_norm": 1.017582443976892, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 5565 + }, + { + "epoch": 0.05566, + "grad_norm": 0.9196943277209378, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 5566 + }, + { + "epoch": 0.05567, + "grad_norm": 0.887778482100592, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 5567 + }, + { + "epoch": 0.05568, + "grad_norm": 0.9446867885667367, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 5568 + }, + { + "epoch": 0.05569, + "grad_norm": 1.2048355489642966, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 5569 + }, + { + "epoch": 0.0557, + "grad_norm": 0.9947727024541948, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 5570 + }, + { + "epoch": 0.05571, + "grad_norm": 0.9242279105036012, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 5571 + }, + { + "epoch": 0.05572, + "grad_norm": 0.8718726616570441, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 5572 + }, + { + "epoch": 0.05573, + "grad_norm": 0.9988632134064598, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 5573 + }, + { + "epoch": 0.05574, + "grad_norm": 1.0858108953319112, + "learning_rate": 0.003, + "loss": 4.067, + "step": 5574 + }, + { + "epoch": 0.05575, + "grad_norm": 0.89341861731438, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 5575 + }, + { + "epoch": 0.05576, + "grad_norm": 0.9927093004321265, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 5576 + }, + { + "epoch": 0.05577, + "grad_norm": 1.0561976942629276, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 5577 + }, + { + "epoch": 0.05578, + "grad_norm": 0.9226854378392683, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 5578 + }, + { + "epoch": 0.05579, + "grad_norm": 0.9660492552781841, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 5579 + }, + { + "epoch": 0.0558, + "grad_norm": 1.0219080757366508, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 5580 + }, + { + "epoch": 0.05581, + "grad_norm": 0.9262442274506089, + "learning_rate": 0.003, + "loss": 4.084, + "step": 5581 + }, + { + "epoch": 0.05582, + "grad_norm": 0.9589806371639085, + "learning_rate": 0.003, + "loss": 4.096, + "step": 5582 + }, + { + "epoch": 0.05583, + "grad_norm": 1.0645509578183838, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 5583 + }, + { + "epoch": 0.05584, + "grad_norm": 1.051226914948146, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 5584 + }, + { + "epoch": 0.05585, + "grad_norm": 1.0467720353310674, + "learning_rate": 0.003, + "loss": 4.043, + "step": 5585 + }, + { + "epoch": 0.05586, + "grad_norm": 0.9663212590556503, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 5586 + }, + { + "epoch": 0.05587, + "grad_norm": 0.9185379338813185, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 5587 + }, + { + "epoch": 0.05588, + "grad_norm": 0.9694934031853484, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 5588 + }, + { + "epoch": 0.05589, + "grad_norm": 1.0526451937838062, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 5589 + }, + { + "epoch": 0.0559, + "grad_norm": 1.1085070368071448, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 5590 + }, + { + "epoch": 0.05591, + "grad_norm": 1.0162203009874562, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5591 + }, + { + "epoch": 0.05592, + "grad_norm": 0.951694969481166, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 5592 + }, + { + "epoch": 0.05593, + "grad_norm": 0.8175074193854779, + "learning_rate": 0.003, + "loss": 4.064, + "step": 5593 + }, + { + "epoch": 0.05594, + "grad_norm": 0.9255694773963952, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 5594 + }, + { + "epoch": 0.05595, + "grad_norm": 1.0156609990647611, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 5595 + }, + { + "epoch": 0.05596, + "grad_norm": 0.932955278279892, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 5596 + }, + { + "epoch": 0.05597, + "grad_norm": 0.920908250325548, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 5597 + }, + { + "epoch": 0.05598, + "grad_norm": 1.0626157594613586, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 5598 + }, + { + "epoch": 0.05599, + "grad_norm": 0.9591722885772912, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 5599 + }, + { + "epoch": 0.056, + "grad_norm": 0.8972051686796552, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 5600 + }, + { + "epoch": 0.05601, + "grad_norm": 0.94039167026587, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 5601 + }, + { + "epoch": 0.05602, + "grad_norm": 0.8789138951678084, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 5602 + }, + { + "epoch": 0.05603, + "grad_norm": 1.0980048694137343, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 5603 + }, + { + "epoch": 0.05604, + "grad_norm": 1.132678130199623, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 5604 + }, + { + "epoch": 0.05605, + "grad_norm": 0.9376010958362765, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 5605 + }, + { + "epoch": 0.05606, + "grad_norm": 1.012601875884022, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 5606 + }, + { + "epoch": 0.05607, + "grad_norm": 1.078658031552116, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 5607 + }, + { + "epoch": 0.05608, + "grad_norm": 1.0104480034658165, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 5608 + }, + { + "epoch": 0.05609, + "grad_norm": 1.1587851237585396, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 5609 + }, + { + "epoch": 0.0561, + "grad_norm": 0.9960041589297254, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 5610 + }, + { + "epoch": 0.05611, + "grad_norm": 0.9361590301922911, + "learning_rate": 0.003, + "loss": 4.072, + "step": 5611 + }, + { + "epoch": 0.05612, + "grad_norm": 0.9764370565765741, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 5612 + }, + { + "epoch": 0.05613, + "grad_norm": 1.0014577790700079, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 5613 + }, + { + "epoch": 0.05614, + "grad_norm": 0.9831216849472494, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 5614 + }, + { + "epoch": 0.05615, + "grad_norm": 1.1586083324328103, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 5615 + }, + { + "epoch": 0.05616, + "grad_norm": 0.8890921040375313, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 5616 + }, + { + "epoch": 0.05617, + "grad_norm": 0.7957949051675782, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 5617 + }, + { + "epoch": 0.05618, + "grad_norm": 0.7215932182196926, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 5618 + }, + { + "epoch": 0.05619, + "grad_norm": 0.7741180783292809, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 5619 + }, + { + "epoch": 0.0562, + "grad_norm": 0.7511553658323501, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 5620 + }, + { + "epoch": 0.05621, + "grad_norm": 0.6396239807569799, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 5621 + }, + { + "epoch": 0.05622, + "grad_norm": 0.6074326667753387, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5622 + }, + { + "epoch": 0.05623, + "grad_norm": 0.6362168788902988, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 5623 + }, + { + "epoch": 0.05624, + "grad_norm": 0.7164948632194134, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 5624 + }, + { + "epoch": 0.05625, + "grad_norm": 0.7858917280239912, + "learning_rate": 0.003, + "loss": 4.084, + "step": 5625 + }, + { + "epoch": 0.05626, + "grad_norm": 0.9610068641748187, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 5626 + }, + { + "epoch": 0.05627, + "grad_norm": 1.0076786147054788, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 5627 + }, + { + "epoch": 0.05628, + "grad_norm": 1.1184221088760653, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 5628 + }, + { + "epoch": 0.05629, + "grad_norm": 1.009005087989754, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 5629 + }, + { + "epoch": 0.0563, + "grad_norm": 0.9675845365105762, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 5630 + }, + { + "epoch": 0.05631, + "grad_norm": 0.9722083145337348, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 5631 + }, + { + "epoch": 0.05632, + "grad_norm": 0.9663539092020825, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 5632 + }, + { + "epoch": 0.05633, + "grad_norm": 0.9758861223474846, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 5633 + }, + { + "epoch": 0.05634, + "grad_norm": 1.1099380799726346, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 5634 + }, + { + "epoch": 0.05635, + "grad_norm": 0.9793289159407238, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 5635 + }, + { + "epoch": 0.05636, + "grad_norm": 1.0928319649450002, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 5636 + }, + { + "epoch": 0.05637, + "grad_norm": 0.9411278378658644, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 5637 + }, + { + "epoch": 0.05638, + "grad_norm": 1.006716505320645, + "learning_rate": 0.003, + "loss": 4.056, + "step": 5638 + }, + { + "epoch": 0.05639, + "grad_norm": 1.140827528278389, + "learning_rate": 0.003, + "loss": 4.076, + "step": 5639 + }, + { + "epoch": 0.0564, + "grad_norm": 1.1625088389769833, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 5640 + }, + { + "epoch": 0.05641, + "grad_norm": 0.9893427923995295, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 5641 + }, + { + "epoch": 0.05642, + "grad_norm": 0.9716727488546525, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 5642 + }, + { + "epoch": 0.05643, + "grad_norm": 0.8388359388426512, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 5643 + }, + { + "epoch": 0.05644, + "grad_norm": 0.8864776479628189, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 5644 + }, + { + "epoch": 0.05645, + "grad_norm": 1.0289020783924894, + "learning_rate": 0.003, + "loss": 4.058, + "step": 5645 + }, + { + "epoch": 0.05646, + "grad_norm": 1.1004794764123877, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 5646 + }, + { + "epoch": 0.05647, + "grad_norm": 0.9547139445319417, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 5647 + }, + { + "epoch": 0.05648, + "grad_norm": 0.967189977084837, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 5648 + }, + { + "epoch": 0.05649, + "grad_norm": 0.973718841402612, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5649 + }, + { + "epoch": 0.0565, + "grad_norm": 1.02998820647941, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 5650 + }, + { + "epoch": 0.05651, + "grad_norm": 0.8716926811029921, + "learning_rate": 0.003, + "loss": 4.048, + "step": 5651 + }, + { + "epoch": 0.05652, + "grad_norm": 0.8882622031970813, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 5652 + }, + { + "epoch": 0.05653, + "grad_norm": 0.8791179996457095, + "learning_rate": 0.003, + "loss": 4.077, + "step": 5653 + }, + { + "epoch": 0.05654, + "grad_norm": 0.9428467074127278, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 5654 + }, + { + "epoch": 0.05655, + "grad_norm": 0.9168230086253641, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 5655 + }, + { + "epoch": 0.05656, + "grad_norm": 0.9563246120783195, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 5656 + }, + { + "epoch": 0.05657, + "grad_norm": 1.0720010551454098, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 5657 + }, + { + "epoch": 0.05658, + "grad_norm": 0.8856156383346027, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 5658 + }, + { + "epoch": 0.05659, + "grad_norm": 0.8368418786449191, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 5659 + }, + { + "epoch": 0.0566, + "grad_norm": 0.8374298075537798, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 5660 + }, + { + "epoch": 0.05661, + "grad_norm": 0.9364487912430249, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 5661 + }, + { + "epoch": 0.05662, + "grad_norm": 1.000052264217599, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 5662 + }, + { + "epoch": 0.05663, + "grad_norm": 1.0446835856273087, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 5663 + }, + { + "epoch": 0.05664, + "grad_norm": 0.9023506375796267, + "learning_rate": 0.003, + "loss": 4.045, + "step": 5664 + }, + { + "epoch": 0.05665, + "grad_norm": 0.977889227291688, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 5665 + }, + { + "epoch": 0.05666, + "grad_norm": 1.130926840966814, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 5666 + }, + { + "epoch": 0.05667, + "grad_norm": 1.0607816508483603, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 5667 + }, + { + "epoch": 0.05668, + "grad_norm": 0.9792158523254157, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 5668 + }, + { + "epoch": 0.05669, + "grad_norm": 0.9055981740448197, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 5669 + }, + { + "epoch": 0.0567, + "grad_norm": 1.1476920087687061, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 5670 + }, + { + "epoch": 0.05671, + "grad_norm": 0.8960616098490328, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5671 + }, + { + "epoch": 0.05672, + "grad_norm": 0.99321915369349, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 5672 + }, + { + "epoch": 0.05673, + "grad_norm": 1.0447752766651002, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 5673 + }, + { + "epoch": 0.05674, + "grad_norm": 1.077079251419245, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 5674 + }, + { + "epoch": 0.05675, + "grad_norm": 0.919109346681063, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 5675 + }, + { + "epoch": 0.05676, + "grad_norm": 0.9790547358628968, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 5676 + }, + { + "epoch": 0.05677, + "grad_norm": 1.1793670646509047, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 5677 + }, + { + "epoch": 0.05678, + "grad_norm": 0.7472606166861204, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 5678 + }, + { + "epoch": 0.05679, + "grad_norm": 0.8050508896957376, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 5679 + }, + { + "epoch": 0.0568, + "grad_norm": 0.9344025333558529, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 5680 + }, + { + "epoch": 0.05681, + "grad_norm": 1.2348092828821307, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 5681 + }, + { + "epoch": 0.05682, + "grad_norm": 1.1959381020011484, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 5682 + }, + { + "epoch": 0.05683, + "grad_norm": 0.8976257998161941, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 5683 + }, + { + "epoch": 0.05684, + "grad_norm": 0.7252712065821193, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 5684 + }, + { + "epoch": 0.05685, + "grad_norm": 0.8829690678207763, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 5685 + }, + { + "epoch": 0.05686, + "grad_norm": 1.1058469730277019, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 5686 + }, + { + "epoch": 0.05687, + "grad_norm": 0.9611307746153006, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 5687 + }, + { + "epoch": 0.05688, + "grad_norm": 0.972565534030192, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 5688 + }, + { + "epoch": 0.05689, + "grad_norm": 1.096241549462336, + "learning_rate": 0.003, + "loss": 4.044, + "step": 5689 + }, + { + "epoch": 0.0569, + "grad_norm": 0.8973939245294639, + "learning_rate": 0.003, + "loss": 4.061, + "step": 5690 + }, + { + "epoch": 0.05691, + "grad_norm": 1.1419881519139168, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 5691 + }, + { + "epoch": 0.05692, + "grad_norm": 0.9949113970884143, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 5692 + }, + { + "epoch": 0.05693, + "grad_norm": 0.8512031165359729, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 5693 + }, + { + "epoch": 0.05694, + "grad_norm": 0.9426191438665297, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 5694 + }, + { + "epoch": 0.05695, + "grad_norm": 0.9020092132644848, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 5695 + }, + { + "epoch": 0.05696, + "grad_norm": 0.8485145899081942, + "learning_rate": 0.003, + "loss": 4.053, + "step": 5696 + }, + { + "epoch": 0.05697, + "grad_norm": 0.8561615924770996, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 5697 + }, + { + "epoch": 0.05698, + "grad_norm": 0.8260144341160308, + "learning_rate": 0.003, + "loss": 4.085, + "step": 5698 + }, + { + "epoch": 0.05699, + "grad_norm": 0.966693143570733, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 5699 + }, + { + "epoch": 0.057, + "grad_norm": 1.1487198582029357, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 5700 + }, + { + "epoch": 0.05701, + "grad_norm": 1.0854758797164454, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 5701 + }, + { + "epoch": 0.05702, + "grad_norm": 0.8835018328849502, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 5702 + }, + { + "epoch": 0.05703, + "grad_norm": 0.7845742412565407, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 5703 + }, + { + "epoch": 0.05704, + "grad_norm": 0.8498642388489823, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 5704 + }, + { + "epoch": 0.05705, + "grad_norm": 1.1326574547883799, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 5705 + }, + { + "epoch": 0.05706, + "grad_norm": 1.1434161313327718, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 5706 + }, + { + "epoch": 0.05707, + "grad_norm": 0.8510990859187535, + "learning_rate": 0.003, + "loss": 4.058, + "step": 5707 + }, + { + "epoch": 0.05708, + "grad_norm": 0.8230771160938593, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 5708 + }, + { + "epoch": 0.05709, + "grad_norm": 0.8626308606823649, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 5709 + }, + { + "epoch": 0.0571, + "grad_norm": 0.8749237183120739, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 5710 + }, + { + "epoch": 0.05711, + "grad_norm": 1.0435884220325964, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 5711 + }, + { + "epoch": 0.05712, + "grad_norm": 1.1753101650891566, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 5712 + }, + { + "epoch": 0.05713, + "grad_norm": 0.9392540388426064, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 5713 + }, + { + "epoch": 0.05714, + "grad_norm": 1.1219818951823304, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 5714 + }, + { + "epoch": 0.05715, + "grad_norm": 0.9027198476231381, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 5715 + }, + { + "epoch": 0.05716, + "grad_norm": 0.9350712749494232, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 5716 + }, + { + "epoch": 0.05717, + "grad_norm": 1.078949022904377, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 5717 + }, + { + "epoch": 0.05718, + "grad_norm": 0.9320114628230994, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 5718 + }, + { + "epoch": 0.05719, + "grad_norm": 1.0737725128729791, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 5719 + }, + { + "epoch": 0.0572, + "grad_norm": 1.4370901793089597, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 5720 + }, + { + "epoch": 0.05721, + "grad_norm": 0.6835699725834176, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5721 + }, + { + "epoch": 0.05722, + "grad_norm": 0.9019947466995998, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 5722 + }, + { + "epoch": 0.05723, + "grad_norm": 0.9478261688039891, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 5723 + }, + { + "epoch": 0.05724, + "grad_norm": 0.9063450900670779, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 5724 + }, + { + "epoch": 0.05725, + "grad_norm": 0.8956563510530542, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 5725 + }, + { + "epoch": 0.05726, + "grad_norm": 0.944931440902404, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 5726 + }, + { + "epoch": 0.05727, + "grad_norm": 0.9785206525274993, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 5727 + }, + { + "epoch": 0.05728, + "grad_norm": 0.9750029413661264, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 5728 + }, + { + "epoch": 0.05729, + "grad_norm": 0.9458877993555904, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 5729 + }, + { + "epoch": 0.0573, + "grad_norm": 0.8220018329351471, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 5730 + }, + { + "epoch": 0.05731, + "grad_norm": 0.8935219092820262, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 5731 + }, + { + "epoch": 0.05732, + "grad_norm": 1.011416798805237, + "learning_rate": 0.003, + "loss": 4.083, + "step": 5732 + }, + { + "epoch": 0.05733, + "grad_norm": 1.191217640923215, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 5733 + }, + { + "epoch": 0.05734, + "grad_norm": 0.9117001751188784, + "learning_rate": 0.003, + "loss": 4.087, + "step": 5734 + }, + { + "epoch": 0.05735, + "grad_norm": 0.7434085878525878, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 5735 + }, + { + "epoch": 0.05736, + "grad_norm": 0.7886018670166627, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 5736 + }, + { + "epoch": 0.05737, + "grad_norm": 0.8276223745361921, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 5737 + }, + { + "epoch": 0.05738, + "grad_norm": 0.9579259513547983, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 5738 + }, + { + "epoch": 0.05739, + "grad_norm": 0.9985511509251391, + "learning_rate": 0.003, + "loss": 4.054, + "step": 5739 + }, + { + "epoch": 0.0574, + "grad_norm": 0.9372357444408635, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 5740 + }, + { + "epoch": 0.05741, + "grad_norm": 1.037244120040037, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 5741 + }, + { + "epoch": 0.05742, + "grad_norm": 1.0925704013324562, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 5742 + }, + { + "epoch": 0.05743, + "grad_norm": 0.9822101766913746, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 5743 + }, + { + "epoch": 0.05744, + "grad_norm": 0.9415350376527191, + "learning_rate": 0.003, + "loss": 4.071, + "step": 5744 + }, + { + "epoch": 0.05745, + "grad_norm": 0.8514513969086831, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 5745 + }, + { + "epoch": 0.05746, + "grad_norm": 0.9590813051602415, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 5746 + }, + { + "epoch": 0.05747, + "grad_norm": 1.033811045759274, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5747 + }, + { + "epoch": 0.05748, + "grad_norm": 0.9881694198935187, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 5748 + }, + { + "epoch": 0.05749, + "grad_norm": 1.2197505780445184, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 5749 + }, + { + "epoch": 0.0575, + "grad_norm": 0.8960825057299955, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 5750 + }, + { + "epoch": 0.05751, + "grad_norm": 0.9380756538662388, + "learning_rate": 0.003, + "loss": 4.09, + "step": 5751 + }, + { + "epoch": 0.05752, + "grad_norm": 1.0494754059465994, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 5752 + }, + { + "epoch": 0.05753, + "grad_norm": 0.9737106865451927, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 5753 + }, + { + "epoch": 0.05754, + "grad_norm": 1.14044841129864, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 5754 + }, + { + "epoch": 0.05755, + "grad_norm": 0.9859781548984715, + "learning_rate": 0.003, + "loss": 4.054, + "step": 5755 + }, + { + "epoch": 0.05756, + "grad_norm": 1.1440582623675561, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 5756 + }, + { + "epoch": 0.05757, + "grad_norm": 1.0784582404787597, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 5757 + }, + { + "epoch": 0.05758, + "grad_norm": 1.0238978196194495, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 5758 + }, + { + "epoch": 0.05759, + "grad_norm": 1.1962555833625972, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 5759 + }, + { + "epoch": 0.0576, + "grad_norm": 0.8606928693615246, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 5760 + }, + { + "epoch": 0.05761, + "grad_norm": 0.8065313954609131, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 5761 + }, + { + "epoch": 0.05762, + "grad_norm": 1.006617466336666, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 5762 + }, + { + "epoch": 0.05763, + "grad_norm": 1.1736455692173053, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 5763 + }, + { + "epoch": 0.05764, + "grad_norm": 1.0156358262996825, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 5764 + }, + { + "epoch": 0.05765, + "grad_norm": 1.1676317429304073, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 5765 + }, + { + "epoch": 0.05766, + "grad_norm": 0.9296797123737459, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 5766 + }, + { + "epoch": 0.05767, + "grad_norm": 1.108411398313796, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 5767 + }, + { + "epoch": 0.05768, + "grad_norm": 0.8873418356321205, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 5768 + }, + { + "epoch": 0.05769, + "grad_norm": 0.8943391072499615, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 5769 + }, + { + "epoch": 0.0577, + "grad_norm": 0.7643635046640151, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 5770 + }, + { + "epoch": 0.05771, + "grad_norm": 0.8171732047022162, + "learning_rate": 0.003, + "loss": 4.073, + "step": 5771 + }, + { + "epoch": 0.05772, + "grad_norm": 0.9195825274621192, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 5772 + }, + { + "epoch": 0.05773, + "grad_norm": 1.1643144465255408, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 5773 + }, + { + "epoch": 0.05774, + "grad_norm": 1.0670738905589834, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 5774 + }, + { + "epoch": 0.05775, + "grad_norm": 1.1379995423607696, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 5775 + }, + { + "epoch": 0.05776, + "grad_norm": 1.0629804758835848, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 5776 + }, + { + "epoch": 0.05777, + "grad_norm": 0.9806929321830686, + "learning_rate": 0.003, + "loss": 4.075, + "step": 5777 + }, + { + "epoch": 0.05778, + "grad_norm": 1.0104200584581524, + "learning_rate": 0.003, + "loss": 4.061, + "step": 5778 + }, + { + "epoch": 0.05779, + "grad_norm": 1.0501375318209158, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5779 + }, + { + "epoch": 0.0578, + "grad_norm": 1.0936198888153104, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 5780 + }, + { + "epoch": 0.05781, + "grad_norm": 1.0599398702570635, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 5781 + }, + { + "epoch": 0.05782, + "grad_norm": 0.8124457081867787, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 5782 + }, + { + "epoch": 0.05783, + "grad_norm": 0.8785108285243248, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 5783 + }, + { + "epoch": 0.05784, + "grad_norm": 0.9665672722542055, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 5784 + }, + { + "epoch": 0.05785, + "grad_norm": 1.0852885954128955, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 5785 + }, + { + "epoch": 0.05786, + "grad_norm": 0.9695946116446583, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 5786 + }, + { + "epoch": 0.05787, + "grad_norm": 1.0650628416583132, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 5787 + }, + { + "epoch": 0.05788, + "grad_norm": 0.7594393281247598, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 5788 + }, + { + "epoch": 0.05789, + "grad_norm": 0.7541805830813078, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5789 + }, + { + "epoch": 0.0579, + "grad_norm": 0.8652244813315739, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5790 + }, + { + "epoch": 0.05791, + "grad_norm": 0.8818398092099728, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 5791 + }, + { + "epoch": 0.05792, + "grad_norm": 0.7291575921298588, + "learning_rate": 0.003, + "loss": 4.054, + "step": 5792 + }, + { + "epoch": 0.05793, + "grad_norm": 0.7978555275990008, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 5793 + }, + { + "epoch": 0.05794, + "grad_norm": 0.9171123664036167, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 5794 + }, + { + "epoch": 0.05795, + "grad_norm": 0.8607991972557187, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 5795 + }, + { + "epoch": 0.05796, + "grad_norm": 0.7463660684838923, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 5796 + }, + { + "epoch": 0.05797, + "grad_norm": 0.7284446091050844, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 5797 + }, + { + "epoch": 0.05798, + "grad_norm": 0.9660475572533875, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 5798 + }, + { + "epoch": 0.05799, + "grad_norm": 1.5107968538940575, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 5799 + }, + { + "epoch": 0.058, + "grad_norm": 0.8675842372107159, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 5800 + }, + { + "epoch": 0.05801, + "grad_norm": 1.0182860429769818, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 5801 + }, + { + "epoch": 0.05802, + "grad_norm": 0.8847441154270007, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 5802 + }, + { + "epoch": 0.05803, + "grad_norm": 0.9726482455890011, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 5803 + }, + { + "epoch": 0.05804, + "grad_norm": 1.0017317638615917, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 5804 + }, + { + "epoch": 0.05805, + "grad_norm": 1.1460034384721858, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 5805 + }, + { + "epoch": 0.05806, + "grad_norm": 0.8265329809662282, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 5806 + }, + { + "epoch": 0.05807, + "grad_norm": 0.7999918311654125, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 5807 + }, + { + "epoch": 0.05808, + "grad_norm": 0.7978362275046719, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 5808 + }, + { + "epoch": 0.05809, + "grad_norm": 0.8854482959350243, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 5809 + }, + { + "epoch": 0.0581, + "grad_norm": 1.0871463000956632, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 5810 + }, + { + "epoch": 0.05811, + "grad_norm": 1.1086945676399056, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 5811 + }, + { + "epoch": 0.05812, + "grad_norm": 0.9304552613092574, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 5812 + }, + { + "epoch": 0.05813, + "grad_norm": 0.9983452335160318, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 5813 + }, + { + "epoch": 0.05814, + "grad_norm": 0.9975951286224842, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 5814 + }, + { + "epoch": 0.05815, + "grad_norm": 0.8924647102734461, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 5815 + }, + { + "epoch": 0.05816, + "grad_norm": 0.9225284620016561, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 5816 + }, + { + "epoch": 0.05817, + "grad_norm": 0.9979994283490636, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 5817 + }, + { + "epoch": 0.05818, + "grad_norm": 1.0439768447035485, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 5818 + }, + { + "epoch": 0.05819, + "grad_norm": 1.033649032838979, + "learning_rate": 0.003, + "loss": 4.06, + "step": 5819 + }, + { + "epoch": 0.0582, + "grad_norm": 1.1096343593879232, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 5820 + }, + { + "epoch": 0.05821, + "grad_norm": 0.9004393239006013, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 5821 + }, + { + "epoch": 0.05822, + "grad_norm": 1.0863173751230966, + "learning_rate": 0.003, + "loss": 4.043, + "step": 5822 + }, + { + "epoch": 0.05823, + "grad_norm": 1.0599324875126475, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 5823 + }, + { + "epoch": 0.05824, + "grad_norm": 0.8876300345105752, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 5824 + }, + { + "epoch": 0.05825, + "grad_norm": 1.0008579052062283, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 5825 + }, + { + "epoch": 0.05826, + "grad_norm": 1.1748225539903427, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 5826 + }, + { + "epoch": 0.05827, + "grad_norm": 0.9593447519304307, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5827 + }, + { + "epoch": 0.05828, + "grad_norm": 1.0225296000471948, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 5828 + }, + { + "epoch": 0.05829, + "grad_norm": 1.0560219694488389, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 5829 + }, + { + "epoch": 0.0583, + "grad_norm": 0.9860898705161171, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 5830 + }, + { + "epoch": 0.05831, + "grad_norm": 1.1250031799560922, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 5831 + }, + { + "epoch": 0.05832, + "grad_norm": 1.157236041461952, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 5832 + }, + { + "epoch": 0.05833, + "grad_norm": 0.978478470266213, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 5833 + }, + { + "epoch": 0.05834, + "grad_norm": 0.9798668444646972, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 5834 + }, + { + "epoch": 0.05835, + "grad_norm": 0.7832001924434391, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 5835 + }, + { + "epoch": 0.05836, + "grad_norm": 0.8249432319588443, + "learning_rate": 0.003, + "loss": 4.074, + "step": 5836 + }, + { + "epoch": 0.05837, + "grad_norm": 0.9729603126849787, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 5837 + }, + { + "epoch": 0.05838, + "grad_norm": 0.9502514262519923, + "learning_rate": 0.003, + "loss": 4.076, + "step": 5838 + }, + { + "epoch": 0.05839, + "grad_norm": 0.9267029498844601, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 5839 + }, + { + "epoch": 0.0584, + "grad_norm": 0.9657962011078538, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 5840 + }, + { + "epoch": 0.05841, + "grad_norm": 1.0401049872356143, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 5841 + }, + { + "epoch": 0.05842, + "grad_norm": 1.1291241708881259, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 5842 + }, + { + "epoch": 0.05843, + "grad_norm": 1.0657328740989145, + "learning_rate": 0.003, + "loss": 4.07, + "step": 5843 + }, + { + "epoch": 0.05844, + "grad_norm": 0.936956030075049, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 5844 + }, + { + "epoch": 0.05845, + "grad_norm": 0.954122564913459, + "learning_rate": 0.003, + "loss": 4.063, + "step": 5845 + }, + { + "epoch": 0.05846, + "grad_norm": 1.0055872032953703, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 5846 + }, + { + "epoch": 0.05847, + "grad_norm": 1.0847001970093764, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 5847 + }, + { + "epoch": 0.05848, + "grad_norm": 0.9328803987155841, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 5848 + }, + { + "epoch": 0.05849, + "grad_norm": 1.0216731452531402, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 5849 + }, + { + "epoch": 0.0585, + "grad_norm": 1.042034861189281, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 5850 + }, + { + "epoch": 0.05851, + "grad_norm": 0.9696375907305427, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 5851 + }, + { + "epoch": 0.05852, + "grad_norm": 0.9548659469705912, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 5852 + }, + { + "epoch": 0.05853, + "grad_norm": 0.9934590507196531, + "learning_rate": 0.003, + "loss": 4.045, + "step": 5853 + }, + { + "epoch": 0.05854, + "grad_norm": 1.109837806391397, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 5854 + }, + { + "epoch": 0.05855, + "grad_norm": 0.9612662531040636, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5855 + }, + { + "epoch": 0.05856, + "grad_norm": 0.9262688978815322, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 5856 + }, + { + "epoch": 0.05857, + "grad_norm": 1.0672841758813956, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 5857 + }, + { + "epoch": 0.05858, + "grad_norm": 0.9924933017958931, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 5858 + }, + { + "epoch": 0.05859, + "grad_norm": 1.1446711154327815, + "learning_rate": 0.003, + "loss": 4.076, + "step": 5859 + }, + { + "epoch": 0.0586, + "grad_norm": 0.9395157838833672, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 5860 + }, + { + "epoch": 0.05861, + "grad_norm": 1.017783268004518, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 5861 + }, + { + "epoch": 0.05862, + "grad_norm": 0.8461181473215044, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 5862 + }, + { + "epoch": 0.05863, + "grad_norm": 0.727622134447459, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 5863 + }, + { + "epoch": 0.05864, + "grad_norm": 0.7787266994865726, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 5864 + }, + { + "epoch": 0.05865, + "grad_norm": 0.9457836140127373, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 5865 + }, + { + "epoch": 0.05866, + "grad_norm": 1.1590380983819644, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5866 + }, + { + "epoch": 0.05867, + "grad_norm": 0.8661814704210331, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 5867 + }, + { + "epoch": 0.05868, + "grad_norm": 0.7841997946264219, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 5868 + }, + { + "epoch": 0.05869, + "grad_norm": 0.9021971418732975, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 5869 + }, + { + "epoch": 0.0587, + "grad_norm": 1.044429489265089, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5870 + }, + { + "epoch": 0.05871, + "grad_norm": 0.9528123707330137, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 5871 + }, + { + "epoch": 0.05872, + "grad_norm": 1.0247575002414453, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 5872 + }, + { + "epoch": 0.05873, + "grad_norm": 1.1855135281087743, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 5873 + }, + { + "epoch": 0.05874, + "grad_norm": 0.9873466030307598, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 5874 + }, + { + "epoch": 0.05875, + "grad_norm": 1.1663397858316737, + "learning_rate": 0.003, + "loss": 4.078, + "step": 5875 + }, + { + "epoch": 0.05876, + "grad_norm": 1.2258462137763335, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 5876 + }, + { + "epoch": 0.05877, + "grad_norm": 1.0386697354012882, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 5877 + }, + { + "epoch": 0.05878, + "grad_norm": 1.183742461564236, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 5878 + }, + { + "epoch": 0.05879, + "grad_norm": 0.940962143351869, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 5879 + }, + { + "epoch": 0.0588, + "grad_norm": 0.927043037474344, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 5880 + }, + { + "epoch": 0.05881, + "grad_norm": 0.8962578128398271, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5881 + }, + { + "epoch": 0.05882, + "grad_norm": 0.9066915820041124, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 5882 + }, + { + "epoch": 0.05883, + "grad_norm": 0.8598805143317192, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 5883 + }, + { + "epoch": 0.05884, + "grad_norm": 1.0114326319466251, + "learning_rate": 0.003, + "loss": 4.063, + "step": 5884 + }, + { + "epoch": 0.05885, + "grad_norm": 1.1324207199773164, + "learning_rate": 0.003, + "loss": 4.084, + "step": 5885 + }, + { + "epoch": 0.05886, + "grad_norm": 0.9079205076328346, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 5886 + }, + { + "epoch": 0.05887, + "grad_norm": 0.9499292804842742, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 5887 + }, + { + "epoch": 0.05888, + "grad_norm": 0.9253742487047153, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 5888 + }, + { + "epoch": 0.05889, + "grad_norm": 0.9756593540292708, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 5889 + }, + { + "epoch": 0.0589, + "grad_norm": 1.2153968005149314, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 5890 + }, + { + "epoch": 0.05891, + "grad_norm": 1.1050096751560372, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 5891 + }, + { + "epoch": 0.05892, + "grad_norm": 1.0123025142968227, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 5892 + }, + { + "epoch": 0.05893, + "grad_norm": 0.9835328234583312, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 5893 + }, + { + "epoch": 0.05894, + "grad_norm": 0.9304265933341809, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 5894 + }, + { + "epoch": 0.05895, + "grad_norm": 0.9274436106212893, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 5895 + }, + { + "epoch": 0.05896, + "grad_norm": 0.8771867029491986, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 5896 + }, + { + "epoch": 0.05897, + "grad_norm": 0.8218952866171018, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 5897 + }, + { + "epoch": 0.05898, + "grad_norm": 0.9430573967743275, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5898 + }, + { + "epoch": 0.05899, + "grad_norm": 0.8911678825157316, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 5899 + }, + { + "epoch": 0.059, + "grad_norm": 0.9174879047193627, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 5900 + }, + { + "epoch": 0.05901, + "grad_norm": 0.9824565061579317, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 5901 + }, + { + "epoch": 0.05902, + "grad_norm": 1.240010288158423, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 5902 + }, + { + "epoch": 0.05903, + "grad_norm": 0.7639178515927444, + "learning_rate": 0.003, + "loss": 4.073, + "step": 5903 + }, + { + "epoch": 0.05904, + "grad_norm": 0.770306018643173, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 5904 + }, + { + "epoch": 0.05905, + "grad_norm": 0.976148342284057, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 5905 + }, + { + "epoch": 0.05906, + "grad_norm": 1.1812920524478123, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 5906 + }, + { + "epoch": 0.05907, + "grad_norm": 1.1654915411869895, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 5907 + }, + { + "epoch": 0.05908, + "grad_norm": 1.120465653132924, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 5908 + }, + { + "epoch": 0.05909, + "grad_norm": 0.8539512494382139, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5909 + }, + { + "epoch": 0.0591, + "grad_norm": 0.8227369203330751, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5910 + }, + { + "epoch": 0.05911, + "grad_norm": 0.8906442288033303, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 5911 + }, + { + "epoch": 0.05912, + "grad_norm": 0.9509533493887127, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 5912 + }, + { + "epoch": 0.05913, + "grad_norm": 1.0981979298361981, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 5913 + }, + { + "epoch": 0.05914, + "grad_norm": 0.9415620615571133, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 5914 + }, + { + "epoch": 0.05915, + "grad_norm": 0.8478332715846145, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5915 + }, + { + "epoch": 0.05916, + "grad_norm": 0.8256880500590459, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 5916 + }, + { + "epoch": 0.05917, + "grad_norm": 0.9790506827233469, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 5917 + }, + { + "epoch": 0.05918, + "grad_norm": 1.0353580913196638, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 5918 + }, + { + "epoch": 0.05919, + "grad_norm": 0.9651964696003633, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5919 + }, + { + "epoch": 0.0592, + "grad_norm": 0.937291761754371, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 5920 + }, + { + "epoch": 0.05921, + "grad_norm": 1.0990872808141368, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5921 + }, + { + "epoch": 0.05922, + "grad_norm": 1.0468886911915498, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 5922 + }, + { + "epoch": 0.05923, + "grad_norm": 1.0323600516250746, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 5923 + }, + { + "epoch": 0.05924, + "grad_norm": 0.954016613180186, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 5924 + }, + { + "epoch": 0.05925, + "grad_norm": 0.9475491801291553, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 5925 + }, + { + "epoch": 0.05926, + "grad_norm": 0.9668145623340975, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 5926 + }, + { + "epoch": 0.05927, + "grad_norm": 1.0595618769436952, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 5927 + }, + { + "epoch": 0.05928, + "grad_norm": 0.9855002059444532, + "learning_rate": 0.003, + "loss": 4.065, + "step": 5928 + }, + { + "epoch": 0.05929, + "grad_norm": 0.945307973831574, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 5929 + }, + { + "epoch": 0.0593, + "grad_norm": 0.9657299567775556, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5930 + }, + { + "epoch": 0.05931, + "grad_norm": 0.9827228555033615, + "learning_rate": 0.003, + "loss": 4.058, + "step": 5931 + }, + { + "epoch": 0.05932, + "grad_norm": 1.156077533419125, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 5932 + }, + { + "epoch": 0.05933, + "grad_norm": 0.847492777721161, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 5933 + }, + { + "epoch": 0.05934, + "grad_norm": 1.0021937329107191, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 5934 + }, + { + "epoch": 0.05935, + "grad_norm": 1.1159732242811857, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 5935 + }, + { + "epoch": 0.05936, + "grad_norm": 0.899401377373719, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 5936 + }, + { + "epoch": 0.05937, + "grad_norm": 0.8301027331575037, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 5937 + }, + { + "epoch": 0.05938, + "grad_norm": 0.8762725606178311, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 5938 + }, + { + "epoch": 0.05939, + "grad_norm": 0.9126590668115435, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 5939 + }, + { + "epoch": 0.0594, + "grad_norm": 1.0407767907462304, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 5940 + }, + { + "epoch": 0.05941, + "grad_norm": 1.1143892345234092, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 5941 + }, + { + "epoch": 0.05942, + "grad_norm": 1.0961434922540771, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5942 + }, + { + "epoch": 0.05943, + "grad_norm": 1.0471390550371067, + "learning_rate": 0.003, + "loss": 4.061, + "step": 5943 + }, + { + "epoch": 0.05944, + "grad_norm": 0.9579622530263362, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 5944 + }, + { + "epoch": 0.05945, + "grad_norm": 0.9926199469417607, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 5945 + }, + { + "epoch": 0.05946, + "grad_norm": 1.0156252547795441, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 5946 + }, + { + "epoch": 0.05947, + "grad_norm": 1.2611887927782788, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 5947 + }, + { + "epoch": 0.05948, + "grad_norm": 0.9815730257128612, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 5948 + }, + { + "epoch": 0.05949, + "grad_norm": 1.0471598623999925, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 5949 + }, + { + "epoch": 0.0595, + "grad_norm": 0.8842194048874816, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 5950 + }, + { + "epoch": 0.05951, + "grad_norm": 0.8826262827071101, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 5951 + }, + { + "epoch": 0.05952, + "grad_norm": 0.8874639709587795, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5952 + }, + { + "epoch": 0.05953, + "grad_norm": 1.094346129739514, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 5953 + }, + { + "epoch": 0.05954, + "grad_norm": 1.0405342399228923, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 5954 + }, + { + "epoch": 0.05955, + "grad_norm": 1.0084545393678892, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 5955 + }, + { + "epoch": 0.05956, + "grad_norm": 0.8681329007495994, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 5956 + }, + { + "epoch": 0.05957, + "grad_norm": 0.9787011417999288, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 5957 + }, + { + "epoch": 0.05958, + "grad_norm": 0.983255913377767, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 5958 + }, + { + "epoch": 0.05959, + "grad_norm": 0.9397652065747355, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 5959 + }, + { + "epoch": 0.0596, + "grad_norm": 0.9197247886041665, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 5960 + }, + { + "epoch": 0.05961, + "grad_norm": 0.975726486158997, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 5961 + }, + { + "epoch": 0.05962, + "grad_norm": 1.0416902503797958, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 5962 + }, + { + "epoch": 0.05963, + "grad_norm": 1.0805802820393233, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 5963 + }, + { + "epoch": 0.05964, + "grad_norm": 0.9125810985534502, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 5964 + }, + { + "epoch": 0.05965, + "grad_norm": 0.7839568155281216, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 5965 + }, + { + "epoch": 0.05966, + "grad_norm": 0.645579903658523, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 5966 + }, + { + "epoch": 0.05967, + "grad_norm": 0.6174059038324572, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 5967 + }, + { + "epoch": 0.05968, + "grad_norm": 0.6892480108767447, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 5968 + }, + { + "epoch": 0.05969, + "grad_norm": 0.7254045272144456, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 5969 + }, + { + "epoch": 0.0597, + "grad_norm": 0.7953745734437678, + "learning_rate": 0.003, + "loss": 4.043, + "step": 5970 + }, + { + "epoch": 0.05971, + "grad_norm": 1.0581208502754886, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 5971 + }, + { + "epoch": 0.05972, + "grad_norm": 1.1798253869062356, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 5972 + }, + { + "epoch": 0.05973, + "grad_norm": 0.7544716901026084, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 5973 + }, + { + "epoch": 0.05974, + "grad_norm": 0.7411406547879388, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 5974 + }, + { + "epoch": 0.05975, + "grad_norm": 0.664439857328635, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 5975 + }, + { + "epoch": 0.05976, + "grad_norm": 0.684227141132107, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5976 + }, + { + "epoch": 0.05977, + "grad_norm": 0.8888845505907672, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 5977 + }, + { + "epoch": 0.05978, + "grad_norm": 1.1416281876950884, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 5978 + }, + { + "epoch": 0.05979, + "grad_norm": 1.008454254020505, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5979 + }, + { + "epoch": 0.0598, + "grad_norm": 1.190632055338276, + "learning_rate": 0.003, + "loss": 4.045, + "step": 5980 + }, + { + "epoch": 0.05981, + "grad_norm": 1.015356754441631, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 5981 + }, + { + "epoch": 0.05982, + "grad_norm": 0.9378130097416635, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 5982 + }, + { + "epoch": 0.05983, + "grad_norm": 0.8053048489511819, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 5983 + }, + { + "epoch": 0.05984, + "grad_norm": 0.9819349814850642, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 5984 + }, + { + "epoch": 0.05985, + "grad_norm": 1.1131918756985963, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 5985 + }, + { + "epoch": 0.05986, + "grad_norm": 0.9174366752153941, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 5986 + }, + { + "epoch": 0.05987, + "grad_norm": 1.0378256913235493, + "learning_rate": 0.003, + "loss": 4.049, + "step": 5987 + }, + { + "epoch": 0.05988, + "grad_norm": 0.9258303467488307, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 5988 + }, + { + "epoch": 0.05989, + "grad_norm": 0.9632304797304597, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 5989 + }, + { + "epoch": 0.0599, + "grad_norm": 0.8935903017412864, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 5990 + }, + { + "epoch": 0.05991, + "grad_norm": 0.6739742445063621, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 5991 + }, + { + "epoch": 0.05992, + "grad_norm": 0.5915614354430911, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 5992 + }, + { + "epoch": 0.05993, + "grad_norm": 0.741024166202095, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5993 + }, + { + "epoch": 0.05994, + "grad_norm": 0.8841179715786665, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 5994 + }, + { + "epoch": 0.05995, + "grad_norm": 1.287538920218187, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 5995 + }, + { + "epoch": 0.05996, + "grad_norm": 1.0684098332314382, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 5996 + }, + { + "epoch": 0.05997, + "grad_norm": 0.9397674955044001, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 5997 + }, + { + "epoch": 0.05998, + "grad_norm": 0.8272889396102959, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5998 + }, + { + "epoch": 0.05999, + "grad_norm": 0.8991148255094739, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 5999 + }, + { + "epoch": 0.06, + "grad_norm": 0.9185220408167984, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 6000 + }, + { + "epoch": 0.06001, + "grad_norm": 0.9908154950299144, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 6001 + }, + { + "epoch": 0.06002, + "grad_norm": 0.9608103918830759, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 6002 + }, + { + "epoch": 0.06003, + "grad_norm": 1.1215832189472665, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 6003 + }, + { + "epoch": 0.06004, + "grad_norm": 1.0570354371095005, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 6004 + }, + { + "epoch": 0.06005, + "grad_norm": 1.0877288777978935, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 6005 + }, + { + "epoch": 0.06006, + "grad_norm": 1.0530975014128934, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 6006 + }, + { + "epoch": 0.06007, + "grad_norm": 0.8755728290132648, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 6007 + }, + { + "epoch": 0.06008, + "grad_norm": 0.9019088575660325, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 6008 + }, + { + "epoch": 0.06009, + "grad_norm": 1.1772437890538816, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 6009 + }, + { + "epoch": 0.0601, + "grad_norm": 0.9527109892195108, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 6010 + }, + { + "epoch": 0.06011, + "grad_norm": 1.1738757885020588, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 6011 + }, + { + "epoch": 0.06012, + "grad_norm": 0.8780691584223036, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 6012 + }, + { + "epoch": 0.06013, + "grad_norm": 0.9179139070213036, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 6013 + }, + { + "epoch": 0.06014, + "grad_norm": 1.00855448654704, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 6014 + }, + { + "epoch": 0.06015, + "grad_norm": 0.9972560180447895, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 6015 + }, + { + "epoch": 0.06016, + "grad_norm": 1.0775312684009026, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 6016 + }, + { + "epoch": 0.06017, + "grad_norm": 1.33609316921924, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 6017 + }, + { + "epoch": 0.06018, + "grad_norm": 1.0280471651185956, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 6018 + }, + { + "epoch": 0.06019, + "grad_norm": 1.0585602447833204, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 6019 + }, + { + "epoch": 0.0602, + "grad_norm": 0.9692400850494943, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 6020 + }, + { + "epoch": 0.06021, + "grad_norm": 1.0415586611820176, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 6021 + }, + { + "epoch": 0.06022, + "grad_norm": 1.055493914814788, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 6022 + }, + { + "epoch": 0.06023, + "grad_norm": 1.1437368963667776, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 6023 + }, + { + "epoch": 0.06024, + "grad_norm": 1.0461727220486359, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 6024 + }, + { + "epoch": 0.06025, + "grad_norm": 0.8495924516845805, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 6025 + }, + { + "epoch": 0.06026, + "grad_norm": 0.8387211090444705, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 6026 + }, + { + "epoch": 0.06027, + "grad_norm": 0.9070801722224514, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 6027 + }, + { + "epoch": 0.06028, + "grad_norm": 0.9625380625910851, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 6028 + }, + { + "epoch": 0.06029, + "grad_norm": 1.0682457520452102, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 6029 + }, + { + "epoch": 0.0603, + "grad_norm": 0.9976915475426508, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 6030 + }, + { + "epoch": 0.06031, + "grad_norm": 1.0453095190042574, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 6031 + }, + { + "epoch": 0.06032, + "grad_norm": 0.9786653778595836, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 6032 + }, + { + "epoch": 0.06033, + "grad_norm": 1.0968968530547012, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 6033 + }, + { + "epoch": 0.06034, + "grad_norm": 0.9514397371995961, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 6034 + }, + { + "epoch": 0.06035, + "grad_norm": 1.0541817557308053, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 6035 + }, + { + "epoch": 0.06036, + "grad_norm": 0.8780263970411091, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 6036 + }, + { + "epoch": 0.06037, + "grad_norm": 1.0193481447935269, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 6037 + }, + { + "epoch": 0.06038, + "grad_norm": 1.3774676565861503, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 6038 + }, + { + "epoch": 0.06039, + "grad_norm": 0.7252838059348526, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 6039 + }, + { + "epoch": 0.0604, + "grad_norm": 0.78601742451761, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 6040 + }, + { + "epoch": 0.06041, + "grad_norm": 0.744786447804636, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 6041 + }, + { + "epoch": 0.06042, + "grad_norm": 0.8660278386592278, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 6042 + }, + { + "epoch": 0.06043, + "grad_norm": 1.1071573404697885, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 6043 + }, + { + "epoch": 0.06044, + "grad_norm": 1.0894101622044046, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 6044 + }, + { + "epoch": 0.06045, + "grad_norm": 1.0493254983595819, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 6045 + }, + { + "epoch": 0.06046, + "grad_norm": 1.0049917198936318, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 6046 + }, + { + "epoch": 0.06047, + "grad_norm": 1.0517277702815098, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 6047 + }, + { + "epoch": 0.06048, + "grad_norm": 1.1809116249025131, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 6048 + }, + { + "epoch": 0.06049, + "grad_norm": 0.9623062437443599, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 6049 + }, + { + "epoch": 0.0605, + "grad_norm": 1.0923522103941121, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 6050 + }, + { + "epoch": 0.06051, + "grad_norm": 1.0265427925273334, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 6051 + }, + { + "epoch": 0.06052, + "grad_norm": 0.9414887816443991, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 6052 + }, + { + "epoch": 0.06053, + "grad_norm": 1.006065905595286, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 6053 + }, + { + "epoch": 0.06054, + "grad_norm": 1.095910787052634, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 6054 + }, + { + "epoch": 0.06055, + "grad_norm": 0.8517266184871414, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 6055 + }, + { + "epoch": 0.06056, + "grad_norm": 1.1390125640686093, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 6056 + }, + { + "epoch": 0.06057, + "grad_norm": 1.0285564034956292, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 6057 + }, + { + "epoch": 0.06058, + "grad_norm": 1.0733695934055711, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 6058 + }, + { + "epoch": 0.06059, + "grad_norm": 0.8565845158282416, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 6059 + }, + { + "epoch": 0.0606, + "grad_norm": 0.8864003378041345, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6060 + }, + { + "epoch": 0.06061, + "grad_norm": 1.0109993363665308, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 6061 + }, + { + "epoch": 0.06062, + "grad_norm": 1.123107920080809, + "learning_rate": 0.003, + "loss": 4.083, + "step": 6062 + }, + { + "epoch": 0.06063, + "grad_norm": 0.949609814613259, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 6063 + }, + { + "epoch": 0.06064, + "grad_norm": 0.9981845823778703, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 6064 + }, + { + "epoch": 0.06065, + "grad_norm": 1.1017349391298685, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 6065 + }, + { + "epoch": 0.06066, + "grad_norm": 1.0732527006387818, + "learning_rate": 0.003, + "loss": 4.098, + "step": 6066 + }, + { + "epoch": 0.06067, + "grad_norm": 0.9294025545570336, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 6067 + }, + { + "epoch": 0.06068, + "grad_norm": 1.0576231179156892, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 6068 + }, + { + "epoch": 0.06069, + "grad_norm": 1.0865552407589456, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 6069 + }, + { + "epoch": 0.0607, + "grad_norm": 1.0230751617256044, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 6070 + }, + { + "epoch": 0.06071, + "grad_norm": 0.999221755293697, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 6071 + }, + { + "epoch": 0.06072, + "grad_norm": 1.066655819343783, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 6072 + }, + { + "epoch": 0.06073, + "grad_norm": 1.0369526156758926, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 6073 + }, + { + "epoch": 0.06074, + "grad_norm": 1.075598582264553, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 6074 + }, + { + "epoch": 0.06075, + "grad_norm": 0.9953552796029778, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 6075 + }, + { + "epoch": 0.06076, + "grad_norm": 0.9778883253174944, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 6076 + }, + { + "epoch": 0.06077, + "grad_norm": 0.8468031501573927, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 6077 + }, + { + "epoch": 0.06078, + "grad_norm": 0.8305408752084648, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 6078 + }, + { + "epoch": 0.06079, + "grad_norm": 0.7960215844617263, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 6079 + }, + { + "epoch": 0.0608, + "grad_norm": 0.8359710386518069, + "learning_rate": 0.003, + "loss": 4.059, + "step": 6080 + }, + { + "epoch": 0.06081, + "grad_norm": 1.0786121941538231, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 6081 + }, + { + "epoch": 0.06082, + "grad_norm": 1.1823183295621211, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 6082 + }, + { + "epoch": 0.06083, + "grad_norm": 0.8837771871886293, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 6083 + }, + { + "epoch": 0.06084, + "grad_norm": 0.7204472009727627, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 6084 + }, + { + "epoch": 0.06085, + "grad_norm": 0.7103714572162915, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 6085 + }, + { + "epoch": 0.06086, + "grad_norm": 0.811997878307769, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 6086 + }, + { + "epoch": 0.06087, + "grad_norm": 0.8073780584157964, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6087 + }, + { + "epoch": 0.06088, + "grad_norm": 0.8544463123489077, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 6088 + }, + { + "epoch": 0.06089, + "grad_norm": 0.9762980822164021, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 6089 + }, + { + "epoch": 0.0609, + "grad_norm": 1.2142304992327353, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 6090 + }, + { + "epoch": 0.06091, + "grad_norm": 1.082695490566579, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 6091 + }, + { + "epoch": 0.06092, + "grad_norm": 0.9278093493414128, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 6092 + }, + { + "epoch": 0.06093, + "grad_norm": 0.8690782403614105, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 6093 + }, + { + "epoch": 0.06094, + "grad_norm": 0.8570778358605303, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 6094 + }, + { + "epoch": 0.06095, + "grad_norm": 0.8138189413941755, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 6095 + }, + { + "epoch": 0.06096, + "grad_norm": 0.9634421284092561, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 6096 + }, + { + "epoch": 0.06097, + "grad_norm": 1.0632007099690335, + "learning_rate": 0.003, + "loss": 4.071, + "step": 6097 + }, + { + "epoch": 0.06098, + "grad_norm": 1.0871230047693712, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 6098 + }, + { + "epoch": 0.06099, + "grad_norm": 1.0812665303937632, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 6099 + }, + { + "epoch": 0.061, + "grad_norm": 0.9904572164775792, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 6100 + }, + { + "epoch": 0.06101, + "grad_norm": 1.2471241797218284, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 6101 + }, + { + "epoch": 0.06102, + "grad_norm": 0.9331918176411176, + "learning_rate": 0.003, + "loss": 4.066, + "step": 6102 + }, + { + "epoch": 0.06103, + "grad_norm": 1.0358350532613663, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 6103 + }, + { + "epoch": 0.06104, + "grad_norm": 0.9863138814409834, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 6104 + }, + { + "epoch": 0.06105, + "grad_norm": 1.2154265908459023, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 6105 + }, + { + "epoch": 0.06106, + "grad_norm": 0.8917304789437333, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 6106 + }, + { + "epoch": 0.06107, + "grad_norm": 0.9427677486805567, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 6107 + }, + { + "epoch": 0.06108, + "grad_norm": 0.9771691970851738, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 6108 + }, + { + "epoch": 0.06109, + "grad_norm": 1.0200829167924297, + "learning_rate": 0.003, + "loss": 4.034, + "step": 6109 + }, + { + "epoch": 0.0611, + "grad_norm": 1.0829855702592046, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 6110 + }, + { + "epoch": 0.06111, + "grad_norm": 1.11913249179738, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 6111 + }, + { + "epoch": 0.06112, + "grad_norm": 0.9956612041062181, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 6112 + }, + { + "epoch": 0.06113, + "grad_norm": 1.1071104171417576, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 6113 + }, + { + "epoch": 0.06114, + "grad_norm": 0.9095207699385705, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 6114 + }, + { + "epoch": 0.06115, + "grad_norm": 1.044965318613425, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 6115 + }, + { + "epoch": 0.06116, + "grad_norm": 1.2061603685848585, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 6116 + }, + { + "epoch": 0.06117, + "grad_norm": 0.9331914752531237, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6117 + }, + { + "epoch": 0.06118, + "grad_norm": 0.9125472343562069, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 6118 + }, + { + "epoch": 0.06119, + "grad_norm": 0.7893638895463538, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 6119 + }, + { + "epoch": 0.0612, + "grad_norm": 0.9244864031831111, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 6120 + }, + { + "epoch": 0.06121, + "grad_norm": 0.9033631978733891, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 6121 + }, + { + "epoch": 0.06122, + "grad_norm": 1.0308394729354522, + "learning_rate": 0.003, + "loss": 4.072, + "step": 6122 + }, + { + "epoch": 0.06123, + "grad_norm": 0.9549347100846058, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 6123 + }, + { + "epoch": 0.06124, + "grad_norm": 1.018979699685598, + "learning_rate": 0.003, + "loss": 4.062, + "step": 6124 + }, + { + "epoch": 0.06125, + "grad_norm": 1.0396926470286123, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 6125 + }, + { + "epoch": 0.06126, + "grad_norm": 0.970338770093608, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 6126 + }, + { + "epoch": 0.06127, + "grad_norm": 0.9481470691411102, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 6127 + }, + { + "epoch": 0.06128, + "grad_norm": 1.0848389247188306, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 6128 + }, + { + "epoch": 0.06129, + "grad_norm": 1.0580408480527546, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 6129 + }, + { + "epoch": 0.0613, + "grad_norm": 0.9682253621336249, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 6130 + }, + { + "epoch": 0.06131, + "grad_norm": 0.9210655524658186, + "learning_rate": 0.003, + "loss": 4.064, + "step": 6131 + }, + { + "epoch": 0.06132, + "grad_norm": 0.9155041280961395, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 6132 + }, + { + "epoch": 0.06133, + "grad_norm": 0.9789482932440452, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6133 + }, + { + "epoch": 0.06134, + "grad_norm": 1.1135362136558498, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 6134 + }, + { + "epoch": 0.06135, + "grad_norm": 1.0670150504515052, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 6135 + }, + { + "epoch": 0.06136, + "grad_norm": 1.009306468929253, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 6136 + }, + { + "epoch": 0.06137, + "grad_norm": 1.1199332436482432, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 6137 + }, + { + "epoch": 0.06138, + "grad_norm": 0.9822960858920264, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 6138 + }, + { + "epoch": 0.06139, + "grad_norm": 1.061123881008166, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 6139 + }, + { + "epoch": 0.0614, + "grad_norm": 1.1067582756640688, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 6140 + }, + { + "epoch": 0.06141, + "grad_norm": 1.1197623616073298, + "learning_rate": 0.003, + "loss": 4.042, + "step": 6141 + }, + { + "epoch": 0.06142, + "grad_norm": 0.9621463377982679, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 6142 + }, + { + "epoch": 0.06143, + "grad_norm": 0.9823524748170486, + "learning_rate": 0.003, + "loss": 4.068, + "step": 6143 + }, + { + "epoch": 0.06144, + "grad_norm": 1.1256975016048572, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6144 + }, + { + "epoch": 0.06145, + "grad_norm": 1.0964390916024676, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 6145 + }, + { + "epoch": 0.06146, + "grad_norm": 0.9340305536724448, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 6146 + }, + { + "epoch": 0.06147, + "grad_norm": 0.9639235550819623, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 6147 + }, + { + "epoch": 0.06148, + "grad_norm": 1.1141344848316928, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 6148 + }, + { + "epoch": 0.06149, + "grad_norm": 1.0008166080107452, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 6149 + }, + { + "epoch": 0.0615, + "grad_norm": 1.291300378500113, + "learning_rate": 0.003, + "loss": 4.077, + "step": 6150 + }, + { + "epoch": 0.06151, + "grad_norm": 0.9179128052204282, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 6151 + }, + { + "epoch": 0.06152, + "grad_norm": 0.83811806335343, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 6152 + }, + { + "epoch": 0.06153, + "grad_norm": 1.0325588836325448, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 6153 + }, + { + "epoch": 0.06154, + "grad_norm": 1.0531062692602984, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 6154 + }, + { + "epoch": 0.06155, + "grad_norm": 0.9258868598309281, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 6155 + }, + { + "epoch": 0.06156, + "grad_norm": 0.8537558991823246, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 6156 + }, + { + "epoch": 0.06157, + "grad_norm": 0.915762781550664, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 6157 + }, + { + "epoch": 0.06158, + "grad_norm": 0.8548951465188689, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 6158 + }, + { + "epoch": 0.06159, + "grad_norm": 0.8765672907860784, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 6159 + }, + { + "epoch": 0.0616, + "grad_norm": 0.9374358381023599, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 6160 + }, + { + "epoch": 0.06161, + "grad_norm": 1.1299972768267692, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 6161 + }, + { + "epoch": 0.06162, + "grad_norm": 0.9317143658408135, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6162 + }, + { + "epoch": 0.06163, + "grad_norm": 0.8061151783232723, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 6163 + }, + { + "epoch": 0.06164, + "grad_norm": 0.7615643215210022, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 6164 + }, + { + "epoch": 0.06165, + "grad_norm": 0.7769881554996289, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 6165 + }, + { + "epoch": 0.06166, + "grad_norm": 0.8085532275590331, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 6166 + }, + { + "epoch": 0.06167, + "grad_norm": 0.855796871730468, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 6167 + }, + { + "epoch": 0.06168, + "grad_norm": 1.138731517223366, + "learning_rate": 0.003, + "loss": 4.065, + "step": 6168 + }, + { + "epoch": 0.06169, + "grad_norm": 1.368396718368792, + "learning_rate": 0.003, + "loss": 4.068, + "step": 6169 + }, + { + "epoch": 0.0617, + "grad_norm": 0.7455936313181897, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6170 + }, + { + "epoch": 0.06171, + "grad_norm": 0.7443727028390393, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 6171 + }, + { + "epoch": 0.06172, + "grad_norm": 0.9360131411969993, + "learning_rate": 0.003, + "loss": 4.038, + "step": 6172 + }, + { + "epoch": 0.06173, + "grad_norm": 1.0491839760469293, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 6173 + }, + { + "epoch": 0.06174, + "grad_norm": 0.9907442372695112, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 6174 + }, + { + "epoch": 0.06175, + "grad_norm": 1.0447539234835719, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 6175 + }, + { + "epoch": 0.06176, + "grad_norm": 1.010073314902336, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 6176 + }, + { + "epoch": 0.06177, + "grad_norm": 1.1559422707750275, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 6177 + }, + { + "epoch": 0.06178, + "grad_norm": 0.9526053774670727, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 6178 + }, + { + "epoch": 0.06179, + "grad_norm": 1.022987693024061, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 6179 + }, + { + "epoch": 0.0618, + "grad_norm": 1.0210445225525222, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 6180 + }, + { + "epoch": 0.06181, + "grad_norm": 0.9583722005427316, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 6181 + }, + { + "epoch": 0.06182, + "grad_norm": 1.0777328688245935, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 6182 + }, + { + "epoch": 0.06183, + "grad_norm": 1.1218720835460165, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 6183 + }, + { + "epoch": 0.06184, + "grad_norm": 0.8901736945496374, + "learning_rate": 0.003, + "loss": 4.055, + "step": 6184 + }, + { + "epoch": 0.06185, + "grad_norm": 0.7821970283133759, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 6185 + }, + { + "epoch": 0.06186, + "grad_norm": 0.8340351304034461, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 6186 + }, + { + "epoch": 0.06187, + "grad_norm": 0.826969525348011, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 6187 + }, + { + "epoch": 0.06188, + "grad_norm": 0.8321316542169037, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 6188 + }, + { + "epoch": 0.06189, + "grad_norm": 1.0085697971600716, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 6189 + }, + { + "epoch": 0.0619, + "grad_norm": 1.2088449819616511, + "learning_rate": 0.003, + "loss": 4.098, + "step": 6190 + }, + { + "epoch": 0.06191, + "grad_norm": 1.052029182606652, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 6191 + }, + { + "epoch": 0.06192, + "grad_norm": 0.9984353604886477, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 6192 + }, + { + "epoch": 0.06193, + "grad_norm": 1.2795032941383804, + "learning_rate": 0.003, + "loss": 4.023, + "step": 6193 + }, + { + "epoch": 0.06194, + "grad_norm": 0.7998102276672221, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 6194 + }, + { + "epoch": 0.06195, + "grad_norm": 0.7435037294270525, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 6195 + }, + { + "epoch": 0.06196, + "grad_norm": 0.9240469454378734, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 6196 + }, + { + "epoch": 0.06197, + "grad_norm": 1.1696222467202784, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 6197 + }, + { + "epoch": 0.06198, + "grad_norm": 0.9340202086200383, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 6198 + }, + { + "epoch": 0.06199, + "grad_norm": 0.8407582922074252, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 6199 + }, + { + "epoch": 0.062, + "grad_norm": 0.8797469108168021, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 6200 + }, + { + "epoch": 0.06201, + "grad_norm": 0.9885503711260389, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 6201 + }, + { + "epoch": 0.06202, + "grad_norm": 1.033995787879555, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 6202 + }, + { + "epoch": 0.06203, + "grad_norm": 0.9934884285362043, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 6203 + }, + { + "epoch": 0.06204, + "grad_norm": 1.2510710857513951, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 6204 + }, + { + "epoch": 0.06205, + "grad_norm": 0.9427630454595206, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 6205 + }, + { + "epoch": 0.06206, + "grad_norm": 0.9038925258511812, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 6206 + }, + { + "epoch": 0.06207, + "grad_norm": 0.86210908819428, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 6207 + }, + { + "epoch": 0.06208, + "grad_norm": 0.8083534046824736, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 6208 + }, + { + "epoch": 0.06209, + "grad_norm": 0.906637525628039, + "learning_rate": 0.003, + "loss": 4.038, + "step": 6209 + }, + { + "epoch": 0.0621, + "grad_norm": 1.1049751485709722, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 6210 + }, + { + "epoch": 0.06211, + "grad_norm": 0.8465215117222046, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 6211 + }, + { + "epoch": 0.06212, + "grad_norm": 0.926150960560891, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 6212 + }, + { + "epoch": 0.06213, + "grad_norm": 1.1848974987642524, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 6213 + }, + { + "epoch": 0.06214, + "grad_norm": 0.9608260526041966, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 6214 + }, + { + "epoch": 0.06215, + "grad_norm": 1.0487901084020896, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 6215 + }, + { + "epoch": 0.06216, + "grad_norm": 1.0157866198312488, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 6216 + }, + { + "epoch": 0.06217, + "grad_norm": 1.1412151022720478, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 6217 + }, + { + "epoch": 0.06218, + "grad_norm": 1.006247649190211, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 6218 + }, + { + "epoch": 0.06219, + "grad_norm": 1.2294549857751111, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 6219 + }, + { + "epoch": 0.0622, + "grad_norm": 0.995208064028826, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 6220 + }, + { + "epoch": 0.06221, + "grad_norm": 0.9162212484056992, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 6221 + }, + { + "epoch": 0.06222, + "grad_norm": 0.8454292925737381, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 6222 + }, + { + "epoch": 0.06223, + "grad_norm": 0.8608341935906129, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 6223 + }, + { + "epoch": 0.06224, + "grad_norm": 1.0517375571967738, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6224 + }, + { + "epoch": 0.06225, + "grad_norm": 1.167553303659172, + "learning_rate": 0.003, + "loss": 4.072, + "step": 6225 + }, + { + "epoch": 0.06226, + "grad_norm": 0.9413390695324375, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 6226 + }, + { + "epoch": 0.06227, + "grad_norm": 1.1275346673821272, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 6227 + }, + { + "epoch": 0.06228, + "grad_norm": 1.071326538063058, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 6228 + }, + { + "epoch": 0.06229, + "grad_norm": 1.0923671993434136, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 6229 + }, + { + "epoch": 0.0623, + "grad_norm": 1.1044316515319128, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 6230 + }, + { + "epoch": 0.06231, + "grad_norm": 1.1192386233956613, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 6231 + }, + { + "epoch": 0.06232, + "grad_norm": 1.0749004721163167, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 6232 + }, + { + "epoch": 0.06233, + "grad_norm": 1.136441483294293, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 6233 + }, + { + "epoch": 0.06234, + "grad_norm": 1.096293829074208, + "learning_rate": 0.003, + "loss": 4.057, + "step": 6234 + }, + { + "epoch": 0.06235, + "grad_norm": 0.9421513147723747, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 6235 + }, + { + "epoch": 0.06236, + "grad_norm": 0.9920661776334326, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 6236 + }, + { + "epoch": 0.06237, + "grad_norm": 1.009784196859223, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 6237 + }, + { + "epoch": 0.06238, + "grad_norm": 0.9870780579077241, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 6238 + }, + { + "epoch": 0.06239, + "grad_norm": 1.2624770661972455, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 6239 + }, + { + "epoch": 0.0624, + "grad_norm": 0.8345122687975308, + "learning_rate": 0.003, + "loss": 4.053, + "step": 6240 + }, + { + "epoch": 0.06241, + "grad_norm": 0.8767674013310682, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 6241 + }, + { + "epoch": 0.06242, + "grad_norm": 0.8694164086668165, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 6242 + }, + { + "epoch": 0.06243, + "grad_norm": 0.8968698662271256, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 6243 + }, + { + "epoch": 0.06244, + "grad_norm": 1.0969155124189833, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 6244 + }, + { + "epoch": 0.06245, + "grad_norm": 0.9267212634506801, + "learning_rate": 0.003, + "loss": 4.059, + "step": 6245 + }, + { + "epoch": 0.06246, + "grad_norm": 1.0070455660568638, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 6246 + }, + { + "epoch": 0.06247, + "grad_norm": 1.1508646699867722, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 6247 + }, + { + "epoch": 0.06248, + "grad_norm": 0.9383548414378287, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 6248 + }, + { + "epoch": 0.06249, + "grad_norm": 0.8885352213986607, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 6249 + }, + { + "epoch": 0.0625, + "grad_norm": 0.8342972923726877, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 6250 + }, + { + "epoch": 0.06251, + "grad_norm": 0.9237428978300573, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 6251 + }, + { + "epoch": 0.06252, + "grad_norm": 1.084993649871814, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 6252 + }, + { + "epoch": 0.06253, + "grad_norm": 1.0323949223042157, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 6253 + }, + { + "epoch": 0.06254, + "grad_norm": 0.971393898605744, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6254 + }, + { + "epoch": 0.06255, + "grad_norm": 0.9706954283600536, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 6255 + }, + { + "epoch": 0.06256, + "grad_norm": 0.9415015290908161, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6256 + }, + { + "epoch": 0.06257, + "grad_norm": 0.8621578985318906, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 6257 + }, + { + "epoch": 0.06258, + "grad_norm": 0.9548440225051628, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 6258 + }, + { + "epoch": 0.06259, + "grad_norm": 0.9263497464951767, + "learning_rate": 0.003, + "loss": 4.088, + "step": 6259 + }, + { + "epoch": 0.0626, + "grad_norm": 0.8908902137670806, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 6260 + }, + { + "epoch": 0.06261, + "grad_norm": 0.8327652914912796, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 6261 + }, + { + "epoch": 0.06262, + "grad_norm": 0.9538403925841564, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 6262 + }, + { + "epoch": 0.06263, + "grad_norm": 1.1599473881323492, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 6263 + }, + { + "epoch": 0.06264, + "grad_norm": 1.0110607933412514, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 6264 + }, + { + "epoch": 0.06265, + "grad_norm": 1.0076482190602307, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 6265 + }, + { + "epoch": 0.06266, + "grad_norm": 0.8737903635942322, + "learning_rate": 0.003, + "loss": 4.056, + "step": 6266 + }, + { + "epoch": 0.06267, + "grad_norm": 0.8712992228434959, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 6267 + }, + { + "epoch": 0.06268, + "grad_norm": 1.017260911789287, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 6268 + }, + { + "epoch": 0.06269, + "grad_norm": 0.997967424239097, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 6269 + }, + { + "epoch": 0.0627, + "grad_norm": 1.0751801208147977, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 6270 + }, + { + "epoch": 0.06271, + "grad_norm": 1.0371657284207103, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 6271 + }, + { + "epoch": 0.06272, + "grad_norm": 1.2803244076660563, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 6272 + }, + { + "epoch": 0.06273, + "grad_norm": 0.9500833272967772, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 6273 + }, + { + "epoch": 0.06274, + "grad_norm": 1.0958729001288292, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 6274 + }, + { + "epoch": 0.06275, + "grad_norm": 1.141758768594646, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 6275 + }, + { + "epoch": 0.06276, + "grad_norm": 0.9628716788683748, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 6276 + }, + { + "epoch": 0.06277, + "grad_norm": 1.0976060075916712, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 6277 + }, + { + "epoch": 0.06278, + "grad_norm": 1.016066298115058, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 6278 + }, + { + "epoch": 0.06279, + "grad_norm": 0.8769401740306906, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 6279 + }, + { + "epoch": 0.0628, + "grad_norm": 0.9170763425191215, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 6280 + }, + { + "epoch": 0.06281, + "grad_norm": 1.111920165775686, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 6281 + }, + { + "epoch": 0.06282, + "grad_norm": 1.0224143144159263, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 6282 + }, + { + "epoch": 0.06283, + "grad_norm": 1.1015506226196767, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 6283 + }, + { + "epoch": 0.06284, + "grad_norm": 0.9844477787774102, + "learning_rate": 0.003, + "loss": 4.066, + "step": 6284 + }, + { + "epoch": 0.06285, + "grad_norm": 1.0020352933530818, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 6285 + }, + { + "epoch": 0.06286, + "grad_norm": 0.8631477379490902, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 6286 + }, + { + "epoch": 0.06287, + "grad_norm": 0.8888287717900007, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 6287 + }, + { + "epoch": 0.06288, + "grad_norm": 0.8183385522469628, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 6288 + }, + { + "epoch": 0.06289, + "grad_norm": 0.7548987473458644, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 6289 + }, + { + "epoch": 0.0629, + "grad_norm": 0.839173523396433, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 6290 + }, + { + "epoch": 0.06291, + "grad_norm": 0.8617471545841462, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 6291 + }, + { + "epoch": 0.06292, + "grad_norm": 0.8638474112750509, + "learning_rate": 0.003, + "loss": 4.05, + "step": 6292 + }, + { + "epoch": 0.06293, + "grad_norm": 0.9541204019425185, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 6293 + }, + { + "epoch": 0.06294, + "grad_norm": 1.1723643539699384, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 6294 + }, + { + "epoch": 0.06295, + "grad_norm": 1.103442251330168, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 6295 + }, + { + "epoch": 0.06296, + "grad_norm": 1.0703680632965407, + "learning_rate": 0.003, + "loss": 4.086, + "step": 6296 + }, + { + "epoch": 0.06297, + "grad_norm": 1.0274650320053247, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 6297 + }, + { + "epoch": 0.06298, + "grad_norm": 0.89154419061986, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6298 + }, + { + "epoch": 0.06299, + "grad_norm": 0.9135944828728849, + "learning_rate": 0.003, + "loss": 4.054, + "step": 6299 + }, + { + "epoch": 0.063, + "grad_norm": 0.8267113445895482, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 6300 + }, + { + "epoch": 0.06301, + "grad_norm": 1.0037936149372095, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 6301 + }, + { + "epoch": 0.06302, + "grad_norm": 1.2860071610961534, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 6302 + }, + { + "epoch": 0.06303, + "grad_norm": 1.0162453119057104, + "learning_rate": 0.003, + "loss": 4.073, + "step": 6303 + }, + { + "epoch": 0.06304, + "grad_norm": 1.0450143375967518, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 6304 + }, + { + "epoch": 0.06305, + "grad_norm": 1.1401136900936193, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 6305 + }, + { + "epoch": 0.06306, + "grad_norm": 1.0466179358101595, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 6306 + }, + { + "epoch": 0.06307, + "grad_norm": 1.1366527233292367, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6307 + }, + { + "epoch": 0.06308, + "grad_norm": 0.8947523569057309, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 6308 + }, + { + "epoch": 0.06309, + "grad_norm": 0.9787036236305489, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 6309 + }, + { + "epoch": 0.0631, + "grad_norm": 1.1060093552214123, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 6310 + }, + { + "epoch": 0.06311, + "grad_norm": 0.934221683083381, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 6311 + }, + { + "epoch": 0.06312, + "grad_norm": 1.0377549576803702, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 6312 + }, + { + "epoch": 0.06313, + "grad_norm": 0.9581325400036763, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 6313 + }, + { + "epoch": 0.06314, + "grad_norm": 1.1035319349717054, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 6314 + }, + { + "epoch": 0.06315, + "grad_norm": 0.9924113339745636, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 6315 + }, + { + "epoch": 0.06316, + "grad_norm": 1.0717645296978155, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 6316 + }, + { + "epoch": 0.06317, + "grad_norm": 1.0557719176935128, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 6317 + }, + { + "epoch": 0.06318, + "grad_norm": 1.151290740982202, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 6318 + }, + { + "epoch": 0.06319, + "grad_norm": 1.1038168869223615, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 6319 + }, + { + "epoch": 0.0632, + "grad_norm": 0.9945691257477356, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 6320 + }, + { + "epoch": 0.06321, + "grad_norm": 1.0211879444413088, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 6321 + }, + { + "epoch": 0.06322, + "grad_norm": 1.074648930120513, + "learning_rate": 0.003, + "loss": 4.065, + "step": 6322 + }, + { + "epoch": 0.06323, + "grad_norm": 1.0531931604114575, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 6323 + }, + { + "epoch": 0.06324, + "grad_norm": 1.0525740567574657, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 6324 + }, + { + "epoch": 0.06325, + "grad_norm": 1.0156601469799102, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 6325 + }, + { + "epoch": 0.06326, + "grad_norm": 1.0885712514526227, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 6326 + }, + { + "epoch": 0.06327, + "grad_norm": 0.9064082305136328, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 6327 + }, + { + "epoch": 0.06328, + "grad_norm": 0.9294707681271361, + "learning_rate": 0.003, + "loss": 4.084, + "step": 6328 + }, + { + "epoch": 0.06329, + "grad_norm": 1.3371931683837754, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 6329 + }, + { + "epoch": 0.0633, + "grad_norm": 0.8781010739013081, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 6330 + }, + { + "epoch": 0.06331, + "grad_norm": 0.8544193926722347, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 6331 + }, + { + "epoch": 0.06332, + "grad_norm": 0.8011724000742859, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 6332 + }, + { + "epoch": 0.06333, + "grad_norm": 0.7119981509484664, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 6333 + }, + { + "epoch": 0.06334, + "grad_norm": 0.6668806340124519, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 6334 + }, + { + "epoch": 0.06335, + "grad_norm": 0.8212002598201582, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 6335 + }, + { + "epoch": 0.06336, + "grad_norm": 0.8088000326714004, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 6336 + }, + { + "epoch": 0.06337, + "grad_norm": 0.8536839468376971, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 6337 + }, + { + "epoch": 0.06338, + "grad_norm": 1.1037572047297124, + "learning_rate": 0.003, + "loss": 4.065, + "step": 6338 + }, + { + "epoch": 0.06339, + "grad_norm": 1.3377815021844122, + "learning_rate": 0.003, + "loss": 4.048, + "step": 6339 + }, + { + "epoch": 0.0634, + "grad_norm": 0.6048765308025956, + "learning_rate": 0.003, + "loss": 4.032, + "step": 6340 + }, + { + "epoch": 0.06341, + "grad_norm": 0.7799193061695396, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 6341 + }, + { + "epoch": 0.06342, + "grad_norm": 1.1151041355238953, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 6342 + }, + { + "epoch": 0.06343, + "grad_norm": 1.1674486948458083, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 6343 + }, + { + "epoch": 0.06344, + "grad_norm": 1.031837089911801, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 6344 + }, + { + "epoch": 0.06345, + "grad_norm": 0.8615252116136146, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 6345 + }, + { + "epoch": 0.06346, + "grad_norm": 0.8103372667424406, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 6346 + }, + { + "epoch": 0.06347, + "grad_norm": 0.8933549042384386, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 6347 + }, + { + "epoch": 0.06348, + "grad_norm": 0.9637829939992113, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 6348 + }, + { + "epoch": 0.06349, + "grad_norm": 0.9135048719984086, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 6349 + }, + { + "epoch": 0.0635, + "grad_norm": 1.1242591431500657, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 6350 + }, + { + "epoch": 0.06351, + "grad_norm": 0.9980317399940769, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 6351 + }, + { + "epoch": 0.06352, + "grad_norm": 1.0451922994538014, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 6352 + }, + { + "epoch": 0.06353, + "grad_norm": 0.9845927560222498, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 6353 + }, + { + "epoch": 0.06354, + "grad_norm": 1.096403795880879, + "learning_rate": 0.003, + "loss": 4.059, + "step": 6354 + }, + { + "epoch": 0.06355, + "grad_norm": 0.9983219209770657, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 6355 + }, + { + "epoch": 0.06356, + "grad_norm": 1.187703733940572, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 6356 + }, + { + "epoch": 0.06357, + "grad_norm": 0.9615101296213491, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 6357 + }, + { + "epoch": 0.06358, + "grad_norm": 1.1451194898172852, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 6358 + }, + { + "epoch": 0.06359, + "grad_norm": 1.0483116074347743, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 6359 + }, + { + "epoch": 0.0636, + "grad_norm": 0.9340547243285371, + "learning_rate": 0.003, + "loss": 4.075, + "step": 6360 + }, + { + "epoch": 0.06361, + "grad_norm": 0.8836803848175515, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 6361 + }, + { + "epoch": 0.06362, + "grad_norm": 0.8297069107111867, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 6362 + }, + { + "epoch": 0.06363, + "grad_norm": 0.9691564696742783, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 6363 + }, + { + "epoch": 0.06364, + "grad_norm": 1.0194286737320961, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6364 + }, + { + "epoch": 0.06365, + "grad_norm": 1.0411408567959355, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 6365 + }, + { + "epoch": 0.06366, + "grad_norm": 0.9916919363829898, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 6366 + }, + { + "epoch": 0.06367, + "grad_norm": 0.9560672684375412, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 6367 + }, + { + "epoch": 0.06368, + "grad_norm": 1.1782070353358838, + "learning_rate": 0.003, + "loss": 4.092, + "step": 6368 + }, + { + "epoch": 0.06369, + "grad_norm": 1.133684868527638, + "learning_rate": 0.003, + "loss": 4.099, + "step": 6369 + }, + { + "epoch": 0.0637, + "grad_norm": 0.8986796581542652, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 6370 + }, + { + "epoch": 0.06371, + "grad_norm": 0.798117831160487, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 6371 + }, + { + "epoch": 0.06372, + "grad_norm": 0.841679881049868, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 6372 + }, + { + "epoch": 0.06373, + "grad_norm": 0.912755996468821, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 6373 + }, + { + "epoch": 0.06374, + "grad_norm": 0.8902631081176816, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 6374 + }, + { + "epoch": 0.06375, + "grad_norm": 0.9176190979451236, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 6375 + }, + { + "epoch": 0.06376, + "grad_norm": 1.0740748411500862, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 6376 + }, + { + "epoch": 0.06377, + "grad_norm": 1.1585312174412248, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 6377 + }, + { + "epoch": 0.06378, + "grad_norm": 1.0560699306688008, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 6378 + }, + { + "epoch": 0.06379, + "grad_norm": 1.2860487561573273, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 6379 + }, + { + "epoch": 0.0638, + "grad_norm": 0.8690291787190746, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 6380 + }, + { + "epoch": 0.06381, + "grad_norm": 0.9436829595338491, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 6381 + }, + { + "epoch": 0.06382, + "grad_norm": 1.0685626630336664, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 6382 + }, + { + "epoch": 0.06383, + "grad_norm": 0.8979552535870967, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 6383 + }, + { + "epoch": 0.06384, + "grad_norm": 0.9809724782902665, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 6384 + }, + { + "epoch": 0.06385, + "grad_norm": 0.9328427920083763, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 6385 + }, + { + "epoch": 0.06386, + "grad_norm": 0.8685113357756273, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 6386 + }, + { + "epoch": 0.06387, + "grad_norm": 0.9155556766963706, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 6387 + }, + { + "epoch": 0.06388, + "grad_norm": 0.9520115095678209, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 6388 + }, + { + "epoch": 0.06389, + "grad_norm": 0.9650777255338291, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 6389 + }, + { + "epoch": 0.0639, + "grad_norm": 0.9711913071461878, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 6390 + }, + { + "epoch": 0.06391, + "grad_norm": 1.0582916428545706, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 6391 + }, + { + "epoch": 0.06392, + "grad_norm": 1.1228513540493317, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 6392 + }, + { + "epoch": 0.06393, + "grad_norm": 0.9824467176596088, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 6393 + }, + { + "epoch": 0.06394, + "grad_norm": 1.0748961169419036, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 6394 + }, + { + "epoch": 0.06395, + "grad_norm": 1.2122348207004991, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 6395 + }, + { + "epoch": 0.06396, + "grad_norm": 0.9525125913234459, + "learning_rate": 0.003, + "loss": 4.095, + "step": 6396 + }, + { + "epoch": 0.06397, + "grad_norm": 1.1157249696277263, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 6397 + }, + { + "epoch": 0.06398, + "grad_norm": 1.0962231868724095, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 6398 + }, + { + "epoch": 0.06399, + "grad_norm": 0.7475394505526716, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 6399 + }, + { + "epoch": 0.064, + "grad_norm": 0.7790863228862557, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 6400 + }, + { + "epoch": 0.06401, + "grad_norm": 0.9202551013299898, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 6401 + }, + { + "epoch": 0.06402, + "grad_norm": 0.9324453718148813, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 6402 + }, + { + "epoch": 0.06403, + "grad_norm": 1.0178636443747968, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 6403 + }, + { + "epoch": 0.06404, + "grad_norm": 1.0889588831558945, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 6404 + }, + { + "epoch": 0.06405, + "grad_norm": 1.0511307110083934, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 6405 + }, + { + "epoch": 0.06406, + "grad_norm": 0.8386442367077682, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 6406 + }, + { + "epoch": 0.06407, + "grad_norm": 0.9385442318393312, + "learning_rate": 0.003, + "loss": 4.045, + "step": 6407 + }, + { + "epoch": 0.06408, + "grad_norm": 1.1880536233711294, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 6408 + }, + { + "epoch": 0.06409, + "grad_norm": 1.0715566126733156, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 6409 + }, + { + "epoch": 0.0641, + "grad_norm": 1.0226807648909895, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 6410 + }, + { + "epoch": 0.06411, + "grad_norm": 1.0144667419036346, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 6411 + }, + { + "epoch": 0.06412, + "grad_norm": 1.0953706363475204, + "learning_rate": 0.003, + "loss": 4.05, + "step": 6412 + }, + { + "epoch": 0.06413, + "grad_norm": 1.0397462803376565, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 6413 + }, + { + "epoch": 0.06414, + "grad_norm": 0.9792075931298253, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 6414 + }, + { + "epoch": 0.06415, + "grad_norm": 1.0282547055284612, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 6415 + }, + { + "epoch": 0.06416, + "grad_norm": 1.159533810090572, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 6416 + }, + { + "epoch": 0.06417, + "grad_norm": 1.1321519339699178, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6417 + }, + { + "epoch": 0.06418, + "grad_norm": 0.9713020938212722, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 6418 + }, + { + "epoch": 0.06419, + "grad_norm": 1.2376733723226676, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6419 + }, + { + "epoch": 0.0642, + "grad_norm": 0.9221140531392306, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 6420 + }, + { + "epoch": 0.06421, + "grad_norm": 0.9969702997793933, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 6421 + }, + { + "epoch": 0.06422, + "grad_norm": 1.0567219061372177, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 6422 + }, + { + "epoch": 0.06423, + "grad_norm": 0.991239045933805, + "learning_rate": 0.003, + "loss": 4.077, + "step": 6423 + }, + { + "epoch": 0.06424, + "grad_norm": 1.0308086601832684, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 6424 + }, + { + "epoch": 0.06425, + "grad_norm": 1.0347557359010413, + "learning_rate": 0.003, + "loss": 4.062, + "step": 6425 + }, + { + "epoch": 0.06426, + "grad_norm": 1.180251291539417, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 6426 + }, + { + "epoch": 0.06427, + "grad_norm": 1.0107602148306123, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 6427 + }, + { + "epoch": 0.06428, + "grad_norm": 1.0541298404940496, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 6428 + }, + { + "epoch": 0.06429, + "grad_norm": 1.0253882188929047, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6429 + }, + { + "epoch": 0.0643, + "grad_norm": 1.143653137235699, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 6430 + }, + { + "epoch": 0.06431, + "grad_norm": 0.9421049550422983, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 6431 + }, + { + "epoch": 0.06432, + "grad_norm": 1.1073171928356662, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 6432 + }, + { + "epoch": 0.06433, + "grad_norm": 1.0407008893715575, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 6433 + }, + { + "epoch": 0.06434, + "grad_norm": 1.1395794148637106, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 6434 + }, + { + "epoch": 0.06435, + "grad_norm": 1.026318482136573, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 6435 + }, + { + "epoch": 0.06436, + "grad_norm": 1.1702677902712457, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 6436 + }, + { + "epoch": 0.06437, + "grad_norm": 1.025273476921687, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 6437 + }, + { + "epoch": 0.06438, + "grad_norm": 1.0119572577344156, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 6438 + }, + { + "epoch": 0.06439, + "grad_norm": 0.8412430606835657, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6439 + }, + { + "epoch": 0.0644, + "grad_norm": 0.9507548790037242, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 6440 + }, + { + "epoch": 0.06441, + "grad_norm": 0.9244681652798047, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 6441 + }, + { + "epoch": 0.06442, + "grad_norm": 0.8807778875679908, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 6442 + }, + { + "epoch": 0.06443, + "grad_norm": 0.827270815956522, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 6443 + }, + { + "epoch": 0.06444, + "grad_norm": 0.803325818969944, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 6444 + }, + { + "epoch": 0.06445, + "grad_norm": 0.9385724441228014, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 6445 + }, + { + "epoch": 0.06446, + "grad_norm": 1.2005144304193571, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 6446 + }, + { + "epoch": 0.06447, + "grad_norm": 0.9501919917514117, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 6447 + }, + { + "epoch": 0.06448, + "grad_norm": 1.0236274372742054, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 6448 + }, + { + "epoch": 0.06449, + "grad_norm": 1.0925032591156019, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 6449 + }, + { + "epoch": 0.0645, + "grad_norm": 0.9065417419150241, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 6450 + }, + { + "epoch": 0.06451, + "grad_norm": 0.8998787170517907, + "learning_rate": 0.003, + "loss": 4.087, + "step": 6451 + }, + { + "epoch": 0.06452, + "grad_norm": 0.9437442931525397, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 6452 + }, + { + "epoch": 0.06453, + "grad_norm": 1.0928923944168785, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 6453 + }, + { + "epoch": 0.06454, + "grad_norm": 1.1495081243301128, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 6454 + }, + { + "epoch": 0.06455, + "grad_norm": 0.9528175723610092, + "learning_rate": 0.003, + "loss": 4.057, + "step": 6455 + }, + { + "epoch": 0.06456, + "grad_norm": 1.0833738182788575, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 6456 + }, + { + "epoch": 0.06457, + "grad_norm": 1.1107357171312071, + "learning_rate": 0.003, + "loss": 4.055, + "step": 6457 + }, + { + "epoch": 0.06458, + "grad_norm": 1.1510295337322964, + "learning_rate": 0.003, + "loss": 4.068, + "step": 6458 + }, + { + "epoch": 0.06459, + "grad_norm": 1.0262159806305138, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 6459 + }, + { + "epoch": 0.0646, + "grad_norm": 1.014576196236502, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 6460 + }, + { + "epoch": 0.06461, + "grad_norm": 1.064471879938064, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 6461 + }, + { + "epoch": 0.06462, + "grad_norm": 0.9222498176393344, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 6462 + }, + { + "epoch": 0.06463, + "grad_norm": 0.8981156217551731, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6463 + }, + { + "epoch": 0.06464, + "grad_norm": 1.1107391697536575, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 6464 + }, + { + "epoch": 0.06465, + "grad_norm": 1.0063336191833832, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 6465 + }, + { + "epoch": 0.06466, + "grad_norm": 1.1961559729990558, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 6466 + }, + { + "epoch": 0.06467, + "grad_norm": 0.8793978826576908, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 6467 + }, + { + "epoch": 0.06468, + "grad_norm": 1.0427733287660477, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6468 + }, + { + "epoch": 0.06469, + "grad_norm": 1.1251364660455194, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 6469 + }, + { + "epoch": 0.0647, + "grad_norm": 1.0043010097247658, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 6470 + }, + { + "epoch": 0.06471, + "grad_norm": 1.1207301616281806, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 6471 + }, + { + "epoch": 0.06472, + "grad_norm": 0.9991580557027551, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 6472 + }, + { + "epoch": 0.06473, + "grad_norm": 1.1171158410309772, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 6473 + }, + { + "epoch": 0.06474, + "grad_norm": 0.8529112213695773, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 6474 + }, + { + "epoch": 0.06475, + "grad_norm": 0.8417759233289197, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 6475 + }, + { + "epoch": 0.06476, + "grad_norm": 0.9302953810525731, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 6476 + }, + { + "epoch": 0.06477, + "grad_norm": 1.0176820311771924, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 6477 + }, + { + "epoch": 0.06478, + "grad_norm": 1.1064098044871353, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 6478 + }, + { + "epoch": 0.06479, + "grad_norm": 1.0450987906387426, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 6479 + }, + { + "epoch": 0.0648, + "grad_norm": 1.0076941555756993, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 6480 + }, + { + "epoch": 0.06481, + "grad_norm": 0.8795541683622259, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 6481 + }, + { + "epoch": 0.06482, + "grad_norm": 0.9850429100701444, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6482 + }, + { + "epoch": 0.06483, + "grad_norm": 1.104728714549408, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 6483 + }, + { + "epoch": 0.06484, + "grad_norm": 1.0625023539317728, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 6484 + }, + { + "epoch": 0.06485, + "grad_norm": 0.9421033844694122, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 6485 + }, + { + "epoch": 0.06486, + "grad_norm": 0.8988596473281786, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 6486 + }, + { + "epoch": 0.06487, + "grad_norm": 0.963516647410238, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 6487 + }, + { + "epoch": 0.06488, + "grad_norm": 1.0333879473021825, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 6488 + }, + { + "epoch": 0.06489, + "grad_norm": 1.0586786599679208, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 6489 + }, + { + "epoch": 0.0649, + "grad_norm": 0.9937287592814337, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 6490 + }, + { + "epoch": 0.06491, + "grad_norm": 1.1726633804612512, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 6491 + }, + { + "epoch": 0.06492, + "grad_norm": 1.0927920818435795, + "learning_rate": 0.003, + "loss": 4.075, + "step": 6492 + }, + { + "epoch": 0.06493, + "grad_norm": 1.0227562445548077, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 6493 + }, + { + "epoch": 0.06494, + "grad_norm": 1.0831459762005358, + "learning_rate": 0.003, + "loss": 4.072, + "step": 6494 + }, + { + "epoch": 0.06495, + "grad_norm": 1.0262098572577179, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 6495 + }, + { + "epoch": 0.06496, + "grad_norm": 1.2236916592367486, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 6496 + }, + { + "epoch": 0.06497, + "grad_norm": 1.0285867021198758, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 6497 + }, + { + "epoch": 0.06498, + "grad_norm": 1.1733459098787142, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 6498 + }, + { + "epoch": 0.06499, + "grad_norm": 1.0531898340962396, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 6499 + }, + { + "epoch": 0.065, + "grad_norm": 1.060199193118051, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6500 + }, + { + "epoch": 0.06501, + "grad_norm": 1.1527686386988991, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 6501 + }, + { + "epoch": 0.06502, + "grad_norm": 1.029295109562913, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 6502 + }, + { + "epoch": 0.06503, + "grad_norm": 0.9836711412915636, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 6503 + }, + { + "epoch": 0.06504, + "grad_norm": 0.9285828574317885, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 6504 + }, + { + "epoch": 0.06505, + "grad_norm": 0.975483354586866, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 6505 + }, + { + "epoch": 0.06506, + "grad_norm": 1.2115222037214375, + "learning_rate": 0.003, + "loss": 4.045, + "step": 6506 + }, + { + "epoch": 0.06507, + "grad_norm": 0.9656432384586594, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 6507 + }, + { + "epoch": 0.06508, + "grad_norm": 1.0366546849866198, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 6508 + }, + { + "epoch": 0.06509, + "grad_norm": 1.009479145061525, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 6509 + }, + { + "epoch": 0.0651, + "grad_norm": 1.0366020609012134, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 6510 + }, + { + "epoch": 0.06511, + "grad_norm": 0.9709239930407754, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 6511 + }, + { + "epoch": 0.06512, + "grad_norm": 0.8644707502418649, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 6512 + }, + { + "epoch": 0.06513, + "grad_norm": 0.8946018112148616, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 6513 + }, + { + "epoch": 0.06514, + "grad_norm": 1.032708726685494, + "learning_rate": 0.003, + "loss": 4.05, + "step": 6514 + }, + { + "epoch": 0.06515, + "grad_norm": 1.0409138603347203, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 6515 + }, + { + "epoch": 0.06516, + "grad_norm": 0.9892609384071499, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 6516 + }, + { + "epoch": 0.06517, + "grad_norm": 0.9148700446008358, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 6517 + }, + { + "epoch": 0.06518, + "grad_norm": 0.9979358270843701, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 6518 + }, + { + "epoch": 0.06519, + "grad_norm": 1.185271592002419, + "learning_rate": 0.003, + "loss": 4.054, + "step": 6519 + }, + { + "epoch": 0.0652, + "grad_norm": 0.9704305412861196, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 6520 + }, + { + "epoch": 0.06521, + "grad_norm": 1.1014185224929063, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 6521 + }, + { + "epoch": 0.06522, + "grad_norm": 1.0703081501278155, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 6522 + }, + { + "epoch": 0.06523, + "grad_norm": 0.8771352485540366, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 6523 + }, + { + "epoch": 0.06524, + "grad_norm": 0.9130638198055455, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 6524 + }, + { + "epoch": 0.06525, + "grad_norm": 1.1326914793850837, + "learning_rate": 0.003, + "loss": 4.078, + "step": 6525 + }, + { + "epoch": 0.06526, + "grad_norm": 1.021254454818802, + "learning_rate": 0.003, + "loss": 4.061, + "step": 6526 + }, + { + "epoch": 0.06527, + "grad_norm": 1.073497230759006, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 6527 + }, + { + "epoch": 0.06528, + "grad_norm": 0.9631221418357153, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 6528 + }, + { + "epoch": 0.06529, + "grad_norm": 0.8809083485174808, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 6529 + }, + { + "epoch": 0.0653, + "grad_norm": 0.9688255439544828, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 6530 + }, + { + "epoch": 0.06531, + "grad_norm": 0.9824939601057029, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 6531 + }, + { + "epoch": 0.06532, + "grad_norm": 0.9436605093416098, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 6532 + }, + { + "epoch": 0.06533, + "grad_norm": 0.8767433235215197, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 6533 + }, + { + "epoch": 0.06534, + "grad_norm": 1.0523478236740318, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 6534 + }, + { + "epoch": 0.06535, + "grad_norm": 1.0000197595533407, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 6535 + }, + { + "epoch": 0.06536, + "grad_norm": 1.2508564779795237, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 6536 + }, + { + "epoch": 0.06537, + "grad_norm": 0.789891181606252, + "learning_rate": 0.003, + "loss": 4.061, + "step": 6537 + }, + { + "epoch": 0.06538, + "grad_norm": 0.7387596226051653, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 6538 + }, + { + "epoch": 0.06539, + "grad_norm": 0.6182415296014235, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 6539 + }, + { + "epoch": 0.0654, + "grad_norm": 0.6896044138716849, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 6540 + }, + { + "epoch": 0.06541, + "grad_norm": 0.9475899668837449, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 6541 + }, + { + "epoch": 0.06542, + "grad_norm": 1.214498343082983, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 6542 + }, + { + "epoch": 0.06543, + "grad_norm": 0.8105078624490595, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 6543 + }, + { + "epoch": 0.06544, + "grad_norm": 0.7169187786636381, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6544 + }, + { + "epoch": 0.06545, + "grad_norm": 0.7645714804284779, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 6545 + }, + { + "epoch": 0.06546, + "grad_norm": 0.7984074302017922, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 6546 + }, + { + "epoch": 0.06547, + "grad_norm": 0.898561711007871, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 6547 + }, + { + "epoch": 0.06548, + "grad_norm": 1.0356346515332355, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 6548 + }, + { + "epoch": 0.06549, + "grad_norm": 1.2065842171050336, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 6549 + }, + { + "epoch": 0.0655, + "grad_norm": 0.9647919793019113, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 6550 + }, + { + "epoch": 0.06551, + "grad_norm": 0.9830823232743289, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 6551 + }, + { + "epoch": 0.06552, + "grad_norm": 1.145414200189479, + "learning_rate": 0.003, + "loss": 4.08, + "step": 6552 + }, + { + "epoch": 0.06553, + "grad_norm": 1.1282546842302874, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 6553 + }, + { + "epoch": 0.06554, + "grad_norm": 1.0724517832235836, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 6554 + }, + { + "epoch": 0.06555, + "grad_norm": 0.9247118456778185, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 6555 + }, + { + "epoch": 0.06556, + "grad_norm": 0.8809168909392743, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 6556 + }, + { + "epoch": 0.06557, + "grad_norm": 1.0027045686558371, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 6557 + }, + { + "epoch": 0.06558, + "grad_norm": 1.1521312374668213, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 6558 + }, + { + "epoch": 0.06559, + "grad_norm": 0.9170120064963928, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 6559 + }, + { + "epoch": 0.0656, + "grad_norm": 0.8925486266426595, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6560 + }, + { + "epoch": 0.06561, + "grad_norm": 0.9709022873633115, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 6561 + }, + { + "epoch": 0.06562, + "grad_norm": 1.0699148932589904, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 6562 + }, + { + "epoch": 0.06563, + "grad_norm": 1.018722166079644, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 6563 + }, + { + "epoch": 0.06564, + "grad_norm": 1.0104563015086112, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 6564 + }, + { + "epoch": 0.06565, + "grad_norm": 0.9932591703232511, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 6565 + }, + { + "epoch": 0.06566, + "grad_norm": 0.9531047599334086, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 6566 + }, + { + "epoch": 0.06567, + "grad_norm": 1.0533507364705403, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 6567 + }, + { + "epoch": 0.06568, + "grad_norm": 1.2316178931504538, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 6568 + }, + { + "epoch": 0.06569, + "grad_norm": 0.9719523712712863, + "learning_rate": 0.003, + "loss": 4.031, + "step": 6569 + }, + { + "epoch": 0.0657, + "grad_norm": 0.9929746213073647, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 6570 + }, + { + "epoch": 0.06571, + "grad_norm": 1.2226520601321686, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 6571 + }, + { + "epoch": 0.06572, + "grad_norm": 1.0093561514710232, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 6572 + }, + { + "epoch": 0.06573, + "grad_norm": 1.0716061347239403, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 6573 + }, + { + "epoch": 0.06574, + "grad_norm": 0.9809951043877417, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 6574 + }, + { + "epoch": 0.06575, + "grad_norm": 0.9974953620162613, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 6575 + }, + { + "epoch": 0.06576, + "grad_norm": 0.9179282772036231, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 6576 + }, + { + "epoch": 0.06577, + "grad_norm": 1.0424561456811468, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 6577 + }, + { + "epoch": 0.06578, + "grad_norm": 1.0920709972772817, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 6578 + }, + { + "epoch": 0.06579, + "grad_norm": 1.2580611117333567, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 6579 + }, + { + "epoch": 0.0658, + "grad_norm": 0.799965431122157, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 6580 + }, + { + "epoch": 0.06581, + "grad_norm": 0.781504567533229, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 6581 + }, + { + "epoch": 0.06582, + "grad_norm": 0.8493771135199062, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 6582 + }, + { + "epoch": 0.06583, + "grad_norm": 1.08382714627964, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 6583 + }, + { + "epoch": 0.06584, + "grad_norm": 0.9060976759116695, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 6584 + }, + { + "epoch": 0.06585, + "grad_norm": 1.0250567461266544, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6585 + }, + { + "epoch": 0.06586, + "grad_norm": 1.2416572702479831, + "learning_rate": 0.003, + "loss": 4.069, + "step": 6586 + }, + { + "epoch": 0.06587, + "grad_norm": 0.8978047159555252, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 6587 + }, + { + "epoch": 0.06588, + "grad_norm": 0.827033589453339, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 6588 + }, + { + "epoch": 0.06589, + "grad_norm": 0.8211186750627676, + "learning_rate": 0.003, + "loss": 4.07, + "step": 6589 + }, + { + "epoch": 0.0659, + "grad_norm": 0.8442088987694915, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 6590 + }, + { + "epoch": 0.06591, + "grad_norm": 1.0094167846920752, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 6591 + }, + { + "epoch": 0.06592, + "grad_norm": 1.13531615910892, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 6592 + }, + { + "epoch": 0.06593, + "grad_norm": 0.8982297119321998, + "learning_rate": 0.003, + "loss": 4.046, + "step": 6593 + }, + { + "epoch": 0.06594, + "grad_norm": 0.946141231020595, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 6594 + }, + { + "epoch": 0.06595, + "grad_norm": 0.9136182259899146, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 6595 + }, + { + "epoch": 0.06596, + "grad_norm": 0.8744151642325578, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 6596 + }, + { + "epoch": 0.06597, + "grad_norm": 0.9533603618553095, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 6597 + }, + { + "epoch": 0.06598, + "grad_norm": 1.0604427104690746, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 6598 + }, + { + "epoch": 0.06599, + "grad_norm": 1.067472601559575, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 6599 + }, + { + "epoch": 0.066, + "grad_norm": 1.0838031919616027, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 6600 + }, + { + "epoch": 0.06601, + "grad_norm": 1.0258280341899866, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 6601 + }, + { + "epoch": 0.06602, + "grad_norm": 1.0270369431845496, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 6602 + }, + { + "epoch": 0.06603, + "grad_norm": 1.0254464189303953, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 6603 + }, + { + "epoch": 0.06604, + "grad_norm": 1.0679326763588755, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6604 + }, + { + "epoch": 0.06605, + "grad_norm": 0.9415976678420839, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 6605 + }, + { + "epoch": 0.06606, + "grad_norm": 0.9849735102421308, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6606 + }, + { + "epoch": 0.06607, + "grad_norm": 1.1302001807963977, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 6607 + }, + { + "epoch": 0.06608, + "grad_norm": 1.0851366334623165, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 6608 + }, + { + "epoch": 0.06609, + "grad_norm": 1.2878538978737437, + "learning_rate": 0.003, + "loss": 4.07, + "step": 6609 + }, + { + "epoch": 0.0661, + "grad_norm": 0.9809189178062531, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 6610 + }, + { + "epoch": 0.06611, + "grad_norm": 0.9774112014402608, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 6611 + }, + { + "epoch": 0.06612, + "grad_norm": 1.074206882798712, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 6612 + }, + { + "epoch": 0.06613, + "grad_norm": 0.9119362233622336, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 6613 + }, + { + "epoch": 0.06614, + "grad_norm": 0.9092970755783418, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 6614 + }, + { + "epoch": 0.06615, + "grad_norm": 0.997529056172159, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 6615 + }, + { + "epoch": 0.06616, + "grad_norm": 1.0127096432931733, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 6616 + }, + { + "epoch": 0.06617, + "grad_norm": 1.047029183095652, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 6617 + }, + { + "epoch": 0.06618, + "grad_norm": 0.9374810538498864, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 6618 + }, + { + "epoch": 0.06619, + "grad_norm": 0.9483864493653071, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 6619 + }, + { + "epoch": 0.0662, + "grad_norm": 1.0543118381581815, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 6620 + }, + { + "epoch": 0.06621, + "grad_norm": 0.9495585297143883, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 6621 + }, + { + "epoch": 0.06622, + "grad_norm": 1.1120232278985485, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 6622 + }, + { + "epoch": 0.06623, + "grad_norm": 1.01560273953621, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 6623 + }, + { + "epoch": 0.06624, + "grad_norm": 1.1763569739613489, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 6624 + }, + { + "epoch": 0.06625, + "grad_norm": 1.1262447069862451, + "learning_rate": 0.003, + "loss": 4.062, + "step": 6625 + }, + { + "epoch": 0.06626, + "grad_norm": 0.9382970303663091, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 6626 + }, + { + "epoch": 0.06627, + "grad_norm": 1.031419952696058, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 6627 + }, + { + "epoch": 0.06628, + "grad_norm": 1.1000233296997506, + "learning_rate": 0.003, + "loss": 4.073, + "step": 6628 + }, + { + "epoch": 0.06629, + "grad_norm": 1.0494015088404511, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 6629 + }, + { + "epoch": 0.0663, + "grad_norm": 0.9410676376451073, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 6630 + }, + { + "epoch": 0.06631, + "grad_norm": 1.0689157449200972, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 6631 + }, + { + "epoch": 0.06632, + "grad_norm": 0.9674796376390861, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 6632 + }, + { + "epoch": 0.06633, + "grad_norm": 1.0093941937371331, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 6633 + }, + { + "epoch": 0.06634, + "grad_norm": 1.2337187993906655, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 6634 + }, + { + "epoch": 0.06635, + "grad_norm": 0.9607834095606316, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 6635 + }, + { + "epoch": 0.06636, + "grad_norm": 0.9317793609051973, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 6636 + }, + { + "epoch": 0.06637, + "grad_norm": 0.9290660104219854, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 6637 + }, + { + "epoch": 0.06638, + "grad_norm": 0.8475879085595186, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 6638 + }, + { + "epoch": 0.06639, + "grad_norm": 0.7941589369777531, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 6639 + }, + { + "epoch": 0.0664, + "grad_norm": 0.8043535640617332, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 6640 + }, + { + "epoch": 0.06641, + "grad_norm": 0.9938652956113684, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 6641 + }, + { + "epoch": 0.06642, + "grad_norm": 1.1838239696711994, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 6642 + }, + { + "epoch": 0.06643, + "grad_norm": 0.8933675359063811, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 6643 + }, + { + "epoch": 0.06644, + "grad_norm": 0.7456469496808978, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 6644 + }, + { + "epoch": 0.06645, + "grad_norm": 0.7991018018930186, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 6645 + }, + { + "epoch": 0.06646, + "grad_norm": 0.9578935057690124, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 6646 + }, + { + "epoch": 0.06647, + "grad_norm": 1.0975543814373472, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 6647 + }, + { + "epoch": 0.06648, + "grad_norm": 0.9324693403913995, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 6648 + }, + { + "epoch": 0.06649, + "grad_norm": 0.783828708420081, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 6649 + }, + { + "epoch": 0.0665, + "grad_norm": 0.8171820344699803, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 6650 + }, + { + "epoch": 0.06651, + "grad_norm": 1.0398397978025293, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 6651 + }, + { + "epoch": 0.06652, + "grad_norm": 1.1078725880727078, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 6652 + }, + { + "epoch": 0.06653, + "grad_norm": 1.05746120078045, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 6653 + }, + { + "epoch": 0.06654, + "grad_norm": 1.1809835092262086, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 6654 + }, + { + "epoch": 0.06655, + "grad_norm": 0.9734402094947942, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6655 + }, + { + "epoch": 0.06656, + "grad_norm": 1.1076276318277252, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 6656 + }, + { + "epoch": 0.06657, + "grad_norm": 0.977985457890705, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 6657 + }, + { + "epoch": 0.06658, + "grad_norm": 0.9396247788800478, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 6658 + }, + { + "epoch": 0.06659, + "grad_norm": 1.0889586998264722, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 6659 + }, + { + "epoch": 0.0666, + "grad_norm": 1.0252758829952033, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6660 + }, + { + "epoch": 0.06661, + "grad_norm": 1.0468601933896104, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 6661 + }, + { + "epoch": 0.06662, + "grad_norm": 1.1000476753001847, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 6662 + }, + { + "epoch": 0.06663, + "grad_norm": 1.1085660118615401, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 6663 + }, + { + "epoch": 0.06664, + "grad_norm": 0.9874276413847503, + "learning_rate": 0.003, + "loss": 4.071, + "step": 6664 + }, + { + "epoch": 0.06665, + "grad_norm": 1.0456032019182708, + "learning_rate": 0.003, + "loss": 4.1, + "step": 6665 + }, + { + "epoch": 0.06666, + "grad_norm": 1.0094947821010092, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 6666 + }, + { + "epoch": 0.06667, + "grad_norm": 1.109076869997086, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 6667 + }, + { + "epoch": 0.06668, + "grad_norm": 1.0088933647503096, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 6668 + }, + { + "epoch": 0.06669, + "grad_norm": 1.1968841584443815, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 6669 + }, + { + "epoch": 0.0667, + "grad_norm": 1.1241320277325377, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 6670 + }, + { + "epoch": 0.06671, + "grad_norm": 1.141904583401614, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 6671 + }, + { + "epoch": 0.06672, + "grad_norm": 1.062629879867344, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 6672 + }, + { + "epoch": 0.06673, + "grad_norm": 0.905571170038888, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 6673 + }, + { + "epoch": 0.06674, + "grad_norm": 0.9626950086254233, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 6674 + }, + { + "epoch": 0.06675, + "grad_norm": 1.0975398719716734, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 6675 + }, + { + "epoch": 0.06676, + "grad_norm": 1.3284543966412756, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 6676 + }, + { + "epoch": 0.06677, + "grad_norm": 0.8172955955655188, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 6677 + }, + { + "epoch": 0.06678, + "grad_norm": 0.9850016002008366, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 6678 + }, + { + "epoch": 0.06679, + "grad_norm": 1.0932275223248946, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 6679 + }, + { + "epoch": 0.0668, + "grad_norm": 1.148365923927585, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 6680 + }, + { + "epoch": 0.06681, + "grad_norm": 1.0338287959641415, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 6681 + }, + { + "epoch": 0.06682, + "grad_norm": 1.179731822743392, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 6682 + }, + { + "epoch": 0.06683, + "grad_norm": 0.8843991533234951, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6683 + }, + { + "epoch": 0.06684, + "grad_norm": 0.9108762252979913, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 6684 + }, + { + "epoch": 0.06685, + "grad_norm": 1.0436969205506577, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 6685 + }, + { + "epoch": 0.06686, + "grad_norm": 1.0267003835359743, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 6686 + }, + { + "epoch": 0.06687, + "grad_norm": 0.9909041191744213, + "learning_rate": 0.003, + "loss": 4.03, + "step": 6687 + }, + { + "epoch": 0.06688, + "grad_norm": 1.0186735696395426, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 6688 + }, + { + "epoch": 0.06689, + "grad_norm": 1.0837124330823527, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 6689 + }, + { + "epoch": 0.0669, + "grad_norm": 0.9374724932078298, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 6690 + }, + { + "epoch": 0.06691, + "grad_norm": 0.9609019141278521, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 6691 + }, + { + "epoch": 0.06692, + "grad_norm": 0.9532183639646817, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 6692 + }, + { + "epoch": 0.06693, + "grad_norm": 0.9948461347977686, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 6693 + }, + { + "epoch": 0.06694, + "grad_norm": 0.927563382693705, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 6694 + }, + { + "epoch": 0.06695, + "grad_norm": 0.8983506006216038, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 6695 + }, + { + "epoch": 0.06696, + "grad_norm": 1.0195737792831439, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 6696 + }, + { + "epoch": 0.06697, + "grad_norm": 1.0461424467719074, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 6697 + }, + { + "epoch": 0.06698, + "grad_norm": 1.1896402560053234, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 6698 + }, + { + "epoch": 0.06699, + "grad_norm": 1.0331768787858662, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 6699 + }, + { + "epoch": 0.067, + "grad_norm": 1.0760979816453098, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 6700 + }, + { + "epoch": 0.06701, + "grad_norm": 0.9030506574904259, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 6701 + }, + { + "epoch": 0.06702, + "grad_norm": 0.9590465759481133, + "learning_rate": 0.003, + "loss": 4.056, + "step": 6702 + }, + { + "epoch": 0.06703, + "grad_norm": 1.0377738190882686, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 6703 + }, + { + "epoch": 0.06704, + "grad_norm": 0.9878474206212229, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 6704 + }, + { + "epoch": 0.06705, + "grad_norm": 1.266667575265577, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 6705 + }, + { + "epoch": 0.06706, + "grad_norm": 0.8290481300791042, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 6706 + }, + { + "epoch": 0.06707, + "grad_norm": 0.8267971520194893, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 6707 + }, + { + "epoch": 0.06708, + "grad_norm": 0.7961467186413875, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 6708 + }, + { + "epoch": 0.06709, + "grad_norm": 0.9824457399373988, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 6709 + }, + { + "epoch": 0.0671, + "grad_norm": 1.1961534967718657, + "learning_rate": 0.003, + "loss": 4.078, + "step": 6710 + }, + { + "epoch": 0.06711, + "grad_norm": 0.9265435402662816, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 6711 + }, + { + "epoch": 0.06712, + "grad_norm": 1.0157357369850826, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 6712 + }, + { + "epoch": 0.06713, + "grad_norm": 0.9690324224443148, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 6713 + }, + { + "epoch": 0.06714, + "grad_norm": 0.897252797835452, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 6714 + }, + { + "epoch": 0.06715, + "grad_norm": 1.0170717220350938, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 6715 + }, + { + "epoch": 0.06716, + "grad_norm": 1.0531655759972556, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 6716 + }, + { + "epoch": 0.06717, + "grad_norm": 1.083960596464857, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 6717 + }, + { + "epoch": 0.06718, + "grad_norm": 0.9551481210518388, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 6718 + }, + { + "epoch": 0.06719, + "grad_norm": 0.935024318023197, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 6719 + }, + { + "epoch": 0.0672, + "grad_norm": 0.9498980025125131, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 6720 + }, + { + "epoch": 0.06721, + "grad_norm": 1.2015504874657223, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 6721 + }, + { + "epoch": 0.06722, + "grad_norm": 0.9347001938513728, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 6722 + }, + { + "epoch": 0.06723, + "grad_norm": 1.018756831416971, + "learning_rate": 0.003, + "loss": 4.07, + "step": 6723 + }, + { + "epoch": 0.06724, + "grad_norm": 1.0987676870246745, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 6724 + }, + { + "epoch": 0.06725, + "grad_norm": 1.0112178470596225, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 6725 + }, + { + "epoch": 0.06726, + "grad_norm": 1.112944901032879, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 6726 + }, + { + "epoch": 0.06727, + "grad_norm": 1.163602379560122, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 6727 + }, + { + "epoch": 0.06728, + "grad_norm": 0.9360909552819368, + "learning_rate": 0.003, + "loss": 4.026, + "step": 6728 + }, + { + "epoch": 0.06729, + "grad_norm": 1.0605315281502332, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 6729 + }, + { + "epoch": 0.0673, + "grad_norm": 1.0349633126182776, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 6730 + }, + { + "epoch": 0.06731, + "grad_norm": 1.2616197215123148, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 6731 + }, + { + "epoch": 0.06732, + "grad_norm": 1.0836805120817457, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 6732 + }, + { + "epoch": 0.06733, + "grad_norm": 0.9975329556874161, + "learning_rate": 0.003, + "loss": 4.066, + "step": 6733 + }, + { + "epoch": 0.06734, + "grad_norm": 1.0166554668806964, + "learning_rate": 0.003, + "loss": 4.063, + "step": 6734 + }, + { + "epoch": 0.06735, + "grad_norm": 0.9860686766932066, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 6735 + }, + { + "epoch": 0.06736, + "grad_norm": 1.1900474439099948, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 6736 + }, + { + "epoch": 0.06737, + "grad_norm": 0.8977382894817897, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 6737 + }, + { + "epoch": 0.06738, + "grad_norm": 0.8986515344833599, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 6738 + }, + { + "epoch": 0.06739, + "grad_norm": 1.0321723301672268, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 6739 + }, + { + "epoch": 0.0674, + "grad_norm": 1.1818737667535166, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 6740 + }, + { + "epoch": 0.06741, + "grad_norm": 0.7306130026600984, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 6741 + }, + { + "epoch": 0.06742, + "grad_norm": 0.671364374797458, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 6742 + }, + { + "epoch": 0.06743, + "grad_norm": 0.7657736082400693, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 6743 + }, + { + "epoch": 0.06744, + "grad_norm": 0.8251136224218921, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 6744 + }, + { + "epoch": 0.06745, + "grad_norm": 0.7693016914924291, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 6745 + }, + { + "epoch": 0.06746, + "grad_norm": 0.8468866611491973, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 6746 + }, + { + "epoch": 0.06747, + "grad_norm": 1.0246297168064884, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 6747 + }, + { + "epoch": 0.06748, + "grad_norm": 1.1135872084646647, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 6748 + }, + { + "epoch": 0.06749, + "grad_norm": 0.8843107432077526, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 6749 + }, + { + "epoch": 0.0675, + "grad_norm": 0.9912073777138863, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 6750 + }, + { + "epoch": 0.06751, + "grad_norm": 1.2591178077510798, + "learning_rate": 0.003, + "loss": 4.08, + "step": 6751 + }, + { + "epoch": 0.06752, + "grad_norm": 0.8082384535393943, + "learning_rate": 0.003, + "loss": 4.065, + "step": 6752 + }, + { + "epoch": 0.06753, + "grad_norm": 0.7511358793452372, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 6753 + }, + { + "epoch": 0.06754, + "grad_norm": 0.867990155758509, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6754 + }, + { + "epoch": 0.06755, + "grad_norm": 0.9955414763715997, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 6755 + }, + { + "epoch": 0.06756, + "grad_norm": 1.2073694787838065, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 6756 + }, + { + "epoch": 0.06757, + "grad_norm": 1.0707480255468453, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 6757 + }, + { + "epoch": 0.06758, + "grad_norm": 1.1122443531121051, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 6758 + }, + { + "epoch": 0.06759, + "grad_norm": 0.9489611527342554, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 6759 + }, + { + "epoch": 0.0676, + "grad_norm": 1.0041886074885658, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 6760 + }, + { + "epoch": 0.06761, + "grad_norm": 1.355937161252965, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 6761 + }, + { + "epoch": 0.06762, + "grad_norm": 0.9894898503815498, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 6762 + }, + { + "epoch": 0.06763, + "grad_norm": 1.050921117069599, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 6763 + }, + { + "epoch": 0.06764, + "grad_norm": 1.067765228974404, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 6764 + }, + { + "epoch": 0.06765, + "grad_norm": 0.9545145035844784, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 6765 + }, + { + "epoch": 0.06766, + "grad_norm": 1.0023904128414394, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 6766 + }, + { + "epoch": 0.06767, + "grad_norm": 0.9593720771665164, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 6767 + }, + { + "epoch": 0.06768, + "grad_norm": 1.1273025407257062, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 6768 + }, + { + "epoch": 0.06769, + "grad_norm": 1.087476946360311, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 6769 + }, + { + "epoch": 0.0677, + "grad_norm": 1.1529163947763248, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 6770 + }, + { + "epoch": 0.06771, + "grad_norm": 0.7314271498801576, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 6771 + }, + { + "epoch": 0.06772, + "grad_norm": 0.7723436133611571, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6772 + }, + { + "epoch": 0.06773, + "grad_norm": 0.9678136783714796, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 6773 + }, + { + "epoch": 0.06774, + "grad_norm": 1.2406355654915857, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 6774 + }, + { + "epoch": 0.06775, + "grad_norm": 0.8156597708290578, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6775 + }, + { + "epoch": 0.06776, + "grad_norm": 0.9530712850122685, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 6776 + }, + { + "epoch": 0.06777, + "grad_norm": 1.081652928525485, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 6777 + }, + { + "epoch": 0.06778, + "grad_norm": 0.9997360227306185, + "learning_rate": 0.003, + "loss": 4.014, + "step": 6778 + }, + { + "epoch": 0.06779, + "grad_norm": 0.9050972292743207, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 6779 + }, + { + "epoch": 0.0678, + "grad_norm": 0.899616606188489, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 6780 + }, + { + "epoch": 0.06781, + "grad_norm": 1.035269328464622, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 6781 + }, + { + "epoch": 0.06782, + "grad_norm": 1.204532202978388, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 6782 + }, + { + "epoch": 0.06783, + "grad_norm": 0.9655678343763224, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 6783 + }, + { + "epoch": 0.06784, + "grad_norm": 0.900065330135723, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 6784 + }, + { + "epoch": 0.06785, + "grad_norm": 0.9737356124056987, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 6785 + }, + { + "epoch": 0.06786, + "grad_norm": 0.9974183872061985, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 6786 + }, + { + "epoch": 0.06787, + "grad_norm": 1.041494125314001, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 6787 + }, + { + "epoch": 0.06788, + "grad_norm": 1.083171903315331, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 6788 + }, + { + "epoch": 0.06789, + "grad_norm": 1.3039936977749973, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 6789 + }, + { + "epoch": 0.0679, + "grad_norm": 0.9733172088754207, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 6790 + }, + { + "epoch": 0.06791, + "grad_norm": 0.8961119962612736, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 6791 + }, + { + "epoch": 0.06792, + "grad_norm": 0.9367283411485818, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 6792 + }, + { + "epoch": 0.06793, + "grad_norm": 1.2662871097017818, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 6793 + }, + { + "epoch": 0.06794, + "grad_norm": 0.9235411725065059, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 6794 + }, + { + "epoch": 0.06795, + "grad_norm": 1.0564098546065013, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 6795 + }, + { + "epoch": 0.06796, + "grad_norm": 1.024685829459076, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 6796 + }, + { + "epoch": 0.06797, + "grad_norm": 1.120134124815487, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 6797 + }, + { + "epoch": 0.06798, + "grad_norm": 0.9517110352700969, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 6798 + }, + { + "epoch": 0.06799, + "grad_norm": 0.9692745939925019, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 6799 + }, + { + "epoch": 0.068, + "grad_norm": 1.0264417600972477, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 6800 + }, + { + "epoch": 0.06801, + "grad_norm": 1.2302920860016315, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 6801 + }, + { + "epoch": 0.06802, + "grad_norm": 1.0369213141884206, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 6802 + }, + { + "epoch": 0.06803, + "grad_norm": 1.0524636853985736, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 6803 + }, + { + "epoch": 0.06804, + "grad_norm": 1.0387365140475044, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 6804 + }, + { + "epoch": 0.06805, + "grad_norm": 1.2150483769149856, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6805 + }, + { + "epoch": 0.06806, + "grad_norm": 0.8807324303080953, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 6806 + }, + { + "epoch": 0.06807, + "grad_norm": 0.8373757284517774, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 6807 + }, + { + "epoch": 0.06808, + "grad_norm": 0.8400667752675757, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 6808 + }, + { + "epoch": 0.06809, + "grad_norm": 0.8369993186398413, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 6809 + }, + { + "epoch": 0.0681, + "grad_norm": 0.9006507393196466, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 6810 + }, + { + "epoch": 0.06811, + "grad_norm": 1.0077180114145714, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 6811 + }, + { + "epoch": 0.06812, + "grad_norm": 1.1521055645070202, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 6812 + }, + { + "epoch": 0.06813, + "grad_norm": 1.0625428176141143, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 6813 + }, + { + "epoch": 0.06814, + "grad_norm": 1.178996563449822, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 6814 + }, + { + "epoch": 0.06815, + "grad_norm": 0.8797766928663225, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 6815 + }, + { + "epoch": 0.06816, + "grad_norm": 0.8289246518934655, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 6816 + }, + { + "epoch": 0.06817, + "grad_norm": 0.8654988446799349, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 6817 + }, + { + "epoch": 0.06818, + "grad_norm": 1.0798387368738107, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 6818 + }, + { + "epoch": 0.06819, + "grad_norm": 1.1105690600865603, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 6819 + }, + { + "epoch": 0.0682, + "grad_norm": 1.0813449551126753, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 6820 + }, + { + "epoch": 0.06821, + "grad_norm": 0.96077795414745, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 6821 + }, + { + "epoch": 0.06822, + "grad_norm": 0.9659380763604074, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 6822 + }, + { + "epoch": 0.06823, + "grad_norm": 1.10215696178397, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 6823 + }, + { + "epoch": 0.06824, + "grad_norm": 1.034364867848419, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 6824 + }, + { + "epoch": 0.06825, + "grad_norm": 0.9937444135494871, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 6825 + }, + { + "epoch": 0.06826, + "grad_norm": 0.9764377761142184, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 6826 + }, + { + "epoch": 0.06827, + "grad_norm": 0.969374372811426, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 6827 + }, + { + "epoch": 0.06828, + "grad_norm": 1.0350502895060627, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 6828 + }, + { + "epoch": 0.06829, + "grad_norm": 1.241961679572367, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6829 + }, + { + "epoch": 0.0683, + "grad_norm": 1.025448801706437, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 6830 + }, + { + "epoch": 0.06831, + "grad_norm": 0.9950647336149414, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 6831 + }, + { + "epoch": 0.06832, + "grad_norm": 1.1012052576359934, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 6832 + }, + { + "epoch": 0.06833, + "grad_norm": 0.8956721313879287, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 6833 + }, + { + "epoch": 0.06834, + "grad_norm": 1.1464044341999087, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 6834 + }, + { + "epoch": 0.06835, + "grad_norm": 1.09845989400467, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 6835 + }, + { + "epoch": 0.06836, + "grad_norm": 1.1577607159035546, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 6836 + }, + { + "epoch": 0.06837, + "grad_norm": 0.9624906863376643, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 6837 + }, + { + "epoch": 0.06838, + "grad_norm": 1.0526594999167607, + "learning_rate": 0.003, + "loss": 4.048, + "step": 6838 + }, + { + "epoch": 0.06839, + "grad_norm": 1.0673019415333493, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 6839 + }, + { + "epoch": 0.0684, + "grad_norm": 1.012103935387283, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 6840 + }, + { + "epoch": 0.06841, + "grad_norm": 0.9147560220218148, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 6841 + }, + { + "epoch": 0.06842, + "grad_norm": 0.9643937321782413, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 6842 + }, + { + "epoch": 0.06843, + "grad_norm": 1.0039478407079416, + "learning_rate": 0.003, + "loss": 4.053, + "step": 6843 + }, + { + "epoch": 0.06844, + "grad_norm": 1.2025234350492864, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 6844 + }, + { + "epoch": 0.06845, + "grad_norm": 1.0176494517204697, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 6845 + }, + { + "epoch": 0.06846, + "grad_norm": 0.9635201111106092, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6846 + }, + { + "epoch": 0.06847, + "grad_norm": 0.9695541827730285, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 6847 + }, + { + "epoch": 0.06848, + "grad_norm": 1.055028109497596, + "learning_rate": 0.003, + "loss": 4.058, + "step": 6848 + }, + { + "epoch": 0.06849, + "grad_norm": 1.2623333403844699, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6849 + }, + { + "epoch": 0.0685, + "grad_norm": 1.0979408883303405, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 6850 + }, + { + "epoch": 0.06851, + "grad_norm": 1.2299990074107543, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 6851 + }, + { + "epoch": 0.06852, + "grad_norm": 0.8622471399851679, + "learning_rate": 0.003, + "loss": 4.038, + "step": 6852 + }, + { + "epoch": 0.06853, + "grad_norm": 0.797053506469073, + "learning_rate": 0.003, + "loss": 4.031, + "step": 6853 + }, + { + "epoch": 0.06854, + "grad_norm": 0.7428713139049907, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 6854 + }, + { + "epoch": 0.06855, + "grad_norm": 0.8702478930012675, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 6855 + }, + { + "epoch": 0.06856, + "grad_norm": 0.9565155211806974, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 6856 + }, + { + "epoch": 0.06857, + "grad_norm": 0.9401346796927762, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 6857 + }, + { + "epoch": 0.06858, + "grad_norm": 1.046352232064668, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 6858 + }, + { + "epoch": 0.06859, + "grad_norm": 1.1113366890222538, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 6859 + }, + { + "epoch": 0.0686, + "grad_norm": 1.0332225044816656, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 6860 + }, + { + "epoch": 0.06861, + "grad_norm": 1.037641447920837, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 6861 + }, + { + "epoch": 0.06862, + "grad_norm": 1.0707793888395856, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 6862 + }, + { + "epoch": 0.06863, + "grad_norm": 1.3025261367467222, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 6863 + }, + { + "epoch": 0.06864, + "grad_norm": 0.7117943067157152, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 6864 + }, + { + "epoch": 0.06865, + "grad_norm": 0.8149012494620742, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 6865 + }, + { + "epoch": 0.06866, + "grad_norm": 0.9768992618939744, + "learning_rate": 0.003, + "loss": 4.061, + "step": 6866 + }, + { + "epoch": 0.06867, + "grad_norm": 1.2876703502703644, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 6867 + }, + { + "epoch": 0.06868, + "grad_norm": 0.8939557694074772, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 6868 + }, + { + "epoch": 0.06869, + "grad_norm": 0.9739018190603045, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 6869 + }, + { + "epoch": 0.0687, + "grad_norm": 1.0808672350561943, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 6870 + }, + { + "epoch": 0.06871, + "grad_norm": 0.9732612594475262, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 6871 + }, + { + "epoch": 0.06872, + "grad_norm": 0.9231941574627566, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 6872 + }, + { + "epoch": 0.06873, + "grad_norm": 1.0038158255901826, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 6873 + }, + { + "epoch": 0.06874, + "grad_norm": 1.1994966137942642, + "learning_rate": 0.003, + "loss": 4.064, + "step": 6874 + }, + { + "epoch": 0.06875, + "grad_norm": 0.9012267255930811, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 6875 + }, + { + "epoch": 0.06876, + "grad_norm": 0.958415530478731, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 6876 + }, + { + "epoch": 0.06877, + "grad_norm": 0.9667858571717155, + "learning_rate": 0.003, + "loss": 4.038, + "step": 6877 + }, + { + "epoch": 0.06878, + "grad_norm": 1.2429535644925198, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 6878 + }, + { + "epoch": 0.06879, + "grad_norm": 0.8678125841491381, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 6879 + }, + { + "epoch": 0.0688, + "grad_norm": 0.9483241231979963, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 6880 + }, + { + "epoch": 0.06881, + "grad_norm": 1.1081275778671038, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 6881 + }, + { + "epoch": 0.06882, + "grad_norm": 0.8941631616443152, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 6882 + }, + { + "epoch": 0.06883, + "grad_norm": 1.0008866943662982, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 6883 + }, + { + "epoch": 0.06884, + "grad_norm": 1.1763212407023795, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 6884 + }, + { + "epoch": 0.06885, + "grad_norm": 0.947660863840069, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 6885 + }, + { + "epoch": 0.06886, + "grad_norm": 1.0445333316933652, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 6886 + }, + { + "epoch": 0.06887, + "grad_norm": 1.2206210118596559, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 6887 + }, + { + "epoch": 0.06888, + "grad_norm": 1.0692166428843015, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 6888 + }, + { + "epoch": 0.06889, + "grad_norm": 0.9355380989508778, + "learning_rate": 0.003, + "loss": 4.05, + "step": 6889 + }, + { + "epoch": 0.0689, + "grad_norm": 1.158851574547141, + "learning_rate": 0.003, + "loss": 4.033, + "step": 6890 + }, + { + "epoch": 0.06891, + "grad_norm": 0.8775458730749034, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 6891 + }, + { + "epoch": 0.06892, + "grad_norm": 1.0821142493586515, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 6892 + }, + { + "epoch": 0.06893, + "grad_norm": 1.0013455149294792, + "learning_rate": 0.003, + "loss": 4.054, + "step": 6893 + }, + { + "epoch": 0.06894, + "grad_norm": 1.1973645911348245, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 6894 + }, + { + "epoch": 0.06895, + "grad_norm": 0.9524668934091803, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 6895 + }, + { + "epoch": 0.06896, + "grad_norm": 1.042772383295928, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 6896 + }, + { + "epoch": 0.06897, + "grad_norm": 1.0973998064718429, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 6897 + }, + { + "epoch": 0.06898, + "grad_norm": 0.941011645940574, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 6898 + }, + { + "epoch": 0.06899, + "grad_norm": 0.9259127376564472, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 6899 + }, + { + "epoch": 0.069, + "grad_norm": 1.0042825475944963, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 6900 + }, + { + "epoch": 0.06901, + "grad_norm": 1.0504783164304958, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 6901 + }, + { + "epoch": 0.06902, + "grad_norm": 1.001395184859859, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 6902 + }, + { + "epoch": 0.06903, + "grad_norm": 0.9533095870741758, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 6903 + }, + { + "epoch": 0.06904, + "grad_norm": 0.8946382452071323, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 6904 + }, + { + "epoch": 0.06905, + "grad_norm": 0.9637064412868039, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 6905 + }, + { + "epoch": 0.06906, + "grad_norm": 1.0077282899066335, + "learning_rate": 0.003, + "loss": 4.023, + "step": 6906 + }, + { + "epoch": 0.06907, + "grad_norm": 1.1560211480862905, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 6907 + }, + { + "epoch": 0.06908, + "grad_norm": 0.9093203754997685, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 6908 + }, + { + "epoch": 0.06909, + "grad_norm": 0.9106719670133724, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 6909 + }, + { + "epoch": 0.0691, + "grad_norm": 1.0352808190093583, + "learning_rate": 0.003, + "loss": 4.027, + "step": 6910 + }, + { + "epoch": 0.06911, + "grad_norm": 1.172505571950034, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 6911 + }, + { + "epoch": 0.06912, + "grad_norm": 1.004717986434819, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 6912 + }, + { + "epoch": 0.06913, + "grad_norm": 1.0346811560718048, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 6913 + }, + { + "epoch": 0.06914, + "grad_norm": 0.9902540554926873, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 6914 + }, + { + "epoch": 0.06915, + "grad_norm": 1.3265154292186019, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 6915 + }, + { + "epoch": 0.06916, + "grad_norm": 1.0096731892843975, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 6916 + }, + { + "epoch": 0.06917, + "grad_norm": 1.1017730347115613, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 6917 + }, + { + "epoch": 0.06918, + "grad_norm": 1.2367019859366197, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 6918 + }, + { + "epoch": 0.06919, + "grad_norm": 0.9219679674649628, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6919 + }, + { + "epoch": 0.0692, + "grad_norm": 0.991754900343874, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 6920 + }, + { + "epoch": 0.06921, + "grad_norm": 1.2175829497274775, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 6921 + }, + { + "epoch": 0.06922, + "grad_norm": 0.9154029403754875, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 6922 + }, + { + "epoch": 0.06923, + "grad_norm": 1.0695987825330893, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 6923 + }, + { + "epoch": 0.06924, + "grad_norm": 0.940173575072378, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 6924 + }, + { + "epoch": 0.06925, + "grad_norm": 1.1072649608493197, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 6925 + }, + { + "epoch": 0.06926, + "grad_norm": 1.0433608354839305, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 6926 + }, + { + "epoch": 0.06927, + "grad_norm": 1.0081972527274148, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 6927 + }, + { + "epoch": 0.06928, + "grad_norm": 1.1110960603111513, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 6928 + }, + { + "epoch": 0.06929, + "grad_norm": 1.020106821680874, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 6929 + }, + { + "epoch": 0.0693, + "grad_norm": 1.1534854540437196, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 6930 + }, + { + "epoch": 0.06931, + "grad_norm": 0.9801283370812117, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 6931 + }, + { + "epoch": 0.06932, + "grad_norm": 1.0759745476144513, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 6932 + }, + { + "epoch": 0.06933, + "grad_norm": 1.1318320501940358, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 6933 + }, + { + "epoch": 0.06934, + "grad_norm": 0.9953954741024389, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 6934 + }, + { + "epoch": 0.06935, + "grad_norm": 1.0404905470053114, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 6935 + }, + { + "epoch": 0.06936, + "grad_norm": 1.0468597561286765, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 6936 + }, + { + "epoch": 0.06937, + "grad_norm": 1.0201183676680066, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 6937 + }, + { + "epoch": 0.06938, + "grad_norm": 1.0787460833199358, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 6938 + }, + { + "epoch": 0.06939, + "grad_norm": 1.0901798974191894, + "learning_rate": 0.003, + "loss": 4.059, + "step": 6939 + }, + { + "epoch": 0.0694, + "grad_norm": 0.9385122020121525, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 6940 + }, + { + "epoch": 0.06941, + "grad_norm": 1.09751042724724, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 6941 + }, + { + "epoch": 0.06942, + "grad_norm": 1.0152502004249075, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 6942 + }, + { + "epoch": 0.06943, + "grad_norm": 1.1106778728034443, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 6943 + }, + { + "epoch": 0.06944, + "grad_norm": 0.9402245095674768, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 6944 + }, + { + "epoch": 0.06945, + "grad_norm": 1.0498365010646888, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 6945 + }, + { + "epoch": 0.06946, + "grad_norm": 1.04938260769981, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 6946 + }, + { + "epoch": 0.06947, + "grad_norm": 1.2040313407275116, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 6947 + }, + { + "epoch": 0.06948, + "grad_norm": 0.9174434519705801, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 6948 + }, + { + "epoch": 0.06949, + "grad_norm": 1.056561705670037, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 6949 + }, + { + "epoch": 0.0695, + "grad_norm": 1.1160879395367673, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 6950 + }, + { + "epoch": 0.06951, + "grad_norm": 0.9311634834480271, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 6951 + }, + { + "epoch": 0.06952, + "grad_norm": 1.0397971869811435, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 6952 + }, + { + "epoch": 0.06953, + "grad_norm": 1.1245034234997564, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 6953 + }, + { + "epoch": 0.06954, + "grad_norm": 1.0386924889956406, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 6954 + }, + { + "epoch": 0.06955, + "grad_norm": 0.9732927657279983, + "learning_rate": 0.003, + "loss": 4.064, + "step": 6955 + }, + { + "epoch": 0.06956, + "grad_norm": 0.9919815526911104, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 6956 + }, + { + "epoch": 0.06957, + "grad_norm": 0.9511426911729295, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 6957 + }, + { + "epoch": 0.06958, + "grad_norm": 0.936900952232894, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 6958 + }, + { + "epoch": 0.06959, + "grad_norm": 0.9506093199052728, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 6959 + }, + { + "epoch": 0.0696, + "grad_norm": 0.894303598657002, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 6960 + }, + { + "epoch": 0.06961, + "grad_norm": 0.860669748877687, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 6961 + }, + { + "epoch": 0.06962, + "grad_norm": 0.9634371666233146, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 6962 + }, + { + "epoch": 0.06963, + "grad_norm": 1.0960373688281182, + "learning_rate": 0.003, + "loss": 4.064, + "step": 6963 + }, + { + "epoch": 0.06964, + "grad_norm": 1.1321187322512247, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 6964 + }, + { + "epoch": 0.06965, + "grad_norm": 1.0060414603670687, + "learning_rate": 0.003, + "loss": 4.043, + "step": 6965 + }, + { + "epoch": 0.06966, + "grad_norm": 1.0380239906833486, + "learning_rate": 0.003, + "loss": 4.066, + "step": 6966 + }, + { + "epoch": 0.06967, + "grad_norm": 1.1084883931180407, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 6967 + }, + { + "epoch": 0.06968, + "grad_norm": 0.9467387430672358, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 6968 + }, + { + "epoch": 0.06969, + "grad_norm": 0.9622808523217883, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 6969 + }, + { + "epoch": 0.0697, + "grad_norm": 1.2314020219415485, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6970 + }, + { + "epoch": 0.06971, + "grad_norm": 1.1260021071930193, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 6971 + }, + { + "epoch": 0.06972, + "grad_norm": 0.9154963484107886, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 6972 + }, + { + "epoch": 0.06973, + "grad_norm": 0.8840670259700667, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 6973 + }, + { + "epoch": 0.06974, + "grad_norm": 0.8851178357236568, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 6974 + }, + { + "epoch": 0.06975, + "grad_norm": 0.9524411300067905, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 6975 + }, + { + "epoch": 0.06976, + "grad_norm": 1.088636962287439, + "learning_rate": 0.003, + "loss": 4.023, + "step": 6976 + }, + { + "epoch": 0.06977, + "grad_norm": 1.1071976636641974, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 6977 + }, + { + "epoch": 0.06978, + "grad_norm": 0.9798522761711383, + "learning_rate": 0.003, + "loss": 4.025, + "step": 6978 + }, + { + "epoch": 0.06979, + "grad_norm": 0.807513012542661, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 6979 + }, + { + "epoch": 0.0698, + "grad_norm": 1.0085245208426825, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 6980 + }, + { + "epoch": 0.06981, + "grad_norm": 1.0861815134074257, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 6981 + }, + { + "epoch": 0.06982, + "grad_norm": 0.7980450645757602, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 6982 + }, + { + "epoch": 0.06983, + "grad_norm": 0.9321566297048406, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 6983 + }, + { + "epoch": 0.06984, + "grad_norm": 1.0721971827592451, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 6984 + }, + { + "epoch": 0.06985, + "grad_norm": 1.0237038863583252, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 6985 + }, + { + "epoch": 0.06986, + "grad_norm": 1.0546187122638413, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 6986 + }, + { + "epoch": 0.06987, + "grad_norm": 1.1037883958736543, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 6987 + }, + { + "epoch": 0.06988, + "grad_norm": 1.0507061025832318, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 6988 + }, + { + "epoch": 0.06989, + "grad_norm": 1.0236366890111934, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 6989 + }, + { + "epoch": 0.0699, + "grad_norm": 1.133782459769843, + "learning_rate": 0.003, + "loss": 4.035, + "step": 6990 + }, + { + "epoch": 0.06991, + "grad_norm": 0.9260651932828348, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 6991 + }, + { + "epoch": 0.06992, + "grad_norm": 0.9925580541567944, + "learning_rate": 0.003, + "loss": 4.031, + "step": 6992 + }, + { + "epoch": 0.06993, + "grad_norm": 1.067990124386425, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 6993 + }, + { + "epoch": 0.06994, + "grad_norm": 0.9360919285448381, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 6994 + }, + { + "epoch": 0.06995, + "grad_norm": 0.96458377970943, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 6995 + }, + { + "epoch": 0.06996, + "grad_norm": 1.148084887236895, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 6996 + }, + { + "epoch": 0.06997, + "grad_norm": 0.9046547601765281, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 6997 + }, + { + "epoch": 0.06998, + "grad_norm": 0.8772615969985239, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 6998 + }, + { + "epoch": 0.06999, + "grad_norm": 0.9340057257128744, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 6999 + }, + { + "epoch": 0.07, + "grad_norm": 1.0565666461050007, + "learning_rate": 0.003, + "loss": 4.039, + "step": 7000 + }, + { + "epoch": 0.07001, + "grad_norm": 0.9426298349737786, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 7001 + }, + { + "epoch": 0.07002, + "grad_norm": 1.0664130531690763, + "learning_rate": 0.003, + "loss": 4.048, + "step": 7002 + }, + { + "epoch": 0.07003, + "grad_norm": 1.1291863119100505, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 7003 + }, + { + "epoch": 0.07004, + "grad_norm": 1.0642470026156752, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 7004 + }, + { + "epoch": 0.07005, + "grad_norm": 1.2390880199207346, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 7005 + }, + { + "epoch": 0.07006, + "grad_norm": 0.9062982273502466, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 7006 + }, + { + "epoch": 0.07007, + "grad_norm": 1.1259354802693449, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 7007 + }, + { + "epoch": 0.07008, + "grad_norm": 1.0839899240911712, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7008 + }, + { + "epoch": 0.07009, + "grad_norm": 1.068714363123779, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 7009 + }, + { + "epoch": 0.0701, + "grad_norm": 0.9452433028423766, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 7010 + }, + { + "epoch": 0.07011, + "grad_norm": 0.9919155661649216, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 7011 + }, + { + "epoch": 0.07012, + "grad_norm": 1.2609870971082537, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 7012 + }, + { + "epoch": 0.07013, + "grad_norm": 0.9993261593044893, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 7013 + }, + { + "epoch": 0.07014, + "grad_norm": 1.0685461199627173, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 7014 + }, + { + "epoch": 0.07015, + "grad_norm": 0.8992285999432262, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 7015 + }, + { + "epoch": 0.07016, + "grad_norm": 0.8983621405017727, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 7016 + }, + { + "epoch": 0.07017, + "grad_norm": 0.9901752576746331, + "learning_rate": 0.003, + "loss": 4.029, + "step": 7017 + }, + { + "epoch": 0.07018, + "grad_norm": 1.1074224986234265, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 7018 + }, + { + "epoch": 0.07019, + "grad_norm": 1.057295544596109, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 7019 + }, + { + "epoch": 0.0702, + "grad_norm": 1.0481341550110743, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 7020 + }, + { + "epoch": 0.07021, + "grad_norm": 1.0407732639272513, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 7021 + }, + { + "epoch": 0.07022, + "grad_norm": 1.1448172616705823, + "learning_rate": 0.003, + "loss": 4.066, + "step": 7022 + }, + { + "epoch": 0.07023, + "grad_norm": 1.0313350981765694, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 7023 + }, + { + "epoch": 0.07024, + "grad_norm": 1.0020252731504273, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 7024 + }, + { + "epoch": 0.07025, + "grad_norm": 1.1359222787122534, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 7025 + }, + { + "epoch": 0.07026, + "grad_norm": 1.1041895748549755, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 7026 + }, + { + "epoch": 0.07027, + "grad_norm": 0.9841999651428623, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 7027 + }, + { + "epoch": 0.07028, + "grad_norm": 1.025891345196206, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 7028 + }, + { + "epoch": 0.07029, + "grad_norm": 0.9948854364695006, + "learning_rate": 0.003, + "loss": 4.037, + "step": 7029 + }, + { + "epoch": 0.0703, + "grad_norm": 0.9668935895325885, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 7030 + }, + { + "epoch": 0.07031, + "grad_norm": 1.0575663087737428, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 7031 + }, + { + "epoch": 0.07032, + "grad_norm": 0.9164666976168099, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 7032 + }, + { + "epoch": 0.07033, + "grad_norm": 0.9214454265285555, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 7033 + }, + { + "epoch": 0.07034, + "grad_norm": 0.9583267727565041, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 7034 + }, + { + "epoch": 0.07035, + "grad_norm": 1.3383936324024557, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 7035 + }, + { + "epoch": 0.07036, + "grad_norm": 0.8658946616429294, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 7036 + }, + { + "epoch": 0.07037, + "grad_norm": 0.9475554646881245, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 7037 + }, + { + "epoch": 0.07038, + "grad_norm": 0.8924840709336554, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 7038 + }, + { + "epoch": 0.07039, + "grad_norm": 0.8951723916341889, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7039 + }, + { + "epoch": 0.0704, + "grad_norm": 0.854416819628114, + "learning_rate": 0.003, + "loss": 4.066, + "step": 7040 + }, + { + "epoch": 0.07041, + "grad_norm": 0.9333352868640866, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 7041 + }, + { + "epoch": 0.07042, + "grad_norm": 1.094975909783682, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 7042 + }, + { + "epoch": 0.07043, + "grad_norm": 1.111332862841996, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 7043 + }, + { + "epoch": 0.07044, + "grad_norm": 1.0657438894896911, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 7044 + }, + { + "epoch": 0.07045, + "grad_norm": 1.214943958168645, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 7045 + }, + { + "epoch": 0.07046, + "grad_norm": 1.0052618021881827, + "learning_rate": 0.003, + "loss": 4.061, + "step": 7046 + }, + { + "epoch": 0.07047, + "grad_norm": 1.2069549174110699, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 7047 + }, + { + "epoch": 0.07048, + "grad_norm": 0.7843964117630299, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 7048 + }, + { + "epoch": 0.07049, + "grad_norm": 0.691374439878814, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 7049 + }, + { + "epoch": 0.0705, + "grad_norm": 0.7306826895178101, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 7050 + }, + { + "epoch": 0.07051, + "grad_norm": 0.8933100091919502, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 7051 + }, + { + "epoch": 0.07052, + "grad_norm": 1.0871930655736466, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 7052 + }, + { + "epoch": 0.07053, + "grad_norm": 1.306112767575483, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 7053 + }, + { + "epoch": 0.07054, + "grad_norm": 0.7906641478755512, + "learning_rate": 0.003, + "loss": 4.041, + "step": 7054 + }, + { + "epoch": 0.07055, + "grad_norm": 0.8522746819689732, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 7055 + }, + { + "epoch": 0.07056, + "grad_norm": 0.9004086354520517, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 7056 + }, + { + "epoch": 0.07057, + "grad_norm": 0.9624496801222439, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 7057 + }, + { + "epoch": 0.07058, + "grad_norm": 0.8808033290776586, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 7058 + }, + { + "epoch": 0.07059, + "grad_norm": 1.0148842321503468, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 7059 + }, + { + "epoch": 0.0706, + "grad_norm": 1.0661395469335966, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 7060 + }, + { + "epoch": 0.07061, + "grad_norm": 1.033267111562011, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 7061 + }, + { + "epoch": 0.07062, + "grad_norm": 1.1023501283703616, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 7062 + }, + { + "epoch": 0.07063, + "grad_norm": 1.1371048285194043, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 7063 + }, + { + "epoch": 0.07064, + "grad_norm": 1.0600698489467515, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 7064 + }, + { + "epoch": 0.07065, + "grad_norm": 1.0845674242904877, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7065 + }, + { + "epoch": 0.07066, + "grad_norm": 1.0117517612735492, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 7066 + }, + { + "epoch": 0.07067, + "grad_norm": 1.1834369979286008, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 7067 + }, + { + "epoch": 0.07068, + "grad_norm": 0.9912589707934546, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 7068 + }, + { + "epoch": 0.07069, + "grad_norm": 1.008241016661584, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 7069 + }, + { + "epoch": 0.0707, + "grad_norm": 1.1771304332271513, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 7070 + }, + { + "epoch": 0.07071, + "grad_norm": 1.1143211837122156, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 7071 + }, + { + "epoch": 0.07072, + "grad_norm": 0.9342779082515223, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 7072 + }, + { + "epoch": 0.07073, + "grad_norm": 0.8837711443464364, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 7073 + }, + { + "epoch": 0.07074, + "grad_norm": 1.0273879438054683, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 7074 + }, + { + "epoch": 0.07075, + "grad_norm": 1.1016591310376989, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 7075 + }, + { + "epoch": 0.07076, + "grad_norm": 1.105334861387564, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 7076 + }, + { + "epoch": 0.07077, + "grad_norm": 0.9146056746290705, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 7077 + }, + { + "epoch": 0.07078, + "grad_norm": 0.8766079673528181, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 7078 + }, + { + "epoch": 0.07079, + "grad_norm": 0.9316844058335465, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 7079 + }, + { + "epoch": 0.0708, + "grad_norm": 1.0866064222511407, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 7080 + }, + { + "epoch": 0.07081, + "grad_norm": 0.9789936615170471, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 7081 + }, + { + "epoch": 0.07082, + "grad_norm": 1.1021827553494439, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 7082 + }, + { + "epoch": 0.07083, + "grad_norm": 1.0245533209037434, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 7083 + }, + { + "epoch": 0.07084, + "grad_norm": 1.0621477148568337, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 7084 + }, + { + "epoch": 0.07085, + "grad_norm": 1.1601015540652249, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 7085 + }, + { + "epoch": 0.07086, + "grad_norm": 1.0854553734016084, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 7086 + }, + { + "epoch": 0.07087, + "grad_norm": 1.3922698043387116, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 7087 + }, + { + "epoch": 0.07088, + "grad_norm": 0.9399003500801127, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 7088 + }, + { + "epoch": 0.07089, + "grad_norm": 1.106468816582972, + "learning_rate": 0.003, + "loss": 4.047, + "step": 7089 + }, + { + "epoch": 0.0709, + "grad_norm": 1.1027532636474677, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 7090 + }, + { + "epoch": 0.07091, + "grad_norm": 1.0374251515271373, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 7091 + }, + { + "epoch": 0.07092, + "grad_norm": 1.1299226085447214, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 7092 + }, + { + "epoch": 0.07093, + "grad_norm": 0.9253623127929507, + "learning_rate": 0.003, + "loss": 4.064, + "step": 7093 + }, + { + "epoch": 0.07094, + "grad_norm": 1.1587332040574374, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 7094 + }, + { + "epoch": 0.07095, + "grad_norm": 0.7769235402798084, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 7095 + }, + { + "epoch": 0.07096, + "grad_norm": 0.8053305671402123, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 7096 + }, + { + "epoch": 0.07097, + "grad_norm": 0.9209272441259044, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 7097 + }, + { + "epoch": 0.07098, + "grad_norm": 1.0625889845082845, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 7098 + }, + { + "epoch": 0.07099, + "grad_norm": 1.0498047751329527, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 7099 + }, + { + "epoch": 0.071, + "grad_norm": 1.224649756474013, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 7100 + }, + { + "epoch": 0.07101, + "grad_norm": 1.008032355385532, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 7101 + }, + { + "epoch": 0.07102, + "grad_norm": 1.0752805417902511, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 7102 + }, + { + "epoch": 0.07103, + "grad_norm": 1.0092223551265274, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 7103 + }, + { + "epoch": 0.07104, + "grad_norm": 1.072487221815657, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 7104 + }, + { + "epoch": 0.07105, + "grad_norm": 1.1260402490166252, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 7105 + }, + { + "epoch": 0.07106, + "grad_norm": 0.9259910792008968, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 7106 + }, + { + "epoch": 0.07107, + "grad_norm": 0.9680226789669217, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 7107 + }, + { + "epoch": 0.07108, + "grad_norm": 0.935158998841827, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 7108 + }, + { + "epoch": 0.07109, + "grad_norm": 1.0085487694764077, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 7109 + }, + { + "epoch": 0.0711, + "grad_norm": 1.196671503733759, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 7110 + }, + { + "epoch": 0.07111, + "grad_norm": 0.9584229952834501, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 7111 + }, + { + "epoch": 0.07112, + "grad_norm": 0.9641919680290156, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 7112 + }, + { + "epoch": 0.07113, + "grad_norm": 0.9993706432398796, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 7113 + }, + { + "epoch": 0.07114, + "grad_norm": 1.1499489634058697, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 7114 + }, + { + "epoch": 0.07115, + "grad_norm": 1.194464795275503, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 7115 + }, + { + "epoch": 0.07116, + "grad_norm": 1.21121303150807, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 7116 + }, + { + "epoch": 0.07117, + "grad_norm": 1.0049044123637882, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 7117 + }, + { + "epoch": 0.07118, + "grad_norm": 1.2298880131028436, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 7118 + }, + { + "epoch": 0.07119, + "grad_norm": 0.8611868402134719, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 7119 + }, + { + "epoch": 0.0712, + "grad_norm": 0.7702510545372283, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 7120 + }, + { + "epoch": 0.07121, + "grad_norm": 0.8954788767538316, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 7121 + }, + { + "epoch": 0.07122, + "grad_norm": 1.0842710437040461, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 7122 + }, + { + "epoch": 0.07123, + "grad_norm": 0.9923425116430409, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 7123 + }, + { + "epoch": 0.07124, + "grad_norm": 1.1813157355705963, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 7124 + }, + { + "epoch": 0.07125, + "grad_norm": 1.2554701276201077, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 7125 + }, + { + "epoch": 0.07126, + "grad_norm": 0.9773001931060757, + "learning_rate": 0.003, + "loss": 4.074, + "step": 7126 + }, + { + "epoch": 0.07127, + "grad_norm": 1.0712269928450333, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 7127 + }, + { + "epoch": 0.07128, + "grad_norm": 1.0739153312552485, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 7128 + }, + { + "epoch": 0.07129, + "grad_norm": 1.041760051339172, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 7129 + }, + { + "epoch": 0.0713, + "grad_norm": 1.079087703256973, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 7130 + }, + { + "epoch": 0.07131, + "grad_norm": 0.9759469792704608, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7131 + }, + { + "epoch": 0.07132, + "grad_norm": 1.048658582228863, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 7132 + }, + { + "epoch": 0.07133, + "grad_norm": 1.0618378394098185, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 7133 + }, + { + "epoch": 0.07134, + "grad_norm": 1.1513266730887497, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 7134 + }, + { + "epoch": 0.07135, + "grad_norm": 0.9259298591427148, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 7135 + }, + { + "epoch": 0.07136, + "grad_norm": 1.0755737256450422, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 7136 + }, + { + "epoch": 0.07137, + "grad_norm": 1.177551226290782, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7137 + }, + { + "epoch": 0.07138, + "grad_norm": 1.22091533451439, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 7138 + }, + { + "epoch": 0.07139, + "grad_norm": 0.7759575681946775, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 7139 + }, + { + "epoch": 0.0714, + "grad_norm": 0.7408826660171673, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 7140 + }, + { + "epoch": 0.07141, + "grad_norm": 0.7146551856791451, + "learning_rate": 0.003, + "loss": 4.031, + "step": 7141 + }, + { + "epoch": 0.07142, + "grad_norm": 0.9506570889697072, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 7142 + }, + { + "epoch": 0.07143, + "grad_norm": 1.30251677781438, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 7143 + }, + { + "epoch": 0.07144, + "grad_norm": 0.8991561050670873, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 7144 + }, + { + "epoch": 0.07145, + "grad_norm": 1.0074973465006982, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 7145 + }, + { + "epoch": 0.07146, + "grad_norm": 1.0771529601001637, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 7146 + }, + { + "epoch": 0.07147, + "grad_norm": 0.8612872764436263, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 7147 + }, + { + "epoch": 0.07148, + "grad_norm": 0.9341113795199162, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 7148 + }, + { + "epoch": 0.07149, + "grad_norm": 1.0062478097452772, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 7149 + }, + { + "epoch": 0.0715, + "grad_norm": 1.0268598684975616, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 7150 + }, + { + "epoch": 0.07151, + "grad_norm": 1.0224439130582545, + "learning_rate": 0.003, + "loss": 4.053, + "step": 7151 + }, + { + "epoch": 0.07152, + "grad_norm": 1.0646975295350594, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 7152 + }, + { + "epoch": 0.07153, + "grad_norm": 1.0171664033211174, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 7153 + }, + { + "epoch": 0.07154, + "grad_norm": 1.0400329068005092, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 7154 + }, + { + "epoch": 0.07155, + "grad_norm": 0.9604129137098717, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 7155 + }, + { + "epoch": 0.07156, + "grad_norm": 1.0703502921341488, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 7156 + }, + { + "epoch": 0.07157, + "grad_norm": 1.2643947181831285, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 7157 + }, + { + "epoch": 0.07158, + "grad_norm": 0.8820140566698819, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 7158 + }, + { + "epoch": 0.07159, + "grad_norm": 0.9999756514750661, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 7159 + }, + { + "epoch": 0.0716, + "grad_norm": 1.2329285675483432, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 7160 + }, + { + "epoch": 0.07161, + "grad_norm": 0.9648860180858708, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 7161 + }, + { + "epoch": 0.07162, + "grad_norm": 1.1598981829481805, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 7162 + }, + { + "epoch": 0.07163, + "grad_norm": 1.0352892191066683, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 7163 + }, + { + "epoch": 0.07164, + "grad_norm": 1.1005133026385345, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 7164 + }, + { + "epoch": 0.07165, + "grad_norm": 1.081455462126825, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 7165 + }, + { + "epoch": 0.07166, + "grad_norm": 1.2353060960241047, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7166 + }, + { + "epoch": 0.07167, + "grad_norm": 0.9483072899067242, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 7167 + }, + { + "epoch": 0.07168, + "grad_norm": 0.9415550285281931, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 7168 + }, + { + "epoch": 0.07169, + "grad_norm": 1.0091462587193163, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 7169 + }, + { + "epoch": 0.0717, + "grad_norm": 1.0220680948634882, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 7170 + }, + { + "epoch": 0.07171, + "grad_norm": 0.9757190374038757, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 7171 + }, + { + "epoch": 0.07172, + "grad_norm": 1.1004971741989973, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 7172 + }, + { + "epoch": 0.07173, + "grad_norm": 0.9543571710406531, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 7173 + }, + { + "epoch": 0.07174, + "grad_norm": 1.087698472674678, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 7174 + }, + { + "epoch": 0.07175, + "grad_norm": 1.1109156909053697, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 7175 + }, + { + "epoch": 0.07176, + "grad_norm": 0.9503187829288862, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 7176 + }, + { + "epoch": 0.07177, + "grad_norm": 0.9936594901006058, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 7177 + }, + { + "epoch": 0.07178, + "grad_norm": 1.061316451230859, + "learning_rate": 0.003, + "loss": 4.021, + "step": 7178 + }, + { + "epoch": 0.07179, + "grad_norm": 0.9212825081885807, + "learning_rate": 0.003, + "loss": 4.053, + "step": 7179 + }, + { + "epoch": 0.0718, + "grad_norm": 1.0771371633803515, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 7180 + }, + { + "epoch": 0.07181, + "grad_norm": 1.0622270106395655, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 7181 + }, + { + "epoch": 0.07182, + "grad_norm": 1.2387354855038064, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 7182 + }, + { + "epoch": 0.07183, + "grad_norm": 0.9639603006489462, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 7183 + }, + { + "epoch": 0.07184, + "grad_norm": 1.0093619599090504, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 7184 + }, + { + "epoch": 0.07185, + "grad_norm": 0.9714974637831857, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 7185 + }, + { + "epoch": 0.07186, + "grad_norm": 0.9153837367744138, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 7186 + }, + { + "epoch": 0.07187, + "grad_norm": 0.9316846582768762, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 7187 + }, + { + "epoch": 0.07188, + "grad_norm": 1.0594153818209886, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 7188 + }, + { + "epoch": 0.07189, + "grad_norm": 0.918217273642505, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 7189 + }, + { + "epoch": 0.0719, + "grad_norm": 1.0085899488533316, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 7190 + }, + { + "epoch": 0.07191, + "grad_norm": 1.1236120434513654, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 7191 + }, + { + "epoch": 0.07192, + "grad_norm": 1.0142883651839383, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 7192 + }, + { + "epoch": 0.07193, + "grad_norm": 0.976978007096748, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 7193 + }, + { + "epoch": 0.07194, + "grad_norm": 1.1423821853316132, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 7194 + }, + { + "epoch": 0.07195, + "grad_norm": 1.105242324637109, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 7195 + }, + { + "epoch": 0.07196, + "grad_norm": 1.050425013065668, + "learning_rate": 0.003, + "loss": 4.049, + "step": 7196 + }, + { + "epoch": 0.07197, + "grad_norm": 1.1236152449610997, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 7197 + }, + { + "epoch": 0.07198, + "grad_norm": 1.0814707926776106, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 7198 + }, + { + "epoch": 0.07199, + "grad_norm": 1.1497210191318876, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 7199 + }, + { + "epoch": 0.072, + "grad_norm": 0.9814431913033573, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 7200 + }, + { + "epoch": 0.07201, + "grad_norm": 0.9399249347861436, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 7201 + }, + { + "epoch": 0.07202, + "grad_norm": 1.0283475742305797, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 7202 + }, + { + "epoch": 0.07203, + "grad_norm": 1.3436302183163868, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 7203 + }, + { + "epoch": 0.07204, + "grad_norm": 0.8118935525393225, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 7204 + }, + { + "epoch": 0.07205, + "grad_norm": 0.6513445176109008, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 7205 + }, + { + "epoch": 0.07206, + "grad_norm": 0.7054255168362077, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 7206 + }, + { + "epoch": 0.07207, + "grad_norm": 0.754052659367641, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 7207 + }, + { + "epoch": 0.07208, + "grad_norm": 0.9086098698061754, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 7208 + }, + { + "epoch": 0.07209, + "grad_norm": 1.1574952413730935, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 7209 + }, + { + "epoch": 0.0721, + "grad_norm": 0.8639251076932414, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 7210 + }, + { + "epoch": 0.07211, + "grad_norm": 0.8049066976652748, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 7211 + }, + { + "epoch": 0.07212, + "grad_norm": 0.9012892110405708, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 7212 + }, + { + "epoch": 0.07213, + "grad_norm": 1.0195248366641139, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 7213 + }, + { + "epoch": 0.07214, + "grad_norm": 1.0957107033402924, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 7214 + }, + { + "epoch": 0.07215, + "grad_norm": 1.130563317224989, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 7215 + }, + { + "epoch": 0.07216, + "grad_norm": 1.0776995902257729, + "learning_rate": 0.003, + "loss": 4.062, + "step": 7216 + }, + { + "epoch": 0.07217, + "grad_norm": 0.9762456922279278, + "learning_rate": 0.003, + "loss": 4.052, + "step": 7217 + }, + { + "epoch": 0.07218, + "grad_norm": 1.0572760232733118, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 7218 + }, + { + "epoch": 0.07219, + "grad_norm": 1.1200235428034362, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 7219 + }, + { + "epoch": 0.0722, + "grad_norm": 0.9279710532680243, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 7220 + }, + { + "epoch": 0.07221, + "grad_norm": 1.0865227213549595, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 7221 + }, + { + "epoch": 0.07222, + "grad_norm": 1.1530440294312476, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 7222 + }, + { + "epoch": 0.07223, + "grad_norm": 1.105728877799717, + "learning_rate": 0.003, + "loss": 4.065, + "step": 7223 + }, + { + "epoch": 0.07224, + "grad_norm": 1.246082917171359, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 7224 + }, + { + "epoch": 0.07225, + "grad_norm": 0.8775383173500949, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 7225 + }, + { + "epoch": 0.07226, + "grad_norm": 0.9055827431051069, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 7226 + }, + { + "epoch": 0.07227, + "grad_norm": 0.9729711259743795, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 7227 + }, + { + "epoch": 0.07228, + "grad_norm": 1.122946918012819, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 7228 + }, + { + "epoch": 0.07229, + "grad_norm": 1.145241716228926, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 7229 + }, + { + "epoch": 0.0723, + "grad_norm": 1.1809619676313377, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 7230 + }, + { + "epoch": 0.07231, + "grad_norm": 1.0638511137158617, + "learning_rate": 0.003, + "loss": 4.063, + "step": 7231 + }, + { + "epoch": 0.07232, + "grad_norm": 1.3174222410179524, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7232 + }, + { + "epoch": 0.07233, + "grad_norm": 0.7594174026538151, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 7233 + }, + { + "epoch": 0.07234, + "grad_norm": 0.7113003269832634, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 7234 + }, + { + "epoch": 0.07235, + "grad_norm": 0.7666612536000506, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 7235 + }, + { + "epoch": 0.07236, + "grad_norm": 1.0228879699076707, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 7236 + }, + { + "epoch": 0.07237, + "grad_norm": 1.0627669746296777, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 7237 + }, + { + "epoch": 0.07238, + "grad_norm": 0.9408632718140275, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 7238 + }, + { + "epoch": 0.07239, + "grad_norm": 1.07854762597468, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 7239 + }, + { + "epoch": 0.0724, + "grad_norm": 1.1116574655009852, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 7240 + }, + { + "epoch": 0.07241, + "grad_norm": 1.0625991735895157, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 7241 + }, + { + "epoch": 0.07242, + "grad_norm": 1.1130412168878459, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 7242 + }, + { + "epoch": 0.07243, + "grad_norm": 1.0523505008259475, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 7243 + }, + { + "epoch": 0.07244, + "grad_norm": 1.1814825129506323, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 7244 + }, + { + "epoch": 0.07245, + "grad_norm": 0.9545116614676915, + "learning_rate": 0.003, + "loss": 4.043, + "step": 7245 + }, + { + "epoch": 0.07246, + "grad_norm": 1.0800472059853927, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 7246 + }, + { + "epoch": 0.07247, + "grad_norm": 0.8747307815336857, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 7247 + }, + { + "epoch": 0.07248, + "grad_norm": 0.9457171551359618, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 7248 + }, + { + "epoch": 0.07249, + "grad_norm": 0.9746074762050819, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 7249 + }, + { + "epoch": 0.0725, + "grad_norm": 1.0830316386031944, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7250 + }, + { + "epoch": 0.07251, + "grad_norm": 1.1295155297093291, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 7251 + }, + { + "epoch": 0.07252, + "grad_norm": 1.0574374191488838, + "learning_rate": 0.003, + "loss": 3.993, + "step": 7252 + }, + { + "epoch": 0.07253, + "grad_norm": 1.1371667240401926, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 7253 + }, + { + "epoch": 0.07254, + "grad_norm": 0.8575120967224161, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 7254 + }, + { + "epoch": 0.07255, + "grad_norm": 1.2040896120213669, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 7255 + }, + { + "epoch": 0.07256, + "grad_norm": 1.0282986308414142, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 7256 + }, + { + "epoch": 0.07257, + "grad_norm": 1.1867368710695958, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 7257 + }, + { + "epoch": 0.07258, + "grad_norm": 0.8751280922843936, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 7258 + }, + { + "epoch": 0.07259, + "grad_norm": 0.9468806915219505, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 7259 + }, + { + "epoch": 0.0726, + "grad_norm": 1.13462604181537, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 7260 + }, + { + "epoch": 0.07261, + "grad_norm": 0.9480119807528755, + "learning_rate": 0.003, + "loss": 4.053, + "step": 7261 + }, + { + "epoch": 0.07262, + "grad_norm": 1.0086785799416162, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 7262 + }, + { + "epoch": 0.07263, + "grad_norm": 0.9776635908283169, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 7263 + }, + { + "epoch": 0.07264, + "grad_norm": 0.9280201688532035, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 7264 + }, + { + "epoch": 0.07265, + "grad_norm": 1.146251437770367, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 7265 + }, + { + "epoch": 0.07266, + "grad_norm": 0.9536640102408181, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 7266 + }, + { + "epoch": 0.07267, + "grad_norm": 1.1810792017759155, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 7267 + }, + { + "epoch": 0.07268, + "grad_norm": 1.0667635725102285, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 7268 + }, + { + "epoch": 0.07269, + "grad_norm": 1.0933436942514816, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 7269 + }, + { + "epoch": 0.0727, + "grad_norm": 0.9075857622412058, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 7270 + }, + { + "epoch": 0.07271, + "grad_norm": 0.9549614482171724, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 7271 + }, + { + "epoch": 0.07272, + "grad_norm": 0.8614083119020385, + "learning_rate": 0.003, + "loss": 4.037, + "step": 7272 + }, + { + "epoch": 0.07273, + "grad_norm": 0.8876688164940039, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 7273 + }, + { + "epoch": 0.07274, + "grad_norm": 0.9248498248735642, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 7274 + }, + { + "epoch": 0.07275, + "grad_norm": 0.9301192282673498, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 7275 + }, + { + "epoch": 0.07276, + "grad_norm": 1.0236905978389461, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 7276 + }, + { + "epoch": 0.07277, + "grad_norm": 1.2555861212594497, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 7277 + }, + { + "epoch": 0.07278, + "grad_norm": 1.0168850674837941, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 7278 + }, + { + "epoch": 0.07279, + "grad_norm": 1.3248139439976, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 7279 + }, + { + "epoch": 0.0728, + "grad_norm": 0.9735763353453082, + "learning_rate": 0.003, + "loss": 4.022, + "step": 7280 + }, + { + "epoch": 0.07281, + "grad_norm": 1.1238701240024835, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7281 + }, + { + "epoch": 0.07282, + "grad_norm": 1.075085316193029, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 7282 + }, + { + "epoch": 0.07283, + "grad_norm": 1.1149909879984707, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 7283 + }, + { + "epoch": 0.07284, + "grad_norm": 1.060193425500397, + "learning_rate": 0.003, + "loss": 4.035, + "step": 7284 + }, + { + "epoch": 0.07285, + "grad_norm": 1.1145512915255615, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 7285 + }, + { + "epoch": 0.07286, + "grad_norm": 1.0821681081123302, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 7286 + }, + { + "epoch": 0.07287, + "grad_norm": 0.9482005317460671, + "learning_rate": 0.003, + "loss": 4.058, + "step": 7287 + }, + { + "epoch": 0.07288, + "grad_norm": 1.1770572136072466, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 7288 + }, + { + "epoch": 0.07289, + "grad_norm": 1.0392539942924648, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 7289 + }, + { + "epoch": 0.0729, + "grad_norm": 1.093723077978444, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 7290 + }, + { + "epoch": 0.07291, + "grad_norm": 1.2345580518798212, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 7291 + }, + { + "epoch": 0.07292, + "grad_norm": 1.066201654699835, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 7292 + }, + { + "epoch": 0.07293, + "grad_norm": 1.1510430611988984, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 7293 + }, + { + "epoch": 0.07294, + "grad_norm": 0.9783586851033524, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 7294 + }, + { + "epoch": 0.07295, + "grad_norm": 1.125641508383537, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 7295 + }, + { + "epoch": 0.07296, + "grad_norm": 0.8931193549630414, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 7296 + }, + { + "epoch": 0.07297, + "grad_norm": 0.9640861362127288, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 7297 + }, + { + "epoch": 0.07298, + "grad_norm": 1.1090770552512412, + "learning_rate": 0.003, + "loss": 4.021, + "step": 7298 + }, + { + "epoch": 0.07299, + "grad_norm": 0.9922013741324502, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 7299 + }, + { + "epoch": 0.073, + "grad_norm": 1.0624960804734376, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 7300 + }, + { + "epoch": 0.07301, + "grad_norm": 1.0296175568733714, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 7301 + }, + { + "epoch": 0.07302, + "grad_norm": 0.877285101418735, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 7302 + }, + { + "epoch": 0.07303, + "grad_norm": 0.8404046231538534, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 7303 + }, + { + "epoch": 0.07304, + "grad_norm": 0.7275922913634015, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 7304 + }, + { + "epoch": 0.07305, + "grad_norm": 0.7396081307701995, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 7305 + }, + { + "epoch": 0.07306, + "grad_norm": 0.8600590405301811, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 7306 + }, + { + "epoch": 0.07307, + "grad_norm": 1.2179767907272885, + "learning_rate": 0.003, + "loss": 4.078, + "step": 7307 + }, + { + "epoch": 0.07308, + "grad_norm": 1.2171863102376617, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7308 + }, + { + "epoch": 0.07309, + "grad_norm": 0.7498319891103764, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 7309 + }, + { + "epoch": 0.0731, + "grad_norm": 0.7626246621724885, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 7310 + }, + { + "epoch": 0.07311, + "grad_norm": 0.9806079646889024, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 7311 + }, + { + "epoch": 0.07312, + "grad_norm": 1.2649393161573326, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 7312 + }, + { + "epoch": 0.07313, + "grad_norm": 1.0944111078253658, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 7313 + }, + { + "epoch": 0.07314, + "grad_norm": 1.0625503664966063, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 7314 + }, + { + "epoch": 0.07315, + "grad_norm": 0.8920266443424979, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 7315 + }, + { + "epoch": 0.07316, + "grad_norm": 0.9411350308589397, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 7316 + }, + { + "epoch": 0.07317, + "grad_norm": 1.0177785717592334, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 7317 + }, + { + "epoch": 0.07318, + "grad_norm": 1.0697061033409336, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 7318 + }, + { + "epoch": 0.07319, + "grad_norm": 1.1560468412807143, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 7319 + }, + { + "epoch": 0.0732, + "grad_norm": 0.9848533350053253, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 7320 + }, + { + "epoch": 0.07321, + "grad_norm": 1.406641344651784, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 7321 + }, + { + "epoch": 0.07322, + "grad_norm": 0.8388161556157484, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 7322 + }, + { + "epoch": 0.07323, + "grad_norm": 0.8844904618419197, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 7323 + }, + { + "epoch": 0.07324, + "grad_norm": 0.9189035571192447, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 7324 + }, + { + "epoch": 0.07325, + "grad_norm": 1.1269111817150788, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 7325 + }, + { + "epoch": 0.07326, + "grad_norm": 1.3034147784515224, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 7326 + }, + { + "epoch": 0.07327, + "grad_norm": 0.9635421663578897, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 7327 + }, + { + "epoch": 0.07328, + "grad_norm": 1.0710552918832759, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 7328 + }, + { + "epoch": 0.07329, + "grad_norm": 0.9097624953128945, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 7329 + }, + { + "epoch": 0.0733, + "grad_norm": 1.0784018937777076, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 7330 + }, + { + "epoch": 0.07331, + "grad_norm": 1.0908726771554587, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 7331 + }, + { + "epoch": 0.07332, + "grad_norm": 0.9641065596068242, + "learning_rate": 0.003, + "loss": 4.062, + "step": 7332 + }, + { + "epoch": 0.07333, + "grad_norm": 1.056358933350289, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 7333 + }, + { + "epoch": 0.07334, + "grad_norm": 1.0825403734635632, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 7334 + }, + { + "epoch": 0.07335, + "grad_norm": 0.996440123195999, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 7335 + }, + { + "epoch": 0.07336, + "grad_norm": 1.194382823583778, + "learning_rate": 0.003, + "loss": 4.075, + "step": 7336 + }, + { + "epoch": 0.07337, + "grad_norm": 1.0010816804287952, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 7337 + }, + { + "epoch": 0.07338, + "grad_norm": 1.1654307114738909, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 7338 + }, + { + "epoch": 0.07339, + "grad_norm": 1.1220075968794316, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 7339 + }, + { + "epoch": 0.0734, + "grad_norm": 1.131259039339433, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 7340 + }, + { + "epoch": 0.07341, + "grad_norm": 1.175969663125831, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 7341 + }, + { + "epoch": 0.07342, + "grad_norm": 0.9696219162610268, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 7342 + }, + { + "epoch": 0.07343, + "grad_norm": 0.9225181238457246, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7343 + }, + { + "epoch": 0.07344, + "grad_norm": 0.9472118247787676, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 7344 + }, + { + "epoch": 0.07345, + "grad_norm": 1.0731847840368303, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 7345 + }, + { + "epoch": 0.07346, + "grad_norm": 1.1511686026889996, + "learning_rate": 0.003, + "loss": 4.023, + "step": 7346 + }, + { + "epoch": 0.07347, + "grad_norm": 0.9402065576595197, + "learning_rate": 0.003, + "loss": 4.042, + "step": 7347 + }, + { + "epoch": 0.07348, + "grad_norm": 0.9633307339305025, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 7348 + }, + { + "epoch": 0.07349, + "grad_norm": 1.0646374377407537, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 7349 + }, + { + "epoch": 0.0735, + "grad_norm": 1.2040620732152902, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 7350 + }, + { + "epoch": 0.07351, + "grad_norm": 0.904348905876666, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 7351 + }, + { + "epoch": 0.07352, + "grad_norm": 1.1063199597987452, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 7352 + }, + { + "epoch": 0.07353, + "grad_norm": 1.027927831165055, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 7353 + }, + { + "epoch": 0.07354, + "grad_norm": 0.9388874030805077, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 7354 + }, + { + "epoch": 0.07355, + "grad_norm": 0.9846783510750604, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 7355 + }, + { + "epoch": 0.07356, + "grad_norm": 0.9691856508427281, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 7356 + }, + { + "epoch": 0.07357, + "grad_norm": 1.0360248831901298, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 7357 + }, + { + "epoch": 0.07358, + "grad_norm": 0.9624035012433214, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 7358 + }, + { + "epoch": 0.07359, + "grad_norm": 1.2494789612717871, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 7359 + }, + { + "epoch": 0.0736, + "grad_norm": 1.1524537864722764, + "learning_rate": 0.003, + "loss": 4.071, + "step": 7360 + }, + { + "epoch": 0.07361, + "grad_norm": 1.0340592810304519, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 7361 + }, + { + "epoch": 0.07362, + "grad_norm": 0.8881109632693943, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 7362 + }, + { + "epoch": 0.07363, + "grad_norm": 1.0007719234565362, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 7363 + }, + { + "epoch": 0.07364, + "grad_norm": 1.1774673252784194, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 7364 + }, + { + "epoch": 0.07365, + "grad_norm": 1.0952657927274845, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 7365 + }, + { + "epoch": 0.07366, + "grad_norm": 1.1457353470828615, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 7366 + }, + { + "epoch": 0.07367, + "grad_norm": 1.068208594063675, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 7367 + }, + { + "epoch": 0.07368, + "grad_norm": 1.1209873829364174, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 7368 + }, + { + "epoch": 0.07369, + "grad_norm": 0.9611042252278179, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 7369 + }, + { + "epoch": 0.0737, + "grad_norm": 1.1681489840918342, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 7370 + }, + { + "epoch": 0.07371, + "grad_norm": 1.1438687370427696, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 7371 + }, + { + "epoch": 0.07372, + "grad_norm": 1.1163570815698207, + "learning_rate": 0.003, + "loss": 4.062, + "step": 7372 + }, + { + "epoch": 0.07373, + "grad_norm": 0.994703138600529, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 7373 + }, + { + "epoch": 0.07374, + "grad_norm": 1.0487042099617259, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 7374 + }, + { + "epoch": 0.07375, + "grad_norm": 0.9349117742647232, + "learning_rate": 0.003, + "loss": 4.054, + "step": 7375 + }, + { + "epoch": 0.07376, + "grad_norm": 0.8520163703396579, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 7376 + }, + { + "epoch": 0.07377, + "grad_norm": 1.0456006728388636, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 7377 + }, + { + "epoch": 0.07378, + "grad_norm": 1.1182333493411414, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 7378 + }, + { + "epoch": 0.07379, + "grad_norm": 1.055202759953318, + "learning_rate": 0.003, + "loss": 4.036, + "step": 7379 + }, + { + "epoch": 0.0738, + "grad_norm": 0.9844943498588187, + "learning_rate": 0.003, + "loss": 4.023, + "step": 7380 + }, + { + "epoch": 0.07381, + "grad_norm": 1.0556758756646913, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 7381 + }, + { + "epoch": 0.07382, + "grad_norm": 0.939993493011405, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 7382 + }, + { + "epoch": 0.07383, + "grad_norm": 0.8597491485943202, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 7383 + }, + { + "epoch": 0.07384, + "grad_norm": 0.921647665125338, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 7384 + }, + { + "epoch": 0.07385, + "grad_norm": 1.2812955999677622, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 7385 + }, + { + "epoch": 0.07386, + "grad_norm": 0.993685614599893, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 7386 + }, + { + "epoch": 0.07387, + "grad_norm": 1.027709481484533, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 7387 + }, + { + "epoch": 0.07388, + "grad_norm": 0.9013082495547424, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 7388 + }, + { + "epoch": 0.07389, + "grad_norm": 0.9109951838256976, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 7389 + }, + { + "epoch": 0.0739, + "grad_norm": 1.2343749977944833, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 7390 + }, + { + "epoch": 0.07391, + "grad_norm": 0.9658322691066191, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 7391 + }, + { + "epoch": 0.07392, + "grad_norm": 1.1277581857498538, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 7392 + }, + { + "epoch": 0.07393, + "grad_norm": 1.0979017569203031, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 7393 + }, + { + "epoch": 0.07394, + "grad_norm": 1.1170278163443845, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 7394 + }, + { + "epoch": 0.07395, + "grad_norm": 1.0993990391688724, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 7395 + }, + { + "epoch": 0.07396, + "grad_norm": 1.0848153750483966, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 7396 + }, + { + "epoch": 0.07397, + "grad_norm": 1.1759764800565051, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 7397 + }, + { + "epoch": 0.07398, + "grad_norm": 0.9099184176925739, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 7398 + }, + { + "epoch": 0.07399, + "grad_norm": 0.9851023987142208, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 7399 + }, + { + "epoch": 0.074, + "grad_norm": 1.216378322428975, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 7400 + }, + { + "epoch": 0.07401, + "grad_norm": 0.880119789550399, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 7401 + }, + { + "epoch": 0.07402, + "grad_norm": 0.9513679960235595, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 7402 + }, + { + "epoch": 0.07403, + "grad_norm": 1.1554462432688264, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 7403 + }, + { + "epoch": 0.07404, + "grad_norm": 1.1015044417016977, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 7404 + }, + { + "epoch": 0.07405, + "grad_norm": 1.3070390701211214, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 7405 + }, + { + "epoch": 0.07406, + "grad_norm": 0.9272149699930327, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 7406 + }, + { + "epoch": 0.07407, + "grad_norm": 1.0566499622445449, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 7407 + }, + { + "epoch": 0.07408, + "grad_norm": 0.9492068898970054, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 7408 + }, + { + "epoch": 0.07409, + "grad_norm": 0.9649105766310203, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 7409 + }, + { + "epoch": 0.0741, + "grad_norm": 1.1043798243043246, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 7410 + }, + { + "epoch": 0.07411, + "grad_norm": 1.1607113803566775, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 7411 + }, + { + "epoch": 0.07412, + "grad_norm": 0.9327351376075126, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 7412 + }, + { + "epoch": 0.07413, + "grad_norm": 1.0657239310703015, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 7413 + }, + { + "epoch": 0.07414, + "grad_norm": 1.0370445327105178, + "learning_rate": 0.003, + "loss": 4.049, + "step": 7414 + }, + { + "epoch": 0.07415, + "grad_norm": 1.1291501330975042, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 7415 + }, + { + "epoch": 0.07416, + "grad_norm": 1.0563748188952578, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 7416 + }, + { + "epoch": 0.07417, + "grad_norm": 0.9426313292975299, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 7417 + }, + { + "epoch": 0.07418, + "grad_norm": 1.138322949259601, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 7418 + }, + { + "epoch": 0.07419, + "grad_norm": 1.213845231951027, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 7419 + }, + { + "epoch": 0.0742, + "grad_norm": 0.9160208863216656, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 7420 + }, + { + "epoch": 0.07421, + "grad_norm": 0.9459537396238892, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 7421 + }, + { + "epoch": 0.07422, + "grad_norm": 0.9896087479545597, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 7422 + }, + { + "epoch": 0.07423, + "grad_norm": 1.029001637107005, + "learning_rate": 0.003, + "loss": 4.032, + "step": 7423 + }, + { + "epoch": 0.07424, + "grad_norm": 0.966524850109166, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 7424 + }, + { + "epoch": 0.07425, + "grad_norm": 1.0858257368928335, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 7425 + }, + { + "epoch": 0.07426, + "grad_norm": 0.9810888238304923, + "learning_rate": 0.003, + "loss": 4.023, + "step": 7426 + }, + { + "epoch": 0.07427, + "grad_norm": 1.1559445528412031, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 7427 + }, + { + "epoch": 0.07428, + "grad_norm": 1.1236073459063627, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 7428 + }, + { + "epoch": 0.07429, + "grad_norm": 1.096676807770383, + "learning_rate": 0.003, + "loss": 4.055, + "step": 7429 + }, + { + "epoch": 0.0743, + "grad_norm": 1.0713521198381804, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 7430 + }, + { + "epoch": 0.07431, + "grad_norm": 1.0003185699822545, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 7431 + }, + { + "epoch": 0.07432, + "grad_norm": 1.242346832405972, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 7432 + }, + { + "epoch": 0.07433, + "grad_norm": 1.0630507457181497, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 7433 + }, + { + "epoch": 0.07434, + "grad_norm": 1.1654066987813088, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 7434 + }, + { + "epoch": 0.07435, + "grad_norm": 1.2683692736396797, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 7435 + }, + { + "epoch": 0.07436, + "grad_norm": 0.945218194969709, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 7436 + }, + { + "epoch": 0.07437, + "grad_norm": 1.097019618654518, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 7437 + }, + { + "epoch": 0.07438, + "grad_norm": 1.2074577366475394, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 7438 + }, + { + "epoch": 0.07439, + "grad_norm": 0.9431501487823388, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 7439 + }, + { + "epoch": 0.0744, + "grad_norm": 1.1411961430256892, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 7440 + }, + { + "epoch": 0.07441, + "grad_norm": 0.945382643731931, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 7441 + }, + { + "epoch": 0.07442, + "grad_norm": 0.9584211239670143, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 7442 + }, + { + "epoch": 0.07443, + "grad_norm": 0.8936679140385173, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 7443 + }, + { + "epoch": 0.07444, + "grad_norm": 0.8616182361179869, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 7444 + }, + { + "epoch": 0.07445, + "grad_norm": 1.112391031977807, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 7445 + }, + { + "epoch": 0.07446, + "grad_norm": 1.260904358901806, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 7446 + }, + { + "epoch": 0.07447, + "grad_norm": 0.9143344109603971, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 7447 + }, + { + "epoch": 0.07448, + "grad_norm": 0.9777877903133408, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 7448 + }, + { + "epoch": 0.07449, + "grad_norm": 1.0092330290166718, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 7449 + }, + { + "epoch": 0.0745, + "grad_norm": 1.116824586523927, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 7450 + }, + { + "epoch": 0.07451, + "grad_norm": 1.0158873706838591, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7451 + }, + { + "epoch": 0.07452, + "grad_norm": 1.0683848996605232, + "learning_rate": 0.003, + "loss": 4.023, + "step": 7452 + }, + { + "epoch": 0.07453, + "grad_norm": 1.0005773823548063, + "learning_rate": 0.003, + "loss": 4.054, + "step": 7453 + }, + { + "epoch": 0.07454, + "grad_norm": 1.1094763576693252, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 7454 + }, + { + "epoch": 0.07455, + "grad_norm": 1.0977415415623035, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 7455 + }, + { + "epoch": 0.07456, + "grad_norm": 1.2070613083727133, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 7456 + }, + { + "epoch": 0.07457, + "grad_norm": 1.033669789154854, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 7457 + }, + { + "epoch": 0.07458, + "grad_norm": 0.9974734919174871, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 7458 + }, + { + "epoch": 0.07459, + "grad_norm": 1.1596224019873076, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 7459 + }, + { + "epoch": 0.0746, + "grad_norm": 1.0558735224515694, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 7460 + }, + { + "epoch": 0.07461, + "grad_norm": 1.1013658395166883, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 7461 + }, + { + "epoch": 0.07462, + "grad_norm": 1.099153298296693, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 7462 + }, + { + "epoch": 0.07463, + "grad_norm": 0.9155991610369465, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 7463 + }, + { + "epoch": 0.07464, + "grad_norm": 0.9379208043653106, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 7464 + }, + { + "epoch": 0.07465, + "grad_norm": 1.0372587578155243, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 7465 + }, + { + "epoch": 0.07466, + "grad_norm": 1.3132084128992856, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 7466 + }, + { + "epoch": 0.07467, + "grad_norm": 0.933995203349761, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 7467 + }, + { + "epoch": 0.07468, + "grad_norm": 1.0967227098443246, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 7468 + }, + { + "epoch": 0.07469, + "grad_norm": 0.8994513242913562, + "learning_rate": 0.003, + "loss": 4.018, + "step": 7469 + }, + { + "epoch": 0.0747, + "grad_norm": 0.8582057247678115, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 7470 + }, + { + "epoch": 0.07471, + "grad_norm": 0.9750254675432543, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 7471 + }, + { + "epoch": 0.07472, + "grad_norm": 1.0758291745603603, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7472 + }, + { + "epoch": 0.07473, + "grad_norm": 1.2870897431186328, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 7473 + }, + { + "epoch": 0.07474, + "grad_norm": 0.8579070231808928, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 7474 + }, + { + "epoch": 0.07475, + "grad_norm": 0.9738156550036234, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 7475 + }, + { + "epoch": 0.07476, + "grad_norm": 1.045108915912123, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 7476 + }, + { + "epoch": 0.07477, + "grad_norm": 1.094542247261385, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 7477 + }, + { + "epoch": 0.07478, + "grad_norm": 0.8629572990188785, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 7478 + }, + { + "epoch": 0.07479, + "grad_norm": 0.9292854361264181, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 7479 + }, + { + "epoch": 0.0748, + "grad_norm": 1.0705526760945827, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 7480 + }, + { + "epoch": 0.07481, + "grad_norm": 1.0359796271584742, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 7481 + }, + { + "epoch": 0.07482, + "grad_norm": 1.232770848270821, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 7482 + }, + { + "epoch": 0.07483, + "grad_norm": 1.0817607398369629, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 7483 + }, + { + "epoch": 0.07484, + "grad_norm": 1.2858180984611243, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 7484 + }, + { + "epoch": 0.07485, + "grad_norm": 0.9867732155173473, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 7485 + }, + { + "epoch": 0.07486, + "grad_norm": 1.0647719783345513, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 7486 + }, + { + "epoch": 0.07487, + "grad_norm": 0.9563519763714589, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7487 + }, + { + "epoch": 0.07488, + "grad_norm": 1.081168855348119, + "learning_rate": 0.003, + "loss": 4.058, + "step": 7488 + }, + { + "epoch": 0.07489, + "grad_norm": 0.9057881307635063, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 7489 + }, + { + "epoch": 0.0749, + "grad_norm": 0.941257228213405, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 7490 + }, + { + "epoch": 0.07491, + "grad_norm": 1.023677883109361, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 7491 + }, + { + "epoch": 0.07492, + "grad_norm": 1.0985843828113449, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 7492 + }, + { + "epoch": 0.07493, + "grad_norm": 1.1291894542659202, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 7493 + }, + { + "epoch": 0.07494, + "grad_norm": 1.046782915344204, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 7494 + }, + { + "epoch": 0.07495, + "grad_norm": 1.2154732983788594, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 7495 + }, + { + "epoch": 0.07496, + "grad_norm": 1.1000284508004632, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 7496 + }, + { + "epoch": 0.07497, + "grad_norm": 1.1040574207112384, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 7497 + }, + { + "epoch": 0.07498, + "grad_norm": 0.8948022242164695, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 7498 + }, + { + "epoch": 0.07499, + "grad_norm": 1.0258546162106021, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 7499 + }, + { + "epoch": 0.075, + "grad_norm": 1.1158118517557325, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 7500 + }, + { + "epoch": 0.07501, + "grad_norm": 1.0543780189380676, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 7501 + }, + { + "epoch": 0.07502, + "grad_norm": 1.1398746708324812, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 7502 + }, + { + "epoch": 0.07503, + "grad_norm": 1.17233757471888, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 7503 + }, + { + "epoch": 0.07504, + "grad_norm": 1.1747348217017677, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 7504 + }, + { + "epoch": 0.07505, + "grad_norm": 0.8995993583616543, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 7505 + }, + { + "epoch": 0.07506, + "grad_norm": 0.9337832043650974, + "learning_rate": 0.003, + "loss": 4.038, + "step": 7506 + }, + { + "epoch": 0.07507, + "grad_norm": 0.9695999584916757, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 7507 + }, + { + "epoch": 0.07508, + "grad_norm": 1.1638077160586862, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 7508 + }, + { + "epoch": 0.07509, + "grad_norm": 0.9874287876840504, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 7509 + }, + { + "epoch": 0.0751, + "grad_norm": 0.9776311060010476, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 7510 + }, + { + "epoch": 0.07511, + "grad_norm": 1.1033486233571166, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 7511 + }, + { + "epoch": 0.07512, + "grad_norm": 0.8907632335623773, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 7512 + }, + { + "epoch": 0.07513, + "grad_norm": 0.9988954487137791, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 7513 + }, + { + "epoch": 0.07514, + "grad_norm": 1.194395962502461, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 7514 + }, + { + "epoch": 0.07515, + "grad_norm": 1.156162880522299, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 7515 + }, + { + "epoch": 0.07516, + "grad_norm": 1.2727170425623708, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 7516 + }, + { + "epoch": 0.07517, + "grad_norm": 1.1565889833590228, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 7517 + }, + { + "epoch": 0.07518, + "grad_norm": 0.8708468314319205, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 7518 + }, + { + "epoch": 0.07519, + "grad_norm": 0.9953433252661223, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 7519 + }, + { + "epoch": 0.0752, + "grad_norm": 1.0940459643263882, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 7520 + }, + { + "epoch": 0.07521, + "grad_norm": 0.928716112619678, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 7521 + }, + { + "epoch": 0.07522, + "grad_norm": 0.9818555888838302, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 7522 + }, + { + "epoch": 0.07523, + "grad_norm": 1.2068474109803409, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 7523 + }, + { + "epoch": 0.07524, + "grad_norm": 1.01597504610466, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 7524 + }, + { + "epoch": 0.07525, + "grad_norm": 1.2715734649300354, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 7525 + }, + { + "epoch": 0.07526, + "grad_norm": 0.7870088972562932, + "learning_rate": 0.003, + "loss": 4.064, + "step": 7526 + }, + { + "epoch": 0.07527, + "grad_norm": 0.9106993971561816, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 7527 + }, + { + "epoch": 0.07528, + "grad_norm": 0.8636142458933644, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 7528 + }, + { + "epoch": 0.07529, + "grad_norm": 0.9616268692995935, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 7529 + }, + { + "epoch": 0.0753, + "grad_norm": 1.114538018638255, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 7530 + }, + { + "epoch": 0.07531, + "grad_norm": 0.9879931527488544, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 7531 + }, + { + "epoch": 0.07532, + "grad_norm": 1.0655313216663054, + "learning_rate": 0.003, + "loss": 4.056, + "step": 7532 + }, + { + "epoch": 0.07533, + "grad_norm": 0.8780149908265403, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 7533 + }, + { + "epoch": 0.07534, + "grad_norm": 0.8784040460131582, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 7534 + }, + { + "epoch": 0.07535, + "grad_norm": 0.8518380464505314, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7535 + }, + { + "epoch": 0.07536, + "grad_norm": 0.9098902162790059, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 7536 + }, + { + "epoch": 0.07537, + "grad_norm": 1.143562497559037, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 7537 + }, + { + "epoch": 0.07538, + "grad_norm": 1.2254729496457664, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 7538 + }, + { + "epoch": 0.07539, + "grad_norm": 1.043842367086055, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 7539 + }, + { + "epoch": 0.0754, + "grad_norm": 1.21725440776156, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 7540 + }, + { + "epoch": 0.07541, + "grad_norm": 0.9295788359152493, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 7541 + }, + { + "epoch": 0.07542, + "grad_norm": 1.130473398219038, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 7542 + }, + { + "epoch": 0.07543, + "grad_norm": 0.9937582722105303, + "learning_rate": 0.003, + "loss": 4.065, + "step": 7543 + }, + { + "epoch": 0.07544, + "grad_norm": 1.2560834000797938, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 7544 + }, + { + "epoch": 0.07545, + "grad_norm": 1.0050728570569878, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 7545 + }, + { + "epoch": 0.07546, + "grad_norm": 1.0720793211640665, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 7546 + }, + { + "epoch": 0.07547, + "grad_norm": 0.9113278300880503, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 7547 + }, + { + "epoch": 0.07548, + "grad_norm": 0.7765738432773017, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 7548 + }, + { + "epoch": 0.07549, + "grad_norm": 0.912954245340519, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 7549 + }, + { + "epoch": 0.0755, + "grad_norm": 1.1357200728898464, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 7550 + }, + { + "epoch": 0.07551, + "grad_norm": 1.1645892480065512, + "learning_rate": 0.003, + "loss": 4.06, + "step": 7551 + }, + { + "epoch": 0.07552, + "grad_norm": 1.0759538967371787, + "learning_rate": 0.003, + "loss": 4.061, + "step": 7552 + }, + { + "epoch": 0.07553, + "grad_norm": 1.113298651168989, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 7553 + }, + { + "epoch": 0.07554, + "grad_norm": 1.3129530629673276, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 7554 + }, + { + "epoch": 0.07555, + "grad_norm": 1.0195391886187795, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 7555 + }, + { + "epoch": 0.07556, + "grad_norm": 1.1261424419589618, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 7556 + }, + { + "epoch": 0.07557, + "grad_norm": 0.9078457564330181, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 7557 + }, + { + "epoch": 0.07558, + "grad_norm": 1.1105037206557775, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 7558 + }, + { + "epoch": 0.07559, + "grad_norm": 1.084878893896586, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 7559 + }, + { + "epoch": 0.0756, + "grad_norm": 0.9422492493376851, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 7560 + }, + { + "epoch": 0.07561, + "grad_norm": 0.9472473971962481, + "learning_rate": 0.003, + "loss": 4.038, + "step": 7561 + }, + { + "epoch": 0.07562, + "grad_norm": 1.0141863361647219, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 7562 + }, + { + "epoch": 0.07563, + "grad_norm": 1.209943999383626, + "learning_rate": 0.003, + "loss": 4.052, + "step": 7563 + }, + { + "epoch": 0.07564, + "grad_norm": 1.040474950950908, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 7564 + }, + { + "epoch": 0.07565, + "grad_norm": 1.041564684801342, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 7565 + }, + { + "epoch": 0.07566, + "grad_norm": 0.9211223111374635, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 7566 + }, + { + "epoch": 0.07567, + "grad_norm": 0.9557313493057552, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 7567 + }, + { + "epoch": 0.07568, + "grad_norm": 1.0157216311700132, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 7568 + }, + { + "epoch": 0.07569, + "grad_norm": 1.0891520113381792, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 7569 + }, + { + "epoch": 0.0757, + "grad_norm": 1.035599436574519, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 7570 + }, + { + "epoch": 0.07571, + "grad_norm": 1.1416705208329894, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 7571 + }, + { + "epoch": 0.07572, + "grad_norm": 1.1983465038651324, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 7572 + }, + { + "epoch": 0.07573, + "grad_norm": 0.8376450895449468, + "learning_rate": 0.003, + "loss": 4.043, + "step": 7573 + }, + { + "epoch": 0.07574, + "grad_norm": 0.7440289348204431, + "learning_rate": 0.003, + "loss": 4.057, + "step": 7574 + }, + { + "epoch": 0.07575, + "grad_norm": 0.6603077120225412, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 7575 + }, + { + "epoch": 0.07576, + "grad_norm": 0.7920785949672058, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 7576 + }, + { + "epoch": 0.07577, + "grad_norm": 1.0268729068383955, + "learning_rate": 0.003, + "loss": 4.066, + "step": 7577 + }, + { + "epoch": 0.07578, + "grad_norm": 1.1368625436467688, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 7578 + }, + { + "epoch": 0.07579, + "grad_norm": 1.008982576200628, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 7579 + }, + { + "epoch": 0.0758, + "grad_norm": 1.3316307903450357, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 7580 + }, + { + "epoch": 0.07581, + "grad_norm": 0.9546023266380341, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 7581 + }, + { + "epoch": 0.07582, + "grad_norm": 1.028351315894607, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 7582 + }, + { + "epoch": 0.07583, + "grad_norm": 1.1084447205745265, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 7583 + }, + { + "epoch": 0.07584, + "grad_norm": 1.083029995624086, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 7584 + }, + { + "epoch": 0.07585, + "grad_norm": 1.0956918067982808, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 7585 + }, + { + "epoch": 0.07586, + "grad_norm": 1.0704482847092711, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 7586 + }, + { + "epoch": 0.07587, + "grad_norm": 1.0248155370988479, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 7587 + }, + { + "epoch": 0.07588, + "grad_norm": 0.9469007252567827, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 7588 + }, + { + "epoch": 0.07589, + "grad_norm": 0.9083602594004783, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 7589 + }, + { + "epoch": 0.0759, + "grad_norm": 1.0888081212751122, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 7590 + }, + { + "epoch": 0.07591, + "grad_norm": 1.1286869611060864, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 7591 + }, + { + "epoch": 0.07592, + "grad_norm": 1.1994073883546317, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 7592 + }, + { + "epoch": 0.07593, + "grad_norm": 0.930854156102424, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 7593 + }, + { + "epoch": 0.07594, + "grad_norm": 0.9949944707269984, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 7594 + }, + { + "epoch": 0.07595, + "grad_norm": 1.0566054746262994, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7595 + }, + { + "epoch": 0.07596, + "grad_norm": 0.9893650668367953, + "learning_rate": 0.003, + "loss": 4.022, + "step": 7596 + }, + { + "epoch": 0.07597, + "grad_norm": 1.1346833349811905, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 7597 + }, + { + "epoch": 0.07598, + "grad_norm": 1.0292720315909252, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 7598 + }, + { + "epoch": 0.07599, + "grad_norm": 1.152689224088955, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 7599 + }, + { + "epoch": 0.076, + "grad_norm": 1.0383151228598986, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 7600 + }, + { + "epoch": 0.07601, + "grad_norm": 1.047117789348393, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 7601 + }, + { + "epoch": 0.07602, + "grad_norm": 1.106052696242171, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 7602 + }, + { + "epoch": 0.07603, + "grad_norm": 1.0073145436596014, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 7603 + }, + { + "epoch": 0.07604, + "grad_norm": 1.0664230896712803, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 7604 + }, + { + "epoch": 0.07605, + "grad_norm": 0.9753888567376868, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 7605 + }, + { + "epoch": 0.07606, + "grad_norm": 1.2030307927813533, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 7606 + }, + { + "epoch": 0.07607, + "grad_norm": 0.9086575357118534, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 7607 + }, + { + "epoch": 0.07608, + "grad_norm": 0.9273700226274223, + "learning_rate": 0.003, + "loss": 4.074, + "step": 7608 + }, + { + "epoch": 0.07609, + "grad_norm": 1.003689493135648, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 7609 + }, + { + "epoch": 0.0761, + "grad_norm": 1.1377269625068565, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 7610 + }, + { + "epoch": 0.07611, + "grad_norm": 0.9384559805000248, + "learning_rate": 0.003, + "loss": 4.07, + "step": 7611 + }, + { + "epoch": 0.07612, + "grad_norm": 1.1271298131099274, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 7612 + }, + { + "epoch": 0.07613, + "grad_norm": 1.1197631113115567, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 7613 + }, + { + "epoch": 0.07614, + "grad_norm": 1.0895475603662517, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 7614 + }, + { + "epoch": 0.07615, + "grad_norm": 1.1379640727683638, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 7615 + }, + { + "epoch": 0.07616, + "grad_norm": 1.0509591762451105, + "learning_rate": 0.003, + "loss": 4.049, + "step": 7616 + }, + { + "epoch": 0.07617, + "grad_norm": 1.0514288577627249, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 7617 + }, + { + "epoch": 0.07618, + "grad_norm": 0.9463880932847941, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 7618 + }, + { + "epoch": 0.07619, + "grad_norm": 1.0151178915603503, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 7619 + }, + { + "epoch": 0.0762, + "grad_norm": 1.2638113293346065, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 7620 + }, + { + "epoch": 0.07621, + "grad_norm": 0.8743906611118283, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 7621 + }, + { + "epoch": 0.07622, + "grad_norm": 0.9376371308176443, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 7622 + }, + { + "epoch": 0.07623, + "grad_norm": 1.2162047119192658, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 7623 + }, + { + "epoch": 0.07624, + "grad_norm": 0.8081909472078085, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 7624 + }, + { + "epoch": 0.07625, + "grad_norm": 0.9250877782038973, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 7625 + }, + { + "epoch": 0.07626, + "grad_norm": 1.105895370815446, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 7626 + }, + { + "epoch": 0.07627, + "grad_norm": 1.0603656758381923, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 7627 + }, + { + "epoch": 0.07628, + "grad_norm": 1.2118848294290039, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 7628 + }, + { + "epoch": 0.07629, + "grad_norm": 0.9491546824024952, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 7629 + }, + { + "epoch": 0.0763, + "grad_norm": 1.1470367440434606, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 7630 + }, + { + "epoch": 0.07631, + "grad_norm": 0.9750737788207199, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 7631 + }, + { + "epoch": 0.07632, + "grad_norm": 0.9223121244351081, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 7632 + }, + { + "epoch": 0.07633, + "grad_norm": 0.9447939771060309, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 7633 + }, + { + "epoch": 0.07634, + "grad_norm": 1.0656262121811246, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 7634 + }, + { + "epoch": 0.07635, + "grad_norm": 1.1293493421594012, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 7635 + }, + { + "epoch": 0.07636, + "grad_norm": 1.1936557881689314, + "learning_rate": 0.003, + "loss": 4.028, + "step": 7636 + }, + { + "epoch": 0.07637, + "grad_norm": 0.9929307055596185, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 7637 + }, + { + "epoch": 0.07638, + "grad_norm": 1.0228636969536977, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 7638 + }, + { + "epoch": 0.07639, + "grad_norm": 0.9823441975383835, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 7639 + }, + { + "epoch": 0.0764, + "grad_norm": 1.1338997019966783, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 7640 + }, + { + "epoch": 0.07641, + "grad_norm": 1.0093913734568616, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 7641 + }, + { + "epoch": 0.07642, + "grad_norm": 1.125461302352058, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 7642 + }, + { + "epoch": 0.07643, + "grad_norm": 1.0514677154433425, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 7643 + }, + { + "epoch": 0.07644, + "grad_norm": 1.0553593637554315, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 7644 + }, + { + "epoch": 0.07645, + "grad_norm": 0.9805977399864129, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 7645 + }, + { + "epoch": 0.07646, + "grad_norm": 1.0299202945868386, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 7646 + }, + { + "epoch": 0.07647, + "grad_norm": 1.3176337185067677, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 7647 + }, + { + "epoch": 0.07648, + "grad_norm": 1.1005654564479557, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 7648 + }, + { + "epoch": 0.07649, + "grad_norm": 1.0905324219896768, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 7649 + }, + { + "epoch": 0.0765, + "grad_norm": 0.8802249156939694, + "learning_rate": 0.003, + "loss": 4.031, + "step": 7650 + }, + { + "epoch": 0.07651, + "grad_norm": 0.9072009257031926, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 7651 + }, + { + "epoch": 0.07652, + "grad_norm": 0.9346164551384647, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 7652 + }, + { + "epoch": 0.07653, + "grad_norm": 0.980298089231196, + "learning_rate": 0.003, + "loss": 4.019, + "step": 7653 + }, + { + "epoch": 0.07654, + "grad_norm": 1.0968929008117443, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 7654 + }, + { + "epoch": 0.07655, + "grad_norm": 0.9601657405610515, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 7655 + }, + { + "epoch": 0.07656, + "grad_norm": 1.258852527855847, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 7656 + }, + { + "epoch": 0.07657, + "grad_norm": 0.8924445280055346, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 7657 + }, + { + "epoch": 0.07658, + "grad_norm": 1.016577242205408, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 7658 + }, + { + "epoch": 0.07659, + "grad_norm": 1.1676547370839572, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 7659 + }, + { + "epoch": 0.0766, + "grad_norm": 1.0842253182708932, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 7660 + }, + { + "epoch": 0.07661, + "grad_norm": 1.233560763252066, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 7661 + }, + { + "epoch": 0.07662, + "grad_norm": 1.0033408895801803, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 7662 + }, + { + "epoch": 0.07663, + "grad_norm": 1.2643790027991915, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 7663 + }, + { + "epoch": 0.07664, + "grad_norm": 1.023325848782995, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 7664 + }, + { + "epoch": 0.07665, + "grad_norm": 1.0799510581720673, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 7665 + }, + { + "epoch": 0.07666, + "grad_norm": 1.108909756809187, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 7666 + }, + { + "epoch": 0.07667, + "grad_norm": 0.8955114573010688, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 7667 + }, + { + "epoch": 0.07668, + "grad_norm": 0.8923297593100042, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 7668 + }, + { + "epoch": 0.07669, + "grad_norm": 1.0249046055145887, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 7669 + }, + { + "epoch": 0.0767, + "grad_norm": 1.2053116597026634, + "learning_rate": 0.003, + "loss": 4.033, + "step": 7670 + }, + { + "epoch": 0.07671, + "grad_norm": 1.0082941639633163, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 7671 + }, + { + "epoch": 0.07672, + "grad_norm": 1.2734345360500605, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 7672 + }, + { + "epoch": 0.07673, + "grad_norm": 0.9281784475022431, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 7673 + }, + { + "epoch": 0.07674, + "grad_norm": 1.0050293522340423, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 7674 + }, + { + "epoch": 0.07675, + "grad_norm": 1.1578682951290742, + "learning_rate": 0.003, + "loss": 4.047, + "step": 7675 + }, + { + "epoch": 0.07676, + "grad_norm": 1.0107218219879721, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 7676 + }, + { + "epoch": 0.07677, + "grad_norm": 1.122086603391903, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 7677 + }, + { + "epoch": 0.07678, + "grad_norm": 1.0817490316966227, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 7678 + }, + { + "epoch": 0.07679, + "grad_norm": 1.0060234752712105, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 7679 + }, + { + "epoch": 0.0768, + "grad_norm": 1.0542727480500718, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 7680 + }, + { + "epoch": 0.07681, + "grad_norm": 0.9906102704846765, + "learning_rate": 0.003, + "loss": 4.049, + "step": 7681 + }, + { + "epoch": 0.07682, + "grad_norm": 1.1799698353197907, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 7682 + }, + { + "epoch": 0.07683, + "grad_norm": 0.9827379813572511, + "learning_rate": 0.003, + "loss": 4.048, + "step": 7683 + }, + { + "epoch": 0.07684, + "grad_norm": 1.0627107640985607, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 7684 + }, + { + "epoch": 0.07685, + "grad_norm": 0.8775795962977772, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 7685 + }, + { + "epoch": 0.07686, + "grad_norm": 0.7639644005133086, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 7686 + }, + { + "epoch": 0.07687, + "grad_norm": 0.8605351853455829, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 7687 + }, + { + "epoch": 0.07688, + "grad_norm": 0.9162241897881261, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 7688 + }, + { + "epoch": 0.07689, + "grad_norm": 1.2441994585763008, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 7689 + }, + { + "epoch": 0.0769, + "grad_norm": 1.0783654477764117, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 7690 + }, + { + "epoch": 0.07691, + "grad_norm": 0.9813259733202012, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 7691 + }, + { + "epoch": 0.07692, + "grad_norm": 1.1132849435616365, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 7692 + }, + { + "epoch": 0.07693, + "grad_norm": 0.968186114392162, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 7693 + }, + { + "epoch": 0.07694, + "grad_norm": 1.1180368633304685, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 7694 + }, + { + "epoch": 0.07695, + "grad_norm": 1.018150504213498, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 7695 + }, + { + "epoch": 0.07696, + "grad_norm": 1.1134933391530109, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 7696 + }, + { + "epoch": 0.07697, + "grad_norm": 1.112740567601611, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 7697 + }, + { + "epoch": 0.07698, + "grad_norm": 1.035646453959349, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 7698 + }, + { + "epoch": 0.07699, + "grad_norm": 1.1776754763679083, + "learning_rate": 0.003, + "loss": 4.089, + "step": 7699 + }, + { + "epoch": 0.077, + "grad_norm": 1.0846924700885932, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 7700 + }, + { + "epoch": 0.07701, + "grad_norm": 1.1355756172933456, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 7701 + }, + { + "epoch": 0.07702, + "grad_norm": 1.2445054567640432, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 7702 + }, + { + "epoch": 0.07703, + "grad_norm": 0.778195828814493, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 7703 + }, + { + "epoch": 0.07704, + "grad_norm": 0.7695651423057185, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 7704 + }, + { + "epoch": 0.07705, + "grad_norm": 0.8208723519118769, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 7705 + }, + { + "epoch": 0.07706, + "grad_norm": 1.0591846703576349, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 7706 + }, + { + "epoch": 0.07707, + "grad_norm": 1.4331501433140448, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 7707 + }, + { + "epoch": 0.07708, + "grad_norm": 0.8586416225127098, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 7708 + }, + { + "epoch": 0.07709, + "grad_norm": 0.9896715342897399, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 7709 + }, + { + "epoch": 0.0771, + "grad_norm": 1.0247102942561306, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 7710 + }, + { + "epoch": 0.07711, + "grad_norm": 0.9573824195950065, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 7711 + }, + { + "epoch": 0.07712, + "grad_norm": 0.9223979651594529, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 7712 + }, + { + "epoch": 0.07713, + "grad_norm": 1.123004149058808, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 7713 + }, + { + "epoch": 0.07714, + "grad_norm": 1.1362346630117264, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 7714 + }, + { + "epoch": 0.07715, + "grad_norm": 0.9873091975110139, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 7715 + }, + { + "epoch": 0.07716, + "grad_norm": 1.1616502010150347, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 7716 + }, + { + "epoch": 0.07717, + "grad_norm": 1.1465651911213932, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 7717 + }, + { + "epoch": 0.07718, + "grad_norm": 1.07816656816876, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 7718 + }, + { + "epoch": 0.07719, + "grad_norm": 0.879915285523726, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 7719 + }, + { + "epoch": 0.0772, + "grad_norm": 0.8694587190603413, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 7720 + }, + { + "epoch": 0.07721, + "grad_norm": 0.9644244720161652, + "learning_rate": 0.003, + "loss": 4.029, + "step": 7721 + }, + { + "epoch": 0.07722, + "grad_norm": 1.0848954562449826, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 7722 + }, + { + "epoch": 0.07723, + "grad_norm": 1.2394893053064837, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 7723 + }, + { + "epoch": 0.07724, + "grad_norm": 0.9558592147764633, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 7724 + }, + { + "epoch": 0.07725, + "grad_norm": 1.1959360495832247, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 7725 + }, + { + "epoch": 0.07726, + "grad_norm": 0.9517967220511826, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 7726 + }, + { + "epoch": 0.07727, + "grad_norm": 1.1579653372604413, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 7727 + }, + { + "epoch": 0.07728, + "grad_norm": 1.1068848248552594, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 7728 + }, + { + "epoch": 0.07729, + "grad_norm": 0.9527663111508775, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 7729 + }, + { + "epoch": 0.0773, + "grad_norm": 0.82211244826477, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 7730 + }, + { + "epoch": 0.07731, + "grad_norm": 0.8528406417127462, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 7731 + }, + { + "epoch": 0.07732, + "grad_norm": 0.9703295806343482, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 7732 + }, + { + "epoch": 0.07733, + "grad_norm": 1.233288598154536, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 7733 + }, + { + "epoch": 0.07734, + "grad_norm": 1.1910360013875985, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 7734 + }, + { + "epoch": 0.07735, + "grad_norm": 1.1528280039800138, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 7735 + }, + { + "epoch": 0.07736, + "grad_norm": 1.163933383869731, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 7736 + }, + { + "epoch": 0.07737, + "grad_norm": 0.9130491112837049, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 7737 + }, + { + "epoch": 0.07738, + "grad_norm": 0.9070020333481247, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 7738 + }, + { + "epoch": 0.07739, + "grad_norm": 0.9956178440672993, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 7739 + }, + { + "epoch": 0.0774, + "grad_norm": 1.2016421176014782, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 7740 + }, + { + "epoch": 0.07741, + "grad_norm": 1.0114172630176164, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 7741 + }, + { + "epoch": 0.07742, + "grad_norm": 1.0767057084476068, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 7742 + }, + { + "epoch": 0.07743, + "grad_norm": 1.140758765621141, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 7743 + }, + { + "epoch": 0.07744, + "grad_norm": 0.941795514834158, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 7744 + }, + { + "epoch": 0.07745, + "grad_norm": 0.947968983480798, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 7745 + }, + { + "epoch": 0.07746, + "grad_norm": 1.1438558351135404, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 7746 + }, + { + "epoch": 0.07747, + "grad_norm": 1.1271433369048343, + "learning_rate": 0.003, + "loss": 4.055, + "step": 7747 + }, + { + "epoch": 0.07748, + "grad_norm": 1.068011422911948, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 7748 + }, + { + "epoch": 0.07749, + "grad_norm": 1.1673301793274091, + "learning_rate": 0.003, + "loss": 4.077, + "step": 7749 + }, + { + "epoch": 0.0775, + "grad_norm": 0.9871201697529424, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 7750 + }, + { + "epoch": 0.07751, + "grad_norm": 1.1865630409028847, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 7751 + }, + { + "epoch": 0.07752, + "grad_norm": 0.9135508252960873, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 7752 + }, + { + "epoch": 0.07753, + "grad_norm": 1.0491466617434224, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 7753 + }, + { + "epoch": 0.07754, + "grad_norm": 1.0597062148757896, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 7754 + }, + { + "epoch": 0.07755, + "grad_norm": 1.0522100404904275, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 7755 + }, + { + "epoch": 0.07756, + "grad_norm": 1.1298394191949892, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 7756 + }, + { + "epoch": 0.07757, + "grad_norm": 0.9473952380257887, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 7757 + }, + { + "epoch": 0.07758, + "grad_norm": 1.0741481447103787, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 7758 + }, + { + "epoch": 0.07759, + "grad_norm": 1.1033185053370784, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 7759 + }, + { + "epoch": 0.0776, + "grad_norm": 0.946764770560814, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 7760 + }, + { + "epoch": 0.07761, + "grad_norm": 1.1024391112725458, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 7761 + }, + { + "epoch": 0.07762, + "grad_norm": 1.0452461471198484, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 7762 + }, + { + "epoch": 0.07763, + "grad_norm": 1.1390912229610954, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 7763 + }, + { + "epoch": 0.07764, + "grad_norm": 1.0090027084277429, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 7764 + }, + { + "epoch": 0.07765, + "grad_norm": 1.1375674959522515, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 7765 + }, + { + "epoch": 0.07766, + "grad_norm": 1.2062883449499884, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7766 + }, + { + "epoch": 0.07767, + "grad_norm": 1.1804562993590413, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7767 + }, + { + "epoch": 0.07768, + "grad_norm": 1.0684635284018162, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 7768 + }, + { + "epoch": 0.07769, + "grad_norm": 1.1159137124746297, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 7769 + }, + { + "epoch": 0.0777, + "grad_norm": 1.0678717136438307, + "learning_rate": 0.003, + "loss": 4.065, + "step": 7770 + }, + { + "epoch": 0.07771, + "grad_norm": 0.9559701342030749, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 7771 + }, + { + "epoch": 0.07772, + "grad_norm": 1.0062287758341482, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 7772 + }, + { + "epoch": 0.07773, + "grad_norm": 1.0966855486122142, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 7773 + }, + { + "epoch": 0.07774, + "grad_norm": 0.9558412183434721, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 7774 + }, + { + "epoch": 0.07775, + "grad_norm": 1.2137613391546822, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 7775 + }, + { + "epoch": 0.07776, + "grad_norm": 0.7820218930391715, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7776 + }, + { + "epoch": 0.07777, + "grad_norm": 0.7196952870861936, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 7777 + }, + { + "epoch": 0.07778, + "grad_norm": 0.9999098617704734, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 7778 + }, + { + "epoch": 0.07779, + "grad_norm": 1.4627741712221116, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 7779 + }, + { + "epoch": 0.0778, + "grad_norm": 0.9529840352238272, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 7780 + }, + { + "epoch": 0.07781, + "grad_norm": 1.2285518753842994, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 7781 + }, + { + "epoch": 0.07782, + "grad_norm": 1.1404528733856336, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 7782 + }, + { + "epoch": 0.07783, + "grad_norm": 1.1975303734486777, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 7783 + }, + { + "epoch": 0.07784, + "grad_norm": 1.0934483916125426, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 7784 + }, + { + "epoch": 0.07785, + "grad_norm": 1.108972454763063, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 7785 + }, + { + "epoch": 0.07786, + "grad_norm": 0.9879792571747996, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 7786 + }, + { + "epoch": 0.07787, + "grad_norm": 1.2023448684879607, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 7787 + }, + { + "epoch": 0.07788, + "grad_norm": 0.9538752828642438, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 7788 + }, + { + "epoch": 0.07789, + "grad_norm": 1.0308214934344375, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 7789 + }, + { + "epoch": 0.0779, + "grad_norm": 1.0650283574186172, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 7790 + }, + { + "epoch": 0.07791, + "grad_norm": 0.9941845094809683, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 7791 + }, + { + "epoch": 0.07792, + "grad_norm": 1.0225769158424693, + "learning_rate": 0.003, + "loss": 4.024, + "step": 7792 + }, + { + "epoch": 0.07793, + "grad_norm": 0.9877404519817934, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 7793 + }, + { + "epoch": 0.07794, + "grad_norm": 1.1151886303616545, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 7794 + }, + { + "epoch": 0.07795, + "grad_norm": 0.8354209579395252, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 7795 + }, + { + "epoch": 0.07796, + "grad_norm": 0.8544590286302752, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 7796 + }, + { + "epoch": 0.07797, + "grad_norm": 0.9428013064341034, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 7797 + }, + { + "epoch": 0.07798, + "grad_norm": 1.1393969133912794, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 7798 + }, + { + "epoch": 0.07799, + "grad_norm": 1.1641475465913624, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 7799 + }, + { + "epoch": 0.078, + "grad_norm": 1.0641337854621304, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 7800 + }, + { + "epoch": 0.07801, + "grad_norm": 1.3641996826697704, + "learning_rate": 0.003, + "loss": 4.055, + "step": 7801 + }, + { + "epoch": 0.07802, + "grad_norm": 1.1452620077902496, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 7802 + }, + { + "epoch": 0.07803, + "grad_norm": 1.1170775188247835, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 7803 + }, + { + "epoch": 0.07804, + "grad_norm": 1.296395122652786, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 7804 + }, + { + "epoch": 0.07805, + "grad_norm": 0.9159841573401555, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 7805 + }, + { + "epoch": 0.07806, + "grad_norm": 1.0076146294344561, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 7806 + }, + { + "epoch": 0.07807, + "grad_norm": 1.0288357992024348, + "learning_rate": 0.003, + "loss": 4.068, + "step": 7807 + }, + { + "epoch": 0.07808, + "grad_norm": 0.9962519995561984, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 7808 + }, + { + "epoch": 0.07809, + "grad_norm": 0.9318878928153395, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 7809 + }, + { + "epoch": 0.0781, + "grad_norm": 0.9578002546946285, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 7810 + }, + { + "epoch": 0.07811, + "grad_norm": 1.2362712301307206, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 7811 + }, + { + "epoch": 0.07812, + "grad_norm": 1.0902092466065818, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 7812 + }, + { + "epoch": 0.07813, + "grad_norm": 1.0656770200110066, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 7813 + }, + { + "epoch": 0.07814, + "grad_norm": 0.9616877242084323, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 7814 + }, + { + "epoch": 0.07815, + "grad_norm": 1.0827894290436153, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 7815 + }, + { + "epoch": 0.07816, + "grad_norm": 0.951863534026633, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 7816 + }, + { + "epoch": 0.07817, + "grad_norm": 1.0693420699978473, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 7817 + }, + { + "epoch": 0.07818, + "grad_norm": 1.1793875665990308, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 7818 + }, + { + "epoch": 0.07819, + "grad_norm": 0.9806237203551665, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 7819 + }, + { + "epoch": 0.0782, + "grad_norm": 1.074778244039972, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 7820 + }, + { + "epoch": 0.07821, + "grad_norm": 0.9930768173537193, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 7821 + }, + { + "epoch": 0.07822, + "grad_norm": 1.0685158401451227, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 7822 + }, + { + "epoch": 0.07823, + "grad_norm": 1.0871181323072148, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 7823 + }, + { + "epoch": 0.07824, + "grad_norm": 1.21966298784846, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 7824 + }, + { + "epoch": 0.07825, + "grad_norm": 1.0804803596808354, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 7825 + }, + { + "epoch": 0.07826, + "grad_norm": 1.019090987969821, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 7826 + }, + { + "epoch": 0.07827, + "grad_norm": 1.0988560306698911, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 7827 + }, + { + "epoch": 0.07828, + "grad_norm": 1.0517679422087889, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 7828 + }, + { + "epoch": 0.07829, + "grad_norm": 0.9621053073017727, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 7829 + }, + { + "epoch": 0.0783, + "grad_norm": 0.9689820333439864, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 7830 + }, + { + "epoch": 0.07831, + "grad_norm": 1.023016739445038, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 7831 + }, + { + "epoch": 0.07832, + "grad_norm": 0.9984156227464857, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 7832 + }, + { + "epoch": 0.07833, + "grad_norm": 0.9919107518305011, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 7833 + }, + { + "epoch": 0.07834, + "grad_norm": 1.1543019969673187, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 7834 + }, + { + "epoch": 0.07835, + "grad_norm": 1.0681590776729737, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 7835 + }, + { + "epoch": 0.07836, + "grad_norm": 1.0574570781003565, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 7836 + }, + { + "epoch": 0.07837, + "grad_norm": 1.2018757893559897, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 7837 + }, + { + "epoch": 0.07838, + "grad_norm": 0.9161356805849349, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 7838 + }, + { + "epoch": 0.07839, + "grad_norm": 0.9907463263104761, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 7839 + }, + { + "epoch": 0.0784, + "grad_norm": 1.1856358777888196, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 7840 + }, + { + "epoch": 0.07841, + "grad_norm": 0.8938988806605744, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 7841 + }, + { + "epoch": 0.07842, + "grad_norm": 1.0190036904110218, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 7842 + }, + { + "epoch": 0.07843, + "grad_norm": 1.2414498860928165, + "learning_rate": 0.003, + "loss": 4.057, + "step": 7843 + }, + { + "epoch": 0.07844, + "grad_norm": 0.8645804084986568, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 7844 + }, + { + "epoch": 0.07845, + "grad_norm": 1.2555968702914302, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 7845 + }, + { + "epoch": 0.07846, + "grad_norm": 1.1827535638877162, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 7846 + }, + { + "epoch": 0.07847, + "grad_norm": 1.0585750613727776, + "learning_rate": 0.003, + "loss": 4.027, + "step": 7847 + }, + { + "epoch": 0.07848, + "grad_norm": 1.0251038257294671, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 7848 + }, + { + "epoch": 0.07849, + "grad_norm": 1.0012756568810086, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 7849 + }, + { + "epoch": 0.0785, + "grad_norm": 1.1087682350572658, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 7850 + }, + { + "epoch": 0.07851, + "grad_norm": 1.0570558235082463, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 7851 + }, + { + "epoch": 0.07852, + "grad_norm": 1.1870988713474258, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 7852 + }, + { + "epoch": 0.07853, + "grad_norm": 1.1877824271372333, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 7853 + }, + { + "epoch": 0.07854, + "grad_norm": 0.9795576119811179, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 7854 + }, + { + "epoch": 0.07855, + "grad_norm": 1.0106904572896982, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 7855 + }, + { + "epoch": 0.07856, + "grad_norm": 1.074012949318807, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 7856 + }, + { + "epoch": 0.07857, + "grad_norm": 1.0682693133140797, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 7857 + }, + { + "epoch": 0.07858, + "grad_norm": 0.9622334233718971, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 7858 + }, + { + "epoch": 0.07859, + "grad_norm": 1.2435642589923817, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 7859 + }, + { + "epoch": 0.0786, + "grad_norm": 0.9713678691004027, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 7860 + }, + { + "epoch": 0.07861, + "grad_norm": 1.1992439198541343, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 7861 + }, + { + "epoch": 0.07862, + "grad_norm": 0.9729097717840791, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 7862 + }, + { + "epoch": 0.07863, + "grad_norm": 1.2123850094787298, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 7863 + }, + { + "epoch": 0.07864, + "grad_norm": 1.0554740247450138, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7864 + }, + { + "epoch": 0.07865, + "grad_norm": 1.2691503318540769, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 7865 + }, + { + "epoch": 0.07866, + "grad_norm": 0.9510762936067487, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 7866 + }, + { + "epoch": 0.07867, + "grad_norm": 1.0063305379540366, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 7867 + }, + { + "epoch": 0.07868, + "grad_norm": 1.0938115833363737, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 7868 + }, + { + "epoch": 0.07869, + "grad_norm": 1.0089952434242428, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 7869 + }, + { + "epoch": 0.0787, + "grad_norm": 1.0654347163848492, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 7870 + }, + { + "epoch": 0.07871, + "grad_norm": 1.1907030044141855, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 7871 + }, + { + "epoch": 0.07872, + "grad_norm": 0.8730754246990861, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 7872 + }, + { + "epoch": 0.07873, + "grad_norm": 1.0021741546132197, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 7873 + }, + { + "epoch": 0.07874, + "grad_norm": 1.4383882439567937, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 7874 + }, + { + "epoch": 0.07875, + "grad_norm": 0.8464322093253535, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 7875 + }, + { + "epoch": 0.07876, + "grad_norm": 0.86475881480814, + "learning_rate": 0.003, + "loss": 4.058, + "step": 7876 + }, + { + "epoch": 0.07877, + "grad_norm": 1.109652921566004, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 7877 + }, + { + "epoch": 0.07878, + "grad_norm": 1.0879946460453025, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 7878 + }, + { + "epoch": 0.07879, + "grad_norm": 1.0666099280508157, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 7879 + }, + { + "epoch": 0.0788, + "grad_norm": 1.1532229023404108, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 7880 + }, + { + "epoch": 0.07881, + "grad_norm": 0.9157988575667765, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7881 + }, + { + "epoch": 0.07882, + "grad_norm": 0.9989329398447048, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 7882 + }, + { + "epoch": 0.07883, + "grad_norm": 1.1179829054728383, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 7883 + }, + { + "epoch": 0.07884, + "grad_norm": 1.0047906870446375, + "learning_rate": 0.003, + "loss": 4.047, + "step": 7884 + }, + { + "epoch": 0.07885, + "grad_norm": 1.111240788046051, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 7885 + }, + { + "epoch": 0.07886, + "grad_norm": 1.0009988839259023, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 7886 + }, + { + "epoch": 0.07887, + "grad_norm": 1.0893230346535585, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 7887 + }, + { + "epoch": 0.07888, + "grad_norm": 1.05660595072792, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 7888 + }, + { + "epoch": 0.07889, + "grad_norm": 1.159409811390816, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 7889 + }, + { + "epoch": 0.0789, + "grad_norm": 0.8841007147309605, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 7890 + }, + { + "epoch": 0.07891, + "grad_norm": 1.1474805660791436, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 7891 + }, + { + "epoch": 0.07892, + "grad_norm": 1.0700800869552938, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 7892 + }, + { + "epoch": 0.07893, + "grad_norm": 1.0506523629123479, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 7893 + }, + { + "epoch": 0.07894, + "grad_norm": 1.1196303889660475, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 7894 + }, + { + "epoch": 0.07895, + "grad_norm": 1.28735191775539, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 7895 + }, + { + "epoch": 0.07896, + "grad_norm": 0.9643317762261, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 7896 + }, + { + "epoch": 0.07897, + "grad_norm": 1.1070256196078418, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 7897 + }, + { + "epoch": 0.07898, + "grad_norm": 1.0355518469983214, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 7898 + }, + { + "epoch": 0.07899, + "grad_norm": 1.2282329250453026, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 7899 + }, + { + "epoch": 0.079, + "grad_norm": 1.0945607454529, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 7900 + }, + { + "epoch": 0.07901, + "grad_norm": 1.0648956548783817, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 7901 + }, + { + "epoch": 0.07902, + "grad_norm": 1.0992859940840842, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 7902 + }, + { + "epoch": 0.07903, + "grad_norm": 0.8834013507865798, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 7903 + }, + { + "epoch": 0.07904, + "grad_norm": 0.9999633090920114, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 7904 + }, + { + "epoch": 0.07905, + "grad_norm": 1.0963454219201827, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 7905 + }, + { + "epoch": 0.07906, + "grad_norm": 0.9167535868840054, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 7906 + }, + { + "epoch": 0.07907, + "grad_norm": 1.0671834336811867, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 7907 + }, + { + "epoch": 0.07908, + "grad_norm": 0.9233899864843036, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 7908 + }, + { + "epoch": 0.07909, + "grad_norm": 1.0062840807702038, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 7909 + }, + { + "epoch": 0.0791, + "grad_norm": 1.235984238772413, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 7910 + }, + { + "epoch": 0.07911, + "grad_norm": 0.9452902767282305, + "learning_rate": 0.003, + "loss": 4.06, + "step": 7911 + }, + { + "epoch": 0.07912, + "grad_norm": 1.170408612990077, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 7912 + }, + { + "epoch": 0.07913, + "grad_norm": 0.96235951994912, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 7913 + }, + { + "epoch": 0.07914, + "grad_norm": 1.0169853669192117, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 7914 + }, + { + "epoch": 0.07915, + "grad_norm": 1.149306172968735, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 7915 + }, + { + "epoch": 0.07916, + "grad_norm": 0.9684272704389931, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 7916 + }, + { + "epoch": 0.07917, + "grad_norm": 1.0099443419186045, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 7917 + }, + { + "epoch": 0.07918, + "grad_norm": 1.2290204445682136, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7918 + }, + { + "epoch": 0.07919, + "grad_norm": 0.950110955160198, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 7919 + }, + { + "epoch": 0.0792, + "grad_norm": 1.1616333982044544, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 7920 + }, + { + "epoch": 0.07921, + "grad_norm": 1.2133949678821083, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 7921 + }, + { + "epoch": 0.07922, + "grad_norm": 0.8406932934598226, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 7922 + }, + { + "epoch": 0.07923, + "grad_norm": 1.000384156582474, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 7923 + }, + { + "epoch": 0.07924, + "grad_norm": 1.102645789500658, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 7924 + }, + { + "epoch": 0.07925, + "grad_norm": 1.0684901214506717, + "learning_rate": 0.003, + "loss": 4.061, + "step": 7925 + }, + { + "epoch": 0.07926, + "grad_norm": 1.293652658373931, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 7926 + }, + { + "epoch": 0.07927, + "grad_norm": 0.9389310994335556, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 7927 + }, + { + "epoch": 0.07928, + "grad_norm": 1.081617761304198, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 7928 + }, + { + "epoch": 0.07929, + "grad_norm": 0.9572759773190526, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 7929 + }, + { + "epoch": 0.0793, + "grad_norm": 1.150932445412838, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 7930 + }, + { + "epoch": 0.07931, + "grad_norm": 1.0459375556090538, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 7931 + }, + { + "epoch": 0.07932, + "grad_norm": 1.2162128355571467, + "learning_rate": 0.003, + "loss": 4.065, + "step": 7932 + }, + { + "epoch": 0.07933, + "grad_norm": 0.8823256635368836, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 7933 + }, + { + "epoch": 0.07934, + "grad_norm": 1.0009325472605832, + "learning_rate": 0.003, + "loss": 4.069, + "step": 7934 + }, + { + "epoch": 0.07935, + "grad_norm": 1.0096305457720214, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 7935 + }, + { + "epoch": 0.07936, + "grad_norm": 1.268310154088615, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 7936 + }, + { + "epoch": 0.07937, + "grad_norm": 0.9021101857561977, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 7937 + }, + { + "epoch": 0.07938, + "grad_norm": 1.0291910822408756, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 7938 + }, + { + "epoch": 0.07939, + "grad_norm": 1.2336928484312266, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 7939 + }, + { + "epoch": 0.0794, + "grad_norm": 0.9200004067230584, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 7940 + }, + { + "epoch": 0.07941, + "grad_norm": 1.2029826125319552, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 7941 + }, + { + "epoch": 0.07942, + "grad_norm": 0.9905351217758338, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 7942 + }, + { + "epoch": 0.07943, + "grad_norm": 1.1672820977045213, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 7943 + }, + { + "epoch": 0.07944, + "grad_norm": 0.9552098668578122, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 7944 + }, + { + "epoch": 0.07945, + "grad_norm": 1.1163448086544936, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 7945 + }, + { + "epoch": 0.07946, + "grad_norm": 1.1575061742112112, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 7946 + }, + { + "epoch": 0.07947, + "grad_norm": 0.9387057322081345, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 7947 + }, + { + "epoch": 0.07948, + "grad_norm": 1.036922515434725, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 7948 + }, + { + "epoch": 0.07949, + "grad_norm": 1.0571677074257047, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 7949 + }, + { + "epoch": 0.0795, + "grad_norm": 0.9485396434466922, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 7950 + }, + { + "epoch": 0.07951, + "grad_norm": 0.9358661587180489, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 7951 + }, + { + "epoch": 0.07952, + "grad_norm": 1.072262899758602, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 7952 + }, + { + "epoch": 0.07953, + "grad_norm": 1.1384679393249275, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 7953 + }, + { + "epoch": 0.07954, + "grad_norm": 1.188750550794104, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 7954 + }, + { + "epoch": 0.07955, + "grad_norm": 1.088794668390209, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 7955 + }, + { + "epoch": 0.07956, + "grad_norm": 0.944395467836122, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 7956 + }, + { + "epoch": 0.07957, + "grad_norm": 0.9876860612041447, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 7957 + }, + { + "epoch": 0.07958, + "grad_norm": 1.007844983649128, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7958 + }, + { + "epoch": 0.07959, + "grad_norm": 1.0039606523415259, + "learning_rate": 0.003, + "loss": 4.05, + "step": 7959 + }, + { + "epoch": 0.0796, + "grad_norm": 1.1452784506766251, + "learning_rate": 0.003, + "loss": 4.048, + "step": 7960 + }, + { + "epoch": 0.07961, + "grad_norm": 1.1250238968213198, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 7961 + }, + { + "epoch": 0.07962, + "grad_norm": 0.8771519916323887, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 7962 + }, + { + "epoch": 0.07963, + "grad_norm": 1.039301187147565, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 7963 + }, + { + "epoch": 0.07964, + "grad_norm": 1.1539629521338333, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 7964 + }, + { + "epoch": 0.07965, + "grad_norm": 1.006514542734939, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 7965 + }, + { + "epoch": 0.07966, + "grad_norm": 1.2419647152063866, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 7966 + }, + { + "epoch": 0.07967, + "grad_norm": 1.0445292324352928, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 7967 + }, + { + "epoch": 0.07968, + "grad_norm": 1.1283302172541068, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 7968 + }, + { + "epoch": 0.07969, + "grad_norm": 0.9667428222328616, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 7969 + }, + { + "epoch": 0.0797, + "grad_norm": 0.9853555487701827, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 7970 + }, + { + "epoch": 0.07971, + "grad_norm": 0.9314086481126423, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 7971 + }, + { + "epoch": 0.07972, + "grad_norm": 1.041017548263576, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 7972 + }, + { + "epoch": 0.07973, + "grad_norm": 1.1632363339227705, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 7973 + }, + { + "epoch": 0.07974, + "grad_norm": 0.9448198238046855, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 7974 + }, + { + "epoch": 0.07975, + "grad_norm": 1.2575451524297467, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 7975 + }, + { + "epoch": 0.07976, + "grad_norm": 0.9254382565643743, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 7976 + }, + { + "epoch": 0.07977, + "grad_norm": 0.9554319748351229, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 7977 + }, + { + "epoch": 0.07978, + "grad_norm": 1.1583747304811194, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 7978 + }, + { + "epoch": 0.07979, + "grad_norm": 1.1000321798363142, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 7979 + }, + { + "epoch": 0.0798, + "grad_norm": 0.8911730010907487, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 7980 + }, + { + "epoch": 0.07981, + "grad_norm": 1.020237000586781, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 7981 + }, + { + "epoch": 0.07982, + "grad_norm": 1.3180466552295464, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 7982 + }, + { + "epoch": 0.07983, + "grad_norm": 1.071553392145814, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 7983 + }, + { + "epoch": 0.07984, + "grad_norm": 1.2117191312115865, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 7984 + }, + { + "epoch": 0.07985, + "grad_norm": 0.7909400534108864, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 7985 + }, + { + "epoch": 0.07986, + "grad_norm": 0.69774180892976, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 7986 + }, + { + "epoch": 0.07987, + "grad_norm": 0.7627415677016504, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 7987 + }, + { + "epoch": 0.07988, + "grad_norm": 0.7977914673045347, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 7988 + }, + { + "epoch": 0.07989, + "grad_norm": 0.9057464007553878, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 7989 + }, + { + "epoch": 0.0799, + "grad_norm": 0.9968132407767837, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 7990 + }, + { + "epoch": 0.07991, + "grad_norm": 1.0049624387001375, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 7991 + }, + { + "epoch": 0.07992, + "grad_norm": 1.2238112277907751, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 7992 + }, + { + "epoch": 0.07993, + "grad_norm": 0.9170174217210549, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 7993 + }, + { + "epoch": 0.07994, + "grad_norm": 0.9120312741869245, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 7994 + }, + { + "epoch": 0.07995, + "grad_norm": 1.0185635782956717, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 7995 + }, + { + "epoch": 0.07996, + "grad_norm": 1.1924384020647054, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 7996 + }, + { + "epoch": 0.07997, + "grad_norm": 1.1865042797171061, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 7997 + }, + { + "epoch": 0.07998, + "grad_norm": 0.9979487786588422, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 7998 + }, + { + "epoch": 0.07999, + "grad_norm": 1.0145256510147855, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 7999 + }, + { + "epoch": 0.08, + "grad_norm": 1.0061604187402993, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 8000 + }, + { + "epoch": 0.08001, + "grad_norm": 1.0877536512304802, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 8001 + }, + { + "epoch": 0.08002, + "grad_norm": 1.103363283471722, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 8002 + }, + { + "epoch": 0.08003, + "grad_norm": 1.1565936510093848, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 8003 + }, + { + "epoch": 0.08004, + "grad_norm": 0.8339485185169685, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 8004 + }, + { + "epoch": 0.08005, + "grad_norm": 0.8957972908506947, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 8005 + }, + { + "epoch": 0.08006, + "grad_norm": 0.9755062166369889, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 8006 + }, + { + "epoch": 0.08007, + "grad_norm": 1.1356741968512036, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 8007 + }, + { + "epoch": 0.08008, + "grad_norm": 0.9798864056276021, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 8008 + }, + { + "epoch": 0.08009, + "grad_norm": 1.1435111910784777, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 8009 + }, + { + "epoch": 0.0801, + "grad_norm": 1.1205942333697474, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 8010 + }, + { + "epoch": 0.08011, + "grad_norm": 1.131476082802707, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 8011 + }, + { + "epoch": 0.08012, + "grad_norm": 1.0461557826019758, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 8012 + }, + { + "epoch": 0.08013, + "grad_norm": 1.0524770267958885, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 8013 + }, + { + "epoch": 0.08014, + "grad_norm": 1.2440328295727094, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 8014 + }, + { + "epoch": 0.08015, + "grad_norm": 1.114392646783546, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 8015 + }, + { + "epoch": 0.08016, + "grad_norm": 1.0812684439503044, + "learning_rate": 0.003, + "loss": 4.019, + "step": 8016 + }, + { + "epoch": 0.08017, + "grad_norm": 1.1433960945173143, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 8017 + }, + { + "epoch": 0.08018, + "grad_norm": 0.9868904103661147, + "learning_rate": 0.003, + "loss": 4.041, + "step": 8018 + }, + { + "epoch": 0.08019, + "grad_norm": 1.084639898239188, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 8019 + }, + { + "epoch": 0.0802, + "grad_norm": 1.185981726128112, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 8020 + }, + { + "epoch": 0.08021, + "grad_norm": 1.0766762166520623, + "learning_rate": 0.003, + "loss": 4.006, + "step": 8021 + }, + { + "epoch": 0.08022, + "grad_norm": 0.9287360641086851, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 8022 + }, + { + "epoch": 0.08023, + "grad_norm": 0.9456722393800089, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 8023 + }, + { + "epoch": 0.08024, + "grad_norm": 1.0828009737424313, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 8024 + }, + { + "epoch": 0.08025, + "grad_norm": 1.2456772261250684, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 8025 + }, + { + "epoch": 0.08026, + "grad_norm": 1.0148217772847923, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 8026 + }, + { + "epoch": 0.08027, + "grad_norm": 0.9138608081544262, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 8027 + }, + { + "epoch": 0.08028, + "grad_norm": 0.9517142464911736, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 8028 + }, + { + "epoch": 0.08029, + "grad_norm": 1.1073434017859498, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 8029 + }, + { + "epoch": 0.0803, + "grad_norm": 0.9290905977228465, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 8030 + }, + { + "epoch": 0.08031, + "grad_norm": 0.9226578478764562, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 8031 + }, + { + "epoch": 0.08032, + "grad_norm": 1.0759256080505066, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 8032 + }, + { + "epoch": 0.08033, + "grad_norm": 0.9511400028043164, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 8033 + }, + { + "epoch": 0.08034, + "grad_norm": 1.1078049780159342, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 8034 + }, + { + "epoch": 0.08035, + "grad_norm": 0.9535032742600911, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 8035 + }, + { + "epoch": 0.08036, + "grad_norm": 1.1614455582425816, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 8036 + }, + { + "epoch": 0.08037, + "grad_norm": 0.8834094122204313, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 8037 + }, + { + "epoch": 0.08038, + "grad_norm": 1.1130873275597826, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 8038 + }, + { + "epoch": 0.08039, + "grad_norm": 1.211408515493295, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 8039 + }, + { + "epoch": 0.0804, + "grad_norm": 1.1685320433873827, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 8040 + }, + { + "epoch": 0.08041, + "grad_norm": 1.0277080460599624, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 8041 + }, + { + "epoch": 0.08042, + "grad_norm": 1.1973269882264728, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 8042 + }, + { + "epoch": 0.08043, + "grad_norm": 0.8271614764410427, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8043 + }, + { + "epoch": 0.08044, + "grad_norm": 0.8887802571055234, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 8044 + }, + { + "epoch": 0.08045, + "grad_norm": 1.2110808135009032, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 8045 + }, + { + "epoch": 0.08046, + "grad_norm": 0.9518356758920827, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 8046 + }, + { + "epoch": 0.08047, + "grad_norm": 1.3038849837413196, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 8047 + }, + { + "epoch": 0.08048, + "grad_norm": 0.919567623717858, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 8048 + }, + { + "epoch": 0.08049, + "grad_norm": 0.8954045094329315, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 8049 + }, + { + "epoch": 0.0805, + "grad_norm": 1.0858667351123445, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 8050 + }, + { + "epoch": 0.08051, + "grad_norm": 1.0474964812508383, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 8051 + }, + { + "epoch": 0.08052, + "grad_norm": 0.998847972432671, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 8052 + }, + { + "epoch": 0.08053, + "grad_norm": 1.0859084509999066, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 8053 + }, + { + "epoch": 0.08054, + "grad_norm": 0.9520041929258798, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 8054 + }, + { + "epoch": 0.08055, + "grad_norm": 0.8849900316097494, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 8055 + }, + { + "epoch": 0.08056, + "grad_norm": 1.0200607183606778, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 8056 + }, + { + "epoch": 0.08057, + "grad_norm": 1.385440316855803, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 8057 + }, + { + "epoch": 0.08058, + "grad_norm": 0.7560965761499825, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 8058 + }, + { + "epoch": 0.08059, + "grad_norm": 0.802745797203238, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 8059 + }, + { + "epoch": 0.0806, + "grad_norm": 0.8780307041913072, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 8060 + }, + { + "epoch": 0.08061, + "grad_norm": 0.9388145447642858, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 8061 + }, + { + "epoch": 0.08062, + "grad_norm": 1.1392056930959564, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 8062 + }, + { + "epoch": 0.08063, + "grad_norm": 1.1500771661810032, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 8063 + }, + { + "epoch": 0.08064, + "grad_norm": 1.2259785226656719, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 8064 + }, + { + "epoch": 0.08065, + "grad_norm": 0.9482855137356565, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 8065 + }, + { + "epoch": 0.08066, + "grad_norm": 1.1941185655993347, + "learning_rate": 0.003, + "loss": 4.069, + "step": 8066 + }, + { + "epoch": 0.08067, + "grad_norm": 1.0216574976533404, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 8067 + }, + { + "epoch": 0.08068, + "grad_norm": 1.1815373632833652, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 8068 + }, + { + "epoch": 0.08069, + "grad_norm": 0.8362728337965947, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 8069 + }, + { + "epoch": 0.0807, + "grad_norm": 1.0141924727587812, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 8070 + }, + { + "epoch": 0.08071, + "grad_norm": 1.2085013425881042, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 8071 + }, + { + "epoch": 0.08072, + "grad_norm": 1.1611590433461545, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 8072 + }, + { + "epoch": 0.08073, + "grad_norm": 1.1178371421514612, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 8073 + }, + { + "epoch": 0.08074, + "grad_norm": 0.8685364005438322, + "learning_rate": 0.003, + "loss": 4.022, + "step": 8074 + }, + { + "epoch": 0.08075, + "grad_norm": 0.9730008967866436, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 8075 + }, + { + "epoch": 0.08076, + "grad_norm": 1.2822538457110706, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 8076 + }, + { + "epoch": 0.08077, + "grad_norm": 1.0798964686596582, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 8077 + }, + { + "epoch": 0.08078, + "grad_norm": 1.2985664488989839, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 8078 + }, + { + "epoch": 0.08079, + "grad_norm": 0.9957269652282372, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 8079 + }, + { + "epoch": 0.0808, + "grad_norm": 0.8700008490746549, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 8080 + }, + { + "epoch": 0.08081, + "grad_norm": 0.9054583409173683, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 8081 + }, + { + "epoch": 0.08082, + "grad_norm": 0.9766818304459607, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 8082 + }, + { + "epoch": 0.08083, + "grad_norm": 1.0138171171604249, + "learning_rate": 0.003, + "loss": 4.037, + "step": 8083 + }, + { + "epoch": 0.08084, + "grad_norm": 1.1036070375778735, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 8084 + }, + { + "epoch": 0.08085, + "grad_norm": 1.1006361031316971, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 8085 + }, + { + "epoch": 0.08086, + "grad_norm": 1.2665221292524846, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 8086 + }, + { + "epoch": 0.08087, + "grad_norm": 0.87235683124875, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 8087 + }, + { + "epoch": 0.08088, + "grad_norm": 1.2101069976094259, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 8088 + }, + { + "epoch": 0.08089, + "grad_norm": 1.174020295847284, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 8089 + }, + { + "epoch": 0.0809, + "grad_norm": 1.0782301641574674, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 8090 + }, + { + "epoch": 0.08091, + "grad_norm": 1.3340904191723806, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 8091 + }, + { + "epoch": 0.08092, + "grad_norm": 0.9678394845670897, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 8092 + }, + { + "epoch": 0.08093, + "grad_norm": 1.1614114147755075, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 8093 + }, + { + "epoch": 0.08094, + "grad_norm": 1.1087224012161654, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 8094 + }, + { + "epoch": 0.08095, + "grad_norm": 1.280636136936636, + "learning_rate": 0.003, + "loss": 4.026, + "step": 8095 + }, + { + "epoch": 0.08096, + "grad_norm": 0.9820162541292741, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 8096 + }, + { + "epoch": 0.08097, + "grad_norm": 1.1089399244430402, + "learning_rate": 0.003, + "loss": 4.042, + "step": 8097 + }, + { + "epoch": 0.08098, + "grad_norm": 0.9715722278401574, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 8098 + }, + { + "epoch": 0.08099, + "grad_norm": 1.164836620545108, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 8099 + }, + { + "epoch": 0.081, + "grad_norm": 1.0306069756917127, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 8100 + }, + { + "epoch": 0.08101, + "grad_norm": 1.1971074710915723, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 8101 + }, + { + "epoch": 0.08102, + "grad_norm": 0.9529922879670738, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 8102 + }, + { + "epoch": 0.08103, + "grad_norm": 1.1839304856546549, + "learning_rate": 0.003, + "loss": 4.054, + "step": 8103 + }, + { + "epoch": 0.08104, + "grad_norm": 0.9885159799435862, + "learning_rate": 0.003, + "loss": 4.047, + "step": 8104 + }, + { + "epoch": 0.08105, + "grad_norm": 1.0385726215167874, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 8105 + }, + { + "epoch": 0.08106, + "grad_norm": 0.9513496918238994, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 8106 + }, + { + "epoch": 0.08107, + "grad_norm": 1.1227985010536334, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 8107 + }, + { + "epoch": 0.08108, + "grad_norm": 1.0247223199647508, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 8108 + }, + { + "epoch": 0.08109, + "grad_norm": 1.0113529850755245, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 8109 + }, + { + "epoch": 0.0811, + "grad_norm": 1.1642917754838757, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 8110 + }, + { + "epoch": 0.08111, + "grad_norm": 0.885899156879861, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 8111 + }, + { + "epoch": 0.08112, + "grad_norm": 1.0233932927167342, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 8112 + }, + { + "epoch": 0.08113, + "grad_norm": 1.1478499269861702, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 8113 + }, + { + "epoch": 0.08114, + "grad_norm": 1.191391018906478, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 8114 + }, + { + "epoch": 0.08115, + "grad_norm": 1.0398287191821558, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 8115 + }, + { + "epoch": 0.08116, + "grad_norm": 1.029396069350254, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 8116 + }, + { + "epoch": 0.08117, + "grad_norm": 1.0620529797775633, + "learning_rate": 0.003, + "loss": 4.067, + "step": 8117 + }, + { + "epoch": 0.08118, + "grad_norm": 1.153228275579987, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 8118 + }, + { + "epoch": 0.08119, + "grad_norm": 1.0235694255861916, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 8119 + }, + { + "epoch": 0.0812, + "grad_norm": 1.2920287937035275, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 8120 + }, + { + "epoch": 0.08121, + "grad_norm": 0.798826528722648, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 8121 + }, + { + "epoch": 0.08122, + "grad_norm": 0.8251701793435979, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8122 + }, + { + "epoch": 0.08123, + "grad_norm": 0.9500120105243509, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 8123 + }, + { + "epoch": 0.08124, + "grad_norm": 1.1358148106927366, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 8124 + }, + { + "epoch": 0.08125, + "grad_norm": 0.9701373726642997, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 8125 + }, + { + "epoch": 0.08126, + "grad_norm": 1.303900333086603, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 8126 + }, + { + "epoch": 0.08127, + "grad_norm": 0.8348257183122614, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 8127 + }, + { + "epoch": 0.08128, + "grad_norm": 1.0168038658961494, + "learning_rate": 0.003, + "loss": 4.058, + "step": 8128 + }, + { + "epoch": 0.08129, + "grad_norm": 1.2570226327195435, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 8129 + }, + { + "epoch": 0.0813, + "grad_norm": 0.9835919552429621, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 8130 + }, + { + "epoch": 0.08131, + "grad_norm": 1.4786402848670535, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 8131 + }, + { + "epoch": 0.08132, + "grad_norm": 0.967482520755732, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 8132 + }, + { + "epoch": 0.08133, + "grad_norm": 1.0136679388024379, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 8133 + }, + { + "epoch": 0.08134, + "grad_norm": 1.3243507873969376, + "learning_rate": 0.003, + "loss": 4.04, + "step": 8134 + }, + { + "epoch": 0.08135, + "grad_norm": 1.1659701473075903, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8135 + }, + { + "epoch": 0.08136, + "grad_norm": 1.1146756067652475, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 8136 + }, + { + "epoch": 0.08137, + "grad_norm": 1.0102380938375048, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 8137 + }, + { + "epoch": 0.08138, + "grad_norm": 1.1104761872750726, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 8138 + }, + { + "epoch": 0.08139, + "grad_norm": 0.9806068045939812, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 8139 + }, + { + "epoch": 0.0814, + "grad_norm": 0.9578534939613226, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 8140 + }, + { + "epoch": 0.08141, + "grad_norm": 0.9783961516579012, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 8141 + }, + { + "epoch": 0.08142, + "grad_norm": 1.1102754179258492, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 8142 + }, + { + "epoch": 0.08143, + "grad_norm": 1.013761291015948, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 8143 + }, + { + "epoch": 0.08144, + "grad_norm": 1.0984848305189987, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 8144 + }, + { + "epoch": 0.08145, + "grad_norm": 1.0741146027806776, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 8145 + }, + { + "epoch": 0.08146, + "grad_norm": 1.0850941853188774, + "learning_rate": 0.003, + "loss": 4.07, + "step": 8146 + }, + { + "epoch": 0.08147, + "grad_norm": 1.0900280533229438, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 8147 + }, + { + "epoch": 0.08148, + "grad_norm": 1.1612246962032904, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 8148 + }, + { + "epoch": 0.08149, + "grad_norm": 1.3562111466807953, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 8149 + }, + { + "epoch": 0.0815, + "grad_norm": 0.9650515229317034, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 8150 + }, + { + "epoch": 0.08151, + "grad_norm": 1.1924509745774654, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 8151 + }, + { + "epoch": 0.08152, + "grad_norm": 1.094500106195963, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 8152 + }, + { + "epoch": 0.08153, + "grad_norm": 1.004251942188992, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 8153 + }, + { + "epoch": 0.08154, + "grad_norm": 1.1798560492422325, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 8154 + }, + { + "epoch": 0.08155, + "grad_norm": 1.0157829602982544, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 8155 + }, + { + "epoch": 0.08156, + "grad_norm": 1.2891395825570224, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8156 + }, + { + "epoch": 0.08157, + "grad_norm": 0.9232969874496607, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 8157 + }, + { + "epoch": 0.08158, + "grad_norm": 1.0373678761694196, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 8158 + }, + { + "epoch": 0.08159, + "grad_norm": 1.3218980216269214, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 8159 + }, + { + "epoch": 0.0816, + "grad_norm": 0.9490086271587503, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 8160 + }, + { + "epoch": 0.08161, + "grad_norm": 1.0833199103673854, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8161 + }, + { + "epoch": 0.08162, + "grad_norm": 1.2185544482985367, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 8162 + }, + { + "epoch": 0.08163, + "grad_norm": 0.9763505701602616, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 8163 + }, + { + "epoch": 0.08164, + "grad_norm": 1.066961091347535, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 8164 + }, + { + "epoch": 0.08165, + "grad_norm": 1.0509794131255215, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 8165 + }, + { + "epoch": 0.08166, + "grad_norm": 1.0757821468791668, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 8166 + }, + { + "epoch": 0.08167, + "grad_norm": 1.3397321654362855, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 8167 + }, + { + "epoch": 0.08168, + "grad_norm": 1.1362149152066932, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 8168 + }, + { + "epoch": 0.08169, + "grad_norm": 0.8763737391154698, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 8169 + }, + { + "epoch": 0.0817, + "grad_norm": 0.8325229180364018, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 8170 + }, + { + "epoch": 0.08171, + "grad_norm": 0.9119090572200499, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 8171 + }, + { + "epoch": 0.08172, + "grad_norm": 0.9166054245530775, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 8172 + }, + { + "epoch": 0.08173, + "grad_norm": 1.0922618916234996, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 8173 + }, + { + "epoch": 0.08174, + "grad_norm": 1.0948674878262272, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 8174 + }, + { + "epoch": 0.08175, + "grad_norm": 1.076487219894897, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 8175 + }, + { + "epoch": 0.08176, + "grad_norm": 1.0595377058120148, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 8176 + }, + { + "epoch": 0.08177, + "grad_norm": 1.1274198711334715, + "learning_rate": 0.003, + "loss": 4.034, + "step": 8177 + }, + { + "epoch": 0.08178, + "grad_norm": 1.0171853704517495, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 8178 + }, + { + "epoch": 0.08179, + "grad_norm": 1.3061262539701504, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 8179 + }, + { + "epoch": 0.0818, + "grad_norm": 0.9862361455218118, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 8180 + }, + { + "epoch": 0.08181, + "grad_norm": 1.248761174376701, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 8181 + }, + { + "epoch": 0.08182, + "grad_norm": 1.04459958881984, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 8182 + }, + { + "epoch": 0.08183, + "grad_norm": 1.034490751276438, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8183 + }, + { + "epoch": 0.08184, + "grad_norm": 1.3084515904264853, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 8184 + }, + { + "epoch": 0.08185, + "grad_norm": 0.9554920895816984, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8185 + }, + { + "epoch": 0.08186, + "grad_norm": 1.0543602272809869, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 8186 + }, + { + "epoch": 0.08187, + "grad_norm": 1.1260343697487754, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 8187 + }, + { + "epoch": 0.08188, + "grad_norm": 1.0245037440997349, + "learning_rate": 0.003, + "loss": 4.068, + "step": 8188 + }, + { + "epoch": 0.08189, + "grad_norm": 1.0290894153169319, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 8189 + }, + { + "epoch": 0.0819, + "grad_norm": 0.877806506800442, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 8190 + }, + { + "epoch": 0.08191, + "grad_norm": 0.8759876139628601, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 8191 + }, + { + "epoch": 0.08192, + "grad_norm": 1.0235597494522326, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 8192 + }, + { + "epoch": 0.08193, + "grad_norm": 1.0680156272560604, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 8193 + }, + { + "epoch": 0.08194, + "grad_norm": 0.9866872972933308, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 8194 + }, + { + "epoch": 0.08195, + "grad_norm": 1.0454954573691113, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 8195 + }, + { + "epoch": 0.08196, + "grad_norm": 0.878803096910163, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 8196 + }, + { + "epoch": 0.08197, + "grad_norm": 1.097095395513429, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 8197 + }, + { + "epoch": 0.08198, + "grad_norm": 1.4128502904759463, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 8198 + }, + { + "epoch": 0.08199, + "grad_norm": 0.9349249325846235, + "learning_rate": 0.003, + "loss": 4.049, + "step": 8199 + }, + { + "epoch": 0.082, + "grad_norm": 1.1340924047935328, + "learning_rate": 0.003, + "loss": 4.027, + "step": 8200 + }, + { + "epoch": 0.08201, + "grad_norm": 1.04912739515356, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 8201 + }, + { + "epoch": 0.08202, + "grad_norm": 1.0714645593844108, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 8202 + }, + { + "epoch": 0.08203, + "grad_norm": 1.164096753838808, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 8203 + }, + { + "epoch": 0.08204, + "grad_norm": 1.1316835410725452, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 8204 + }, + { + "epoch": 0.08205, + "grad_norm": 1.2120673109706728, + "learning_rate": 0.003, + "loss": 4.045, + "step": 8205 + }, + { + "epoch": 0.08206, + "grad_norm": 1.11571994095571, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 8206 + }, + { + "epoch": 0.08207, + "grad_norm": 1.0621146376384545, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 8207 + }, + { + "epoch": 0.08208, + "grad_norm": 1.1747546800940614, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 8208 + }, + { + "epoch": 0.08209, + "grad_norm": 0.8502432613824764, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 8209 + }, + { + "epoch": 0.0821, + "grad_norm": 0.9220719569509254, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 8210 + }, + { + "epoch": 0.08211, + "grad_norm": 1.2327782482092768, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8211 + }, + { + "epoch": 0.08212, + "grad_norm": 1.019547573968891, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 8212 + }, + { + "epoch": 0.08213, + "grad_norm": 1.1264918883793147, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 8213 + }, + { + "epoch": 0.08214, + "grad_norm": 1.0153304383270647, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 8214 + }, + { + "epoch": 0.08215, + "grad_norm": 1.172513559366227, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 8215 + }, + { + "epoch": 0.08216, + "grad_norm": 1.1761345458122565, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 8216 + }, + { + "epoch": 0.08217, + "grad_norm": 1.0086245946196835, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 8217 + }, + { + "epoch": 0.08218, + "grad_norm": 1.0192220496141946, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 8218 + }, + { + "epoch": 0.08219, + "grad_norm": 1.1072279418839062, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8219 + }, + { + "epoch": 0.0822, + "grad_norm": 1.2129949090775702, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 8220 + }, + { + "epoch": 0.08221, + "grad_norm": 1.335506660895218, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 8221 + }, + { + "epoch": 0.08222, + "grad_norm": 0.9100821320950854, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 8222 + }, + { + "epoch": 0.08223, + "grad_norm": 1.2359285375629427, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 8223 + }, + { + "epoch": 0.08224, + "grad_norm": 1.1420581603411728, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 8224 + }, + { + "epoch": 0.08225, + "grad_norm": 1.131941471315245, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 8225 + }, + { + "epoch": 0.08226, + "grad_norm": 0.9193337169990431, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 8226 + }, + { + "epoch": 0.08227, + "grad_norm": 0.9322903091927021, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 8227 + }, + { + "epoch": 0.08228, + "grad_norm": 1.015271996777266, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 8228 + }, + { + "epoch": 0.08229, + "grad_norm": 1.134812005676848, + "learning_rate": 0.003, + "loss": 4.025, + "step": 8229 + }, + { + "epoch": 0.0823, + "grad_norm": 0.9067833302627255, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 8230 + }, + { + "epoch": 0.08231, + "grad_norm": 1.1844642289666762, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 8231 + }, + { + "epoch": 0.08232, + "grad_norm": 1.074667737158904, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 8232 + }, + { + "epoch": 0.08233, + "grad_norm": 0.9455096038944548, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 8233 + }, + { + "epoch": 0.08234, + "grad_norm": 0.9451177013419999, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 8234 + }, + { + "epoch": 0.08235, + "grad_norm": 0.9751741426641366, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 8235 + }, + { + "epoch": 0.08236, + "grad_norm": 1.0139644613308296, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 8236 + }, + { + "epoch": 0.08237, + "grad_norm": 1.1102447919282574, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 8237 + }, + { + "epoch": 0.08238, + "grad_norm": 1.090996849580925, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 8238 + }, + { + "epoch": 0.08239, + "grad_norm": 1.1072648040046784, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 8239 + }, + { + "epoch": 0.0824, + "grad_norm": 0.963339170365707, + "learning_rate": 0.003, + "loss": 4.059, + "step": 8240 + }, + { + "epoch": 0.08241, + "grad_norm": 1.1416193830206067, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 8241 + }, + { + "epoch": 0.08242, + "grad_norm": 1.1161581273297645, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 8242 + }, + { + "epoch": 0.08243, + "grad_norm": 0.9836364537147602, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 8243 + }, + { + "epoch": 0.08244, + "grad_norm": 1.0051947565508417, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 8244 + }, + { + "epoch": 0.08245, + "grad_norm": 1.179881665824301, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8245 + }, + { + "epoch": 0.08246, + "grad_norm": 1.0994155612110759, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 8246 + }, + { + "epoch": 0.08247, + "grad_norm": 1.1514528107070527, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 8247 + }, + { + "epoch": 0.08248, + "grad_norm": 1.1175713935986686, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 8248 + }, + { + "epoch": 0.08249, + "grad_norm": 1.022832537141348, + "learning_rate": 0.003, + "loss": 4.056, + "step": 8249 + }, + { + "epoch": 0.0825, + "grad_norm": 1.2208883258896968, + "learning_rate": 0.003, + "loss": 4.035, + "step": 8250 + }, + { + "epoch": 0.08251, + "grad_norm": 0.9539194420518737, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 8251 + }, + { + "epoch": 0.08252, + "grad_norm": 1.0155458495353762, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 8252 + }, + { + "epoch": 0.08253, + "grad_norm": 1.241505078520073, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 8253 + }, + { + "epoch": 0.08254, + "grad_norm": 0.9520715771435223, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 8254 + }, + { + "epoch": 0.08255, + "grad_norm": 1.024935048858687, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 8255 + }, + { + "epoch": 0.08256, + "grad_norm": 1.233328543047205, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 8256 + }, + { + "epoch": 0.08257, + "grad_norm": 1.190460709783594, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 8257 + }, + { + "epoch": 0.08258, + "grad_norm": 1.0367792756754817, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8258 + }, + { + "epoch": 0.08259, + "grad_norm": 1.169563709355769, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 8259 + }, + { + "epoch": 0.0826, + "grad_norm": 0.861527127229618, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 8260 + }, + { + "epoch": 0.08261, + "grad_norm": 0.8201507958780685, + "learning_rate": 0.003, + "loss": 4.028, + "step": 8261 + }, + { + "epoch": 0.08262, + "grad_norm": 0.9386291752296059, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 8262 + }, + { + "epoch": 0.08263, + "grad_norm": 1.0230682657821621, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8263 + }, + { + "epoch": 0.08264, + "grad_norm": 1.2158135102547172, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 8264 + }, + { + "epoch": 0.08265, + "grad_norm": 0.9858027331224795, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 8265 + }, + { + "epoch": 0.08266, + "grad_norm": 1.1514602100198823, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 8266 + }, + { + "epoch": 0.08267, + "grad_norm": 1.19867896784868, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 8267 + }, + { + "epoch": 0.08268, + "grad_norm": 0.8924446930144582, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 8268 + }, + { + "epoch": 0.08269, + "grad_norm": 0.9332387923779609, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 8269 + }, + { + "epoch": 0.0827, + "grad_norm": 1.0395207333244374, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8270 + }, + { + "epoch": 0.08271, + "grad_norm": 1.1217717636669475, + "learning_rate": 0.003, + "loss": 4.058, + "step": 8271 + }, + { + "epoch": 0.08272, + "grad_norm": 1.158015224233601, + "learning_rate": 0.003, + "loss": 4.049, + "step": 8272 + }, + { + "epoch": 0.08273, + "grad_norm": 1.2991517374950396, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 8273 + }, + { + "epoch": 0.08274, + "grad_norm": 1.081398113932392, + "learning_rate": 0.003, + "loss": 4.059, + "step": 8274 + }, + { + "epoch": 0.08275, + "grad_norm": 1.1176536482392028, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 8275 + }, + { + "epoch": 0.08276, + "grad_norm": 0.9350722584845963, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 8276 + }, + { + "epoch": 0.08277, + "grad_norm": 0.9894919434780528, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 8277 + }, + { + "epoch": 0.08278, + "grad_norm": 1.1769258157634674, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 8278 + }, + { + "epoch": 0.08279, + "grad_norm": 0.9433131457871408, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 8279 + }, + { + "epoch": 0.0828, + "grad_norm": 0.9576999263892194, + "learning_rate": 0.003, + "loss": 4.062, + "step": 8280 + }, + { + "epoch": 0.08281, + "grad_norm": 1.209178534584828, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 8281 + }, + { + "epoch": 0.08282, + "grad_norm": 0.9614787804606547, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8282 + }, + { + "epoch": 0.08283, + "grad_norm": 1.0971465514681484, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 8283 + }, + { + "epoch": 0.08284, + "grad_norm": 1.0107622943623467, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 8284 + }, + { + "epoch": 0.08285, + "grad_norm": 1.15514741446281, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 8285 + }, + { + "epoch": 0.08286, + "grad_norm": 1.0676521040179816, + "learning_rate": 0.003, + "loss": 4.065, + "step": 8286 + }, + { + "epoch": 0.08287, + "grad_norm": 1.2393097291371007, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 8287 + }, + { + "epoch": 0.08288, + "grad_norm": 1.229284881723444, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 8288 + }, + { + "epoch": 0.08289, + "grad_norm": 0.8030678208057539, + "learning_rate": 0.003, + "loss": 4.043, + "step": 8289 + }, + { + "epoch": 0.0829, + "grad_norm": 0.9234071041203235, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 8290 + }, + { + "epoch": 0.08291, + "grad_norm": 1.2192535372086108, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 8291 + }, + { + "epoch": 0.08292, + "grad_norm": 1.0516037727758725, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 8292 + }, + { + "epoch": 0.08293, + "grad_norm": 0.964068579815504, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 8293 + }, + { + "epoch": 0.08294, + "grad_norm": 1.07685445953297, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 8294 + }, + { + "epoch": 0.08295, + "grad_norm": 1.2651219299645922, + "learning_rate": 0.003, + "loss": 4.042, + "step": 8295 + }, + { + "epoch": 0.08296, + "grad_norm": 0.9626687885048552, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 8296 + }, + { + "epoch": 0.08297, + "grad_norm": 0.9687755374419604, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8297 + }, + { + "epoch": 0.08298, + "grad_norm": 1.3012272212633311, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 8298 + }, + { + "epoch": 0.08299, + "grad_norm": 0.9471340632405811, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 8299 + }, + { + "epoch": 0.083, + "grad_norm": 1.1889456485828054, + "learning_rate": 0.003, + "loss": 4.078, + "step": 8300 + }, + { + "epoch": 0.08301, + "grad_norm": 1.0926624866225692, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 8301 + }, + { + "epoch": 0.08302, + "grad_norm": 1.061747983105285, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 8302 + }, + { + "epoch": 0.08303, + "grad_norm": 1.2124144409859756, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 8303 + }, + { + "epoch": 0.08304, + "grad_norm": 1.0062093800648917, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 8304 + }, + { + "epoch": 0.08305, + "grad_norm": 1.0797635537801393, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 8305 + }, + { + "epoch": 0.08306, + "grad_norm": 1.2165041896744566, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 8306 + }, + { + "epoch": 0.08307, + "grad_norm": 0.9371964802211156, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 8307 + }, + { + "epoch": 0.08308, + "grad_norm": 1.0824545701121477, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 8308 + }, + { + "epoch": 0.08309, + "grad_norm": 1.0702785761306515, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 8309 + }, + { + "epoch": 0.0831, + "grad_norm": 0.9850867353661309, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 8310 + }, + { + "epoch": 0.08311, + "grad_norm": 1.059401767428526, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 8311 + }, + { + "epoch": 0.08312, + "grad_norm": 1.1525254824845346, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 8312 + }, + { + "epoch": 0.08313, + "grad_norm": 0.8651922401808925, + "learning_rate": 0.003, + "loss": 4.027, + "step": 8313 + }, + { + "epoch": 0.08314, + "grad_norm": 1.0602609432745576, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 8314 + }, + { + "epoch": 0.08315, + "grad_norm": 0.8182680455954303, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 8315 + }, + { + "epoch": 0.08316, + "grad_norm": 0.898921435803555, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 8316 + }, + { + "epoch": 0.08317, + "grad_norm": 1.262794585028179, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 8317 + }, + { + "epoch": 0.08318, + "grad_norm": 1.215745444075188, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 8318 + }, + { + "epoch": 0.08319, + "grad_norm": 1.019233573877823, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 8319 + }, + { + "epoch": 0.0832, + "grad_norm": 1.0398703718664923, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 8320 + }, + { + "epoch": 0.08321, + "grad_norm": 1.1361754781004614, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 8321 + }, + { + "epoch": 0.08322, + "grad_norm": 0.9093568729986535, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 8322 + }, + { + "epoch": 0.08323, + "grad_norm": 1.1369192324437696, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 8323 + }, + { + "epoch": 0.08324, + "grad_norm": 1.217916598721158, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8324 + }, + { + "epoch": 0.08325, + "grad_norm": 1.0915850365976298, + "learning_rate": 0.003, + "loss": 4.051, + "step": 8325 + }, + { + "epoch": 0.08326, + "grad_norm": 1.1565123938632185, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8326 + }, + { + "epoch": 0.08327, + "grad_norm": 0.8586661169517376, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 8327 + }, + { + "epoch": 0.08328, + "grad_norm": 0.8431082244999714, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 8328 + }, + { + "epoch": 0.08329, + "grad_norm": 0.831022312396331, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 8329 + }, + { + "epoch": 0.0833, + "grad_norm": 0.8854225855827627, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8330 + }, + { + "epoch": 0.08331, + "grad_norm": 0.9489920933886857, + "learning_rate": 0.003, + "loss": 4.025, + "step": 8331 + }, + { + "epoch": 0.08332, + "grad_norm": 1.0014721691993675, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 8332 + }, + { + "epoch": 0.08333, + "grad_norm": 1.1572373116887549, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 8333 + }, + { + "epoch": 0.08334, + "grad_norm": 1.0696738130518826, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 8334 + }, + { + "epoch": 0.08335, + "grad_norm": 1.134011488039698, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 8335 + }, + { + "epoch": 0.08336, + "grad_norm": 1.171624442165235, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 8336 + }, + { + "epoch": 0.08337, + "grad_norm": 0.9808780299153776, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 8337 + }, + { + "epoch": 0.08338, + "grad_norm": 1.1552726886292524, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 8338 + }, + { + "epoch": 0.08339, + "grad_norm": 1.119935444567191, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 8339 + }, + { + "epoch": 0.0834, + "grad_norm": 1.1254896342797662, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 8340 + }, + { + "epoch": 0.08341, + "grad_norm": 1.1341861913032731, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 8341 + }, + { + "epoch": 0.08342, + "grad_norm": 1.025162475919914, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 8342 + }, + { + "epoch": 0.08343, + "grad_norm": 1.205937843754898, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 8343 + }, + { + "epoch": 0.08344, + "grad_norm": 0.9902520923365202, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8344 + }, + { + "epoch": 0.08345, + "grad_norm": 1.288987691741149, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 8345 + }, + { + "epoch": 0.08346, + "grad_norm": 0.9551763232371178, + "learning_rate": 0.003, + "loss": 4.046, + "step": 8346 + }, + { + "epoch": 0.08347, + "grad_norm": 0.9880135377663202, + "learning_rate": 0.003, + "loss": 4.043, + "step": 8347 + }, + { + "epoch": 0.08348, + "grad_norm": 1.2888100524027102, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 8348 + }, + { + "epoch": 0.08349, + "grad_norm": 0.9525381368631001, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 8349 + }, + { + "epoch": 0.0835, + "grad_norm": 1.1902573539688603, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 8350 + }, + { + "epoch": 0.08351, + "grad_norm": 1.0469716149508292, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 8351 + }, + { + "epoch": 0.08352, + "grad_norm": 1.189680771320039, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 8352 + }, + { + "epoch": 0.08353, + "grad_norm": 0.9911168111334406, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 8353 + }, + { + "epoch": 0.08354, + "grad_norm": 1.0276631861654357, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 8354 + }, + { + "epoch": 0.08355, + "grad_norm": 0.9849728479941415, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 8355 + }, + { + "epoch": 0.08356, + "grad_norm": 1.01503709707315, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 8356 + }, + { + "epoch": 0.08357, + "grad_norm": 1.2200149399048952, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 8357 + }, + { + "epoch": 0.08358, + "grad_norm": 0.9309969506083857, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 8358 + }, + { + "epoch": 0.08359, + "grad_norm": 1.1903852595304782, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 8359 + }, + { + "epoch": 0.0836, + "grad_norm": 1.1339893715722242, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 8360 + }, + { + "epoch": 0.08361, + "grad_norm": 1.0643831400781818, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 8361 + }, + { + "epoch": 0.08362, + "grad_norm": 1.249532330084765, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 8362 + }, + { + "epoch": 0.08363, + "grad_norm": 0.8661252025928181, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 8363 + }, + { + "epoch": 0.08364, + "grad_norm": 0.90648079035645, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 8364 + }, + { + "epoch": 0.08365, + "grad_norm": 0.9973227794538887, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 8365 + }, + { + "epoch": 0.08366, + "grad_norm": 1.185845321832341, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 8366 + }, + { + "epoch": 0.08367, + "grad_norm": 1.0527110344332224, + "learning_rate": 0.003, + "loss": 4.038, + "step": 8367 + }, + { + "epoch": 0.08368, + "grad_norm": 1.1213177889830799, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 8368 + }, + { + "epoch": 0.08369, + "grad_norm": 0.8532852276611194, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 8369 + }, + { + "epoch": 0.0837, + "grad_norm": 0.9223226425587234, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 8370 + }, + { + "epoch": 0.08371, + "grad_norm": 1.0781243641141955, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8371 + }, + { + "epoch": 0.08372, + "grad_norm": 0.9339538120598732, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 8372 + }, + { + "epoch": 0.08373, + "grad_norm": 1.4452876155141794, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 8373 + }, + { + "epoch": 0.08374, + "grad_norm": 0.9063853871722677, + "learning_rate": 0.003, + "loss": 4.024, + "step": 8374 + }, + { + "epoch": 0.08375, + "grad_norm": 0.993278332875694, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 8375 + }, + { + "epoch": 0.08376, + "grad_norm": 1.1855350373423543, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 8376 + }, + { + "epoch": 0.08377, + "grad_norm": 1.1147746898162996, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 8377 + }, + { + "epoch": 0.08378, + "grad_norm": 1.138214657903677, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 8378 + }, + { + "epoch": 0.08379, + "grad_norm": 1.2479011609022754, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 8379 + }, + { + "epoch": 0.0838, + "grad_norm": 1.1416308067638565, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 8380 + }, + { + "epoch": 0.08381, + "grad_norm": 0.8970625864687897, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 8381 + }, + { + "epoch": 0.08382, + "grad_norm": 1.1310709176053892, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 8382 + }, + { + "epoch": 0.08383, + "grad_norm": 1.2920665241918188, + "learning_rate": 0.003, + "loss": 4.028, + "step": 8383 + }, + { + "epoch": 0.08384, + "grad_norm": 1.0350723875968093, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 8384 + }, + { + "epoch": 0.08385, + "grad_norm": 1.103275398901523, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 8385 + }, + { + "epoch": 0.08386, + "grad_norm": 0.9681045513350701, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 8386 + }, + { + "epoch": 0.08387, + "grad_norm": 1.355083711142688, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 8387 + }, + { + "epoch": 0.08388, + "grad_norm": 0.7934403977447374, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 8388 + }, + { + "epoch": 0.08389, + "grad_norm": 0.9516574206564951, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 8389 + }, + { + "epoch": 0.0839, + "grad_norm": 1.1290072369116717, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 8390 + }, + { + "epoch": 0.08391, + "grad_norm": 1.0983041546433672, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 8391 + }, + { + "epoch": 0.08392, + "grad_norm": 1.1676787780611126, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 8392 + }, + { + "epoch": 0.08393, + "grad_norm": 0.941880845010308, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 8393 + }, + { + "epoch": 0.08394, + "grad_norm": 1.0752262942555577, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 8394 + }, + { + "epoch": 0.08395, + "grad_norm": 1.0717439798578736, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 8395 + }, + { + "epoch": 0.08396, + "grad_norm": 0.923502718166825, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 8396 + }, + { + "epoch": 0.08397, + "grad_norm": 1.03568380249952, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 8397 + }, + { + "epoch": 0.08398, + "grad_norm": 1.2308296952539537, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 8398 + }, + { + "epoch": 0.08399, + "grad_norm": 0.9074167959796567, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8399 + }, + { + "epoch": 0.084, + "grad_norm": 1.0676950120843756, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 8400 + }, + { + "epoch": 0.08401, + "grad_norm": 1.1876371217178374, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 8401 + }, + { + "epoch": 0.08402, + "grad_norm": 0.9744971214519869, + "learning_rate": 0.003, + "loss": 4.017, + "step": 8402 + }, + { + "epoch": 0.08403, + "grad_norm": 1.1750219156987762, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 8403 + }, + { + "epoch": 0.08404, + "grad_norm": 1.0407406867360138, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 8404 + }, + { + "epoch": 0.08405, + "grad_norm": 1.2236517472521589, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 8405 + }, + { + "epoch": 0.08406, + "grad_norm": 1.0737329492962817, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 8406 + }, + { + "epoch": 0.08407, + "grad_norm": 1.1563501503982763, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 8407 + }, + { + "epoch": 0.08408, + "grad_norm": 1.151672035981857, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 8408 + }, + { + "epoch": 0.08409, + "grad_norm": 1.052519356480212, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 8409 + }, + { + "epoch": 0.0841, + "grad_norm": 1.3255584980401665, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 8410 + }, + { + "epoch": 0.08411, + "grad_norm": 0.7725053175360581, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 8411 + }, + { + "epoch": 0.08412, + "grad_norm": 0.9932518870157506, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 8412 + }, + { + "epoch": 0.08413, + "grad_norm": 1.3097609444125302, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 8413 + }, + { + "epoch": 0.08414, + "grad_norm": 0.8806084620399991, + "learning_rate": 0.003, + "loss": 4.057, + "step": 8414 + }, + { + "epoch": 0.08415, + "grad_norm": 1.1049915504817205, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 8415 + }, + { + "epoch": 0.08416, + "grad_norm": 1.0825647357972656, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 8416 + }, + { + "epoch": 0.08417, + "grad_norm": 0.982855881055315, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 8417 + }, + { + "epoch": 0.08418, + "grad_norm": 0.9521266964049695, + "learning_rate": 0.003, + "loss": 4.033, + "step": 8418 + }, + { + "epoch": 0.08419, + "grad_norm": 0.9807370150605913, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 8419 + }, + { + "epoch": 0.0842, + "grad_norm": 1.1284553332406668, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 8420 + }, + { + "epoch": 0.08421, + "grad_norm": 1.0701358935667225, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 8421 + }, + { + "epoch": 0.08422, + "grad_norm": 1.110363256131976, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 8422 + }, + { + "epoch": 0.08423, + "grad_norm": 1.0301265941795035, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 8423 + }, + { + "epoch": 0.08424, + "grad_norm": 1.2241888249026445, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 8424 + }, + { + "epoch": 0.08425, + "grad_norm": 1.2132223786728582, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 8425 + }, + { + "epoch": 0.08426, + "grad_norm": 1.053345472153612, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 8426 + }, + { + "epoch": 0.08427, + "grad_norm": 1.062648403306003, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 8427 + }, + { + "epoch": 0.08428, + "grad_norm": 0.9969338637013739, + "learning_rate": 0.003, + "loss": 4.044, + "step": 8428 + }, + { + "epoch": 0.08429, + "grad_norm": 1.1167478557553572, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 8429 + }, + { + "epoch": 0.0843, + "grad_norm": 0.9976920929387573, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 8430 + }, + { + "epoch": 0.08431, + "grad_norm": 1.068620820351791, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 8431 + }, + { + "epoch": 0.08432, + "grad_norm": 1.055723055328784, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 8432 + }, + { + "epoch": 0.08433, + "grad_norm": 1.135026949998236, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 8433 + }, + { + "epoch": 0.08434, + "grad_norm": 1.046614082716392, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 8434 + }, + { + "epoch": 0.08435, + "grad_norm": 1.2281345462765678, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 8435 + }, + { + "epoch": 0.08436, + "grad_norm": 1.0815704353084468, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 8436 + }, + { + "epoch": 0.08437, + "grad_norm": 1.0822120658950827, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 8437 + }, + { + "epoch": 0.08438, + "grad_norm": 1.0610781680082266, + "learning_rate": 0.003, + "loss": 4.039, + "step": 8438 + }, + { + "epoch": 0.08439, + "grad_norm": 1.1633949783616182, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 8439 + }, + { + "epoch": 0.0844, + "grad_norm": 1.2351950497416953, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 8440 + }, + { + "epoch": 0.08441, + "grad_norm": 0.9160963850716966, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 8441 + }, + { + "epoch": 0.08442, + "grad_norm": 1.0403778744976664, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 8442 + }, + { + "epoch": 0.08443, + "grad_norm": 0.9690769171898936, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 8443 + }, + { + "epoch": 0.08444, + "grad_norm": 1.0046811419165038, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 8444 + }, + { + "epoch": 0.08445, + "grad_norm": 1.1457379760825268, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 8445 + }, + { + "epoch": 0.08446, + "grad_norm": 1.0784653343143604, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 8446 + }, + { + "epoch": 0.08447, + "grad_norm": 0.9955575399367267, + "learning_rate": 0.003, + "loss": 4.053, + "step": 8447 + }, + { + "epoch": 0.08448, + "grad_norm": 1.0185612550385477, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 8448 + }, + { + "epoch": 0.08449, + "grad_norm": 1.1385992623128935, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 8449 + }, + { + "epoch": 0.0845, + "grad_norm": 1.1658036316438427, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 8450 + }, + { + "epoch": 0.08451, + "grad_norm": 1.0612990389711034, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 8451 + }, + { + "epoch": 0.08452, + "grad_norm": 1.1305785397594885, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 8452 + }, + { + "epoch": 0.08453, + "grad_norm": 0.8831441944452132, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 8453 + }, + { + "epoch": 0.08454, + "grad_norm": 1.1184930943973512, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 8454 + }, + { + "epoch": 0.08455, + "grad_norm": 1.3097730422669984, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 8455 + }, + { + "epoch": 0.08456, + "grad_norm": 0.9486933571257935, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8456 + }, + { + "epoch": 0.08457, + "grad_norm": 1.0027555477605425, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 8457 + }, + { + "epoch": 0.08458, + "grad_norm": 1.1962971696576739, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 8458 + }, + { + "epoch": 0.08459, + "grad_norm": 0.9387597135483401, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 8459 + }, + { + "epoch": 0.0846, + "grad_norm": 1.0383172475711102, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 8460 + }, + { + "epoch": 0.08461, + "grad_norm": 1.030965491865683, + "learning_rate": 0.003, + "loss": 4.058, + "step": 8461 + }, + { + "epoch": 0.08462, + "grad_norm": 1.2171453720071779, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 8462 + }, + { + "epoch": 0.08463, + "grad_norm": 1.0376733249187324, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 8463 + }, + { + "epoch": 0.08464, + "grad_norm": 1.0961626196613234, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 8464 + }, + { + "epoch": 0.08465, + "grad_norm": 1.0350888302747112, + "learning_rate": 0.003, + "loss": 4.047, + "step": 8465 + }, + { + "epoch": 0.08466, + "grad_norm": 1.064724029669904, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 8466 + }, + { + "epoch": 0.08467, + "grad_norm": 1.0710400208790887, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 8467 + }, + { + "epoch": 0.08468, + "grad_norm": 1.2789305198093153, + "learning_rate": 0.003, + "loss": 4.063, + "step": 8468 + }, + { + "epoch": 0.08469, + "grad_norm": 1.1452225551848991, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 8469 + }, + { + "epoch": 0.0847, + "grad_norm": 1.13409301786806, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 8470 + }, + { + "epoch": 0.08471, + "grad_norm": 1.1566161039870437, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 8471 + }, + { + "epoch": 0.08472, + "grad_norm": 1.089186577688795, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 8472 + }, + { + "epoch": 0.08473, + "grad_norm": 0.9900808592861112, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 8473 + }, + { + "epoch": 0.08474, + "grad_norm": 1.2998498233820543, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 8474 + }, + { + "epoch": 0.08475, + "grad_norm": 0.9225329005828342, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 8475 + }, + { + "epoch": 0.08476, + "grad_norm": 1.0564013920428952, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 8476 + }, + { + "epoch": 0.08477, + "grad_norm": 0.939864643930408, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 8477 + }, + { + "epoch": 0.08478, + "grad_norm": 0.9883124870718386, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8478 + }, + { + "epoch": 0.08479, + "grad_norm": 0.9957851651749337, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 8479 + }, + { + "epoch": 0.0848, + "grad_norm": 1.178478915544431, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 8480 + }, + { + "epoch": 0.08481, + "grad_norm": 1.0741787634257696, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 8481 + }, + { + "epoch": 0.08482, + "grad_norm": 1.1194627479559824, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 8482 + }, + { + "epoch": 0.08483, + "grad_norm": 0.9960554568951283, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8483 + }, + { + "epoch": 0.08484, + "grad_norm": 1.361511264699175, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 8484 + }, + { + "epoch": 0.08485, + "grad_norm": 1.1007501689045969, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 8485 + }, + { + "epoch": 0.08486, + "grad_norm": 1.174240597139128, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 8486 + }, + { + "epoch": 0.08487, + "grad_norm": 0.9639151938457295, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 8487 + }, + { + "epoch": 0.08488, + "grad_norm": 1.0352335647337247, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 8488 + }, + { + "epoch": 0.08489, + "grad_norm": 1.0493143503227644, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 8489 + }, + { + "epoch": 0.0849, + "grad_norm": 1.1208147184681883, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 8490 + }, + { + "epoch": 0.08491, + "grad_norm": 0.9315118881737822, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 8491 + }, + { + "epoch": 0.08492, + "grad_norm": 1.1354408904086009, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 8492 + }, + { + "epoch": 0.08493, + "grad_norm": 0.9756810751721894, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 8493 + }, + { + "epoch": 0.08494, + "grad_norm": 0.943874329273859, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 8494 + }, + { + "epoch": 0.08495, + "grad_norm": 0.9530790823279345, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 8495 + }, + { + "epoch": 0.08496, + "grad_norm": 1.0476242283034254, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 8496 + }, + { + "epoch": 0.08497, + "grad_norm": 1.2230946630382007, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 8497 + }, + { + "epoch": 0.08498, + "grad_norm": 1.1262934455755145, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 8498 + }, + { + "epoch": 0.08499, + "grad_norm": 1.1704572369620323, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 8499 + }, + { + "epoch": 0.085, + "grad_norm": 1.0437922946142013, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 8500 + }, + { + "epoch": 0.08501, + "grad_norm": 1.0686795257894022, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 8501 + }, + { + "epoch": 0.08502, + "grad_norm": 1.1527737710293182, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 8502 + }, + { + "epoch": 0.08503, + "grad_norm": 1.0146162316100584, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 8503 + }, + { + "epoch": 0.08504, + "grad_norm": 1.117321943006214, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 8504 + }, + { + "epoch": 0.08505, + "grad_norm": 1.1211708270006748, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 8505 + }, + { + "epoch": 0.08506, + "grad_norm": 1.0564903712086136, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 8506 + }, + { + "epoch": 0.08507, + "grad_norm": 1.0674811926990067, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 8507 + }, + { + "epoch": 0.08508, + "grad_norm": 1.361628142349855, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 8508 + }, + { + "epoch": 0.08509, + "grad_norm": 0.9588268292198073, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 8509 + }, + { + "epoch": 0.0851, + "grad_norm": 1.4027990342622492, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 8510 + }, + { + "epoch": 0.08511, + "grad_norm": 1.0313983203700878, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 8511 + }, + { + "epoch": 0.08512, + "grad_norm": 1.014747429488357, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 8512 + }, + { + "epoch": 0.08513, + "grad_norm": 1.1188271554621882, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 8513 + }, + { + "epoch": 0.08514, + "grad_norm": 0.9977123655522062, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 8514 + }, + { + "epoch": 0.08515, + "grad_norm": 1.0545034904025057, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 8515 + }, + { + "epoch": 0.08516, + "grad_norm": 1.2561317636760059, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 8516 + }, + { + "epoch": 0.08517, + "grad_norm": 1.015842099395916, + "learning_rate": 0.003, + "loss": 4.038, + "step": 8517 + }, + { + "epoch": 0.08518, + "grad_norm": 1.1154744018152205, + "learning_rate": 0.003, + "loss": 4.032, + "step": 8518 + }, + { + "epoch": 0.08519, + "grad_norm": 0.9308156900461738, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 8519 + }, + { + "epoch": 0.0852, + "grad_norm": 1.0123290179587556, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 8520 + }, + { + "epoch": 0.08521, + "grad_norm": 0.9696253911148752, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 8521 + }, + { + "epoch": 0.08522, + "grad_norm": 1.0651444858320227, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 8522 + }, + { + "epoch": 0.08523, + "grad_norm": 1.221416691858745, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8523 + }, + { + "epoch": 0.08524, + "grad_norm": 1.0694054720010744, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 8524 + }, + { + "epoch": 0.08525, + "grad_norm": 1.206105613858707, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8525 + }, + { + "epoch": 0.08526, + "grad_norm": 1.125746365698348, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 8526 + }, + { + "epoch": 0.08527, + "grad_norm": 1.021213535127077, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 8527 + }, + { + "epoch": 0.08528, + "grad_norm": 1.0514486808333348, + "learning_rate": 0.003, + "loss": 4.02, + "step": 8528 + }, + { + "epoch": 0.08529, + "grad_norm": 0.8891881765729944, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8529 + }, + { + "epoch": 0.0853, + "grad_norm": 1.0090942600896682, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 8530 + }, + { + "epoch": 0.08531, + "grad_norm": 1.077165689433922, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 8531 + }, + { + "epoch": 0.08532, + "grad_norm": 0.8886817388250743, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 8532 + }, + { + "epoch": 0.08533, + "grad_norm": 0.9499157917354463, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 8533 + }, + { + "epoch": 0.08534, + "grad_norm": 1.0462406691660049, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 8534 + }, + { + "epoch": 0.08535, + "grad_norm": 1.0427155178314294, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 8535 + }, + { + "epoch": 0.08536, + "grad_norm": 1.3345887932955967, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 8536 + }, + { + "epoch": 0.08537, + "grad_norm": 1.1424324189819963, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 8537 + }, + { + "epoch": 0.08538, + "grad_norm": 1.1087130620488592, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 8538 + }, + { + "epoch": 0.08539, + "grad_norm": 0.9423257796614134, + "learning_rate": 0.003, + "loss": 4.06, + "step": 8539 + }, + { + "epoch": 0.0854, + "grad_norm": 0.9989943262329959, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 8540 + }, + { + "epoch": 0.08541, + "grad_norm": 1.1159602342043562, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 8541 + }, + { + "epoch": 0.08542, + "grad_norm": 1.1750101404966282, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 8542 + }, + { + "epoch": 0.08543, + "grad_norm": 1.1097646268558357, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 8543 + }, + { + "epoch": 0.08544, + "grad_norm": 1.2246197163952683, + "learning_rate": 0.003, + "loss": 4.03, + "step": 8544 + }, + { + "epoch": 0.08545, + "grad_norm": 0.9006431906440999, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 8545 + }, + { + "epoch": 0.08546, + "grad_norm": 0.8535243027708214, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 8546 + }, + { + "epoch": 0.08547, + "grad_norm": 0.9091940967690285, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8547 + }, + { + "epoch": 0.08548, + "grad_norm": 1.1204422349465504, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 8548 + }, + { + "epoch": 0.08549, + "grad_norm": 1.1376178894662337, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 8549 + }, + { + "epoch": 0.0855, + "grad_norm": 0.9004834981881831, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 8550 + }, + { + "epoch": 0.08551, + "grad_norm": 1.0908856802279696, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 8551 + }, + { + "epoch": 0.08552, + "grad_norm": 1.0747606537103462, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 8552 + }, + { + "epoch": 0.08553, + "grad_norm": 1.1824536376723846, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 8553 + }, + { + "epoch": 0.08554, + "grad_norm": 0.9327035296984333, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 8554 + }, + { + "epoch": 0.08555, + "grad_norm": 1.0845301764121567, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 8555 + }, + { + "epoch": 0.08556, + "grad_norm": 1.0781500615217907, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 8556 + }, + { + "epoch": 0.08557, + "grad_norm": 1.1114030326740636, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 8557 + }, + { + "epoch": 0.08558, + "grad_norm": 1.1387705031741684, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 8558 + }, + { + "epoch": 0.08559, + "grad_norm": 1.0373792894377512, + "learning_rate": 0.003, + "loss": 4.048, + "step": 8559 + }, + { + "epoch": 0.0856, + "grad_norm": 1.205780356820795, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 8560 + }, + { + "epoch": 0.08561, + "grad_norm": 0.8516562843948833, + "learning_rate": 0.003, + "loss": 4.032, + "step": 8561 + }, + { + "epoch": 0.08562, + "grad_norm": 0.933449495259518, + "learning_rate": 0.003, + "loss": 4.007, + "step": 8562 + }, + { + "epoch": 0.08563, + "grad_norm": 1.0726820040186111, + "learning_rate": 0.003, + "loss": 4.038, + "step": 8563 + }, + { + "epoch": 0.08564, + "grad_norm": 1.0347716349195137, + "learning_rate": 0.003, + "loss": 4.034, + "step": 8564 + }, + { + "epoch": 0.08565, + "grad_norm": 1.0186874520180855, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 8565 + }, + { + "epoch": 0.08566, + "grad_norm": 1.1233566323209978, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 8566 + }, + { + "epoch": 0.08567, + "grad_norm": 1.1961057328383637, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 8567 + }, + { + "epoch": 0.08568, + "grad_norm": 1.0511200423129479, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 8568 + }, + { + "epoch": 0.08569, + "grad_norm": 1.1119472613821024, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 8569 + }, + { + "epoch": 0.0857, + "grad_norm": 1.1318905094833156, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 8570 + }, + { + "epoch": 0.08571, + "grad_norm": 1.3705938071159904, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 8571 + }, + { + "epoch": 0.08572, + "grad_norm": 0.8350836435173717, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 8572 + }, + { + "epoch": 0.08573, + "grad_norm": 0.9539575731341017, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8573 + }, + { + "epoch": 0.08574, + "grad_norm": 1.2227418528249123, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 8574 + }, + { + "epoch": 0.08575, + "grad_norm": 1.0933729376633516, + "learning_rate": 0.003, + "loss": 4.053, + "step": 8575 + }, + { + "epoch": 0.08576, + "grad_norm": 1.011476248931269, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 8576 + }, + { + "epoch": 0.08577, + "grad_norm": 0.9840937697692118, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 8577 + }, + { + "epoch": 0.08578, + "grad_norm": 1.202980530727398, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 8578 + }, + { + "epoch": 0.08579, + "grad_norm": 0.9372317642711228, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 8579 + }, + { + "epoch": 0.0858, + "grad_norm": 1.1233693209921423, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 8580 + }, + { + "epoch": 0.08581, + "grad_norm": 1.2444195714201942, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 8581 + }, + { + "epoch": 0.08582, + "grad_norm": 1.1497113076403636, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 8582 + }, + { + "epoch": 0.08583, + "grad_norm": 0.8869126729713231, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 8583 + }, + { + "epoch": 0.08584, + "grad_norm": 0.9033828604652565, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 8584 + }, + { + "epoch": 0.08585, + "grad_norm": 1.2413242779151106, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 8585 + }, + { + "epoch": 0.08586, + "grad_norm": 0.8785419954113871, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 8586 + }, + { + "epoch": 0.08587, + "grad_norm": 1.0721276325176368, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 8587 + }, + { + "epoch": 0.08588, + "grad_norm": 1.1679761583906054, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8588 + }, + { + "epoch": 0.08589, + "grad_norm": 1.080344129233115, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 8589 + }, + { + "epoch": 0.0859, + "grad_norm": 0.9870833943241247, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8590 + }, + { + "epoch": 0.08591, + "grad_norm": 1.034639577282747, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 8591 + }, + { + "epoch": 0.08592, + "grad_norm": 1.0710719468181558, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 8592 + }, + { + "epoch": 0.08593, + "grad_norm": 0.9888408563640142, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 8593 + }, + { + "epoch": 0.08594, + "grad_norm": 1.2453584999330138, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 8594 + }, + { + "epoch": 0.08595, + "grad_norm": 1.0196678830693755, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 8595 + }, + { + "epoch": 0.08596, + "grad_norm": 1.1945792959997095, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 8596 + }, + { + "epoch": 0.08597, + "grad_norm": 1.1530581283872166, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 8597 + }, + { + "epoch": 0.08598, + "grad_norm": 1.3926960034791933, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 8598 + }, + { + "epoch": 0.08599, + "grad_norm": 0.826192764112842, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 8599 + }, + { + "epoch": 0.086, + "grad_norm": 1.06077334266566, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 8600 + }, + { + "epoch": 0.08601, + "grad_norm": 1.1871115259087262, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 8601 + }, + { + "epoch": 0.08602, + "grad_norm": 1.2181866405394508, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 8602 + }, + { + "epoch": 0.08603, + "grad_norm": 1.0883340096060563, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 8603 + }, + { + "epoch": 0.08604, + "grad_norm": 0.9503342506965254, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 8604 + }, + { + "epoch": 0.08605, + "grad_norm": 1.2172592642031064, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 8605 + }, + { + "epoch": 0.08606, + "grad_norm": 1.136390303673651, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 8606 + }, + { + "epoch": 0.08607, + "grad_norm": 1.1355782479254777, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 8607 + }, + { + "epoch": 0.08608, + "grad_norm": 1.0517606318333685, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8608 + }, + { + "epoch": 0.08609, + "grad_norm": 0.9420025556721784, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 8609 + }, + { + "epoch": 0.0861, + "grad_norm": 0.9141311653122305, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 8610 + }, + { + "epoch": 0.08611, + "grad_norm": 1.0548509864539932, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 8611 + }, + { + "epoch": 0.08612, + "grad_norm": 1.001800534694237, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 8612 + }, + { + "epoch": 0.08613, + "grad_norm": 1.0972442988257152, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 8613 + }, + { + "epoch": 0.08614, + "grad_norm": 0.9428815080476953, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 8614 + }, + { + "epoch": 0.08615, + "grad_norm": 1.0484637576712974, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 8615 + }, + { + "epoch": 0.08616, + "grad_norm": 1.0917980757047034, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 8616 + }, + { + "epoch": 0.08617, + "grad_norm": 1.2097536460049183, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 8617 + }, + { + "epoch": 0.08618, + "grad_norm": 1.3562182619090937, + "learning_rate": 0.003, + "loss": 4.051, + "step": 8618 + }, + { + "epoch": 0.08619, + "grad_norm": 0.9246195111624731, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 8619 + }, + { + "epoch": 0.0862, + "grad_norm": 0.9824968489316007, + "learning_rate": 0.003, + "loss": 4.046, + "step": 8620 + }, + { + "epoch": 0.08621, + "grad_norm": 1.0258471276787318, + "learning_rate": 0.003, + "loss": 4.035, + "step": 8621 + }, + { + "epoch": 0.08622, + "grad_norm": 1.183578647619695, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 8622 + }, + { + "epoch": 0.08623, + "grad_norm": 1.1381975301836866, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 8623 + }, + { + "epoch": 0.08624, + "grad_norm": 1.1630194311964601, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 8624 + }, + { + "epoch": 0.08625, + "grad_norm": 0.9306587451164217, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 8625 + }, + { + "epoch": 0.08626, + "grad_norm": 1.0498514896735527, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 8626 + }, + { + "epoch": 0.08627, + "grad_norm": 1.1755016236915463, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 8627 + }, + { + "epoch": 0.08628, + "grad_norm": 0.938529239831759, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 8628 + }, + { + "epoch": 0.08629, + "grad_norm": 1.0144244468682728, + "learning_rate": 0.003, + "loss": 4.043, + "step": 8629 + }, + { + "epoch": 0.0863, + "grad_norm": 1.113738693838708, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 8630 + }, + { + "epoch": 0.08631, + "grad_norm": 1.1256675059051566, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 8631 + }, + { + "epoch": 0.08632, + "grad_norm": 0.8974862907760961, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 8632 + }, + { + "epoch": 0.08633, + "grad_norm": 1.0495156855470718, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 8633 + }, + { + "epoch": 0.08634, + "grad_norm": 1.2948675647945798, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 8634 + }, + { + "epoch": 0.08635, + "grad_norm": 1.0184506641521571, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8635 + }, + { + "epoch": 0.08636, + "grad_norm": 1.368584927067475, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 8636 + }, + { + "epoch": 0.08637, + "grad_norm": 0.9162061134202947, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 8637 + }, + { + "epoch": 0.08638, + "grad_norm": 1.0811639530557986, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 8638 + }, + { + "epoch": 0.08639, + "grad_norm": 0.9376836857950404, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 8639 + }, + { + "epoch": 0.0864, + "grad_norm": 1.0873412665068496, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 8640 + }, + { + "epoch": 0.08641, + "grad_norm": 1.1255192902825057, + "learning_rate": 0.003, + "loss": 4.049, + "step": 8641 + }, + { + "epoch": 0.08642, + "grad_norm": 1.1145569844437186, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 8642 + }, + { + "epoch": 0.08643, + "grad_norm": 1.4460777384526706, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 8643 + }, + { + "epoch": 0.08644, + "grad_norm": 0.9245977533541625, + "learning_rate": 0.003, + "loss": 4.022, + "step": 8644 + }, + { + "epoch": 0.08645, + "grad_norm": 1.0181172313173914, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 8645 + }, + { + "epoch": 0.08646, + "grad_norm": 1.1454056707809717, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 8646 + }, + { + "epoch": 0.08647, + "grad_norm": 1.0065775077778298, + "learning_rate": 0.003, + "loss": 4.007, + "step": 8647 + }, + { + "epoch": 0.08648, + "grad_norm": 1.2202424385911437, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 8648 + }, + { + "epoch": 0.08649, + "grad_norm": 0.9257171603914495, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 8649 + }, + { + "epoch": 0.0865, + "grad_norm": 0.9390491681338301, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 8650 + }, + { + "epoch": 0.08651, + "grad_norm": 1.1931127237305668, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 8651 + }, + { + "epoch": 0.08652, + "grad_norm": 0.9274015309476554, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 8652 + }, + { + "epoch": 0.08653, + "grad_norm": 1.02289631221439, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 8653 + }, + { + "epoch": 0.08654, + "grad_norm": 1.3262485339916, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 8654 + }, + { + "epoch": 0.08655, + "grad_norm": 1.2428022427956695, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 8655 + }, + { + "epoch": 0.08656, + "grad_norm": 1.031840617295688, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 8656 + }, + { + "epoch": 0.08657, + "grad_norm": 1.325567296493681, + "learning_rate": 0.003, + "loss": 4.076, + "step": 8657 + }, + { + "epoch": 0.08658, + "grad_norm": 1.0189750997869491, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 8658 + }, + { + "epoch": 0.08659, + "grad_norm": 1.209380922792147, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 8659 + }, + { + "epoch": 0.0866, + "grad_norm": 1.0610985840216802, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 8660 + }, + { + "epoch": 0.08661, + "grad_norm": 1.0561305118599713, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 8661 + }, + { + "epoch": 0.08662, + "grad_norm": 1.0679146127811068, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 8662 + }, + { + "epoch": 0.08663, + "grad_norm": 0.9064928152737026, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 8663 + }, + { + "epoch": 0.08664, + "grad_norm": 1.0923363230039982, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 8664 + }, + { + "epoch": 0.08665, + "grad_norm": 1.1988464238998455, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 8665 + }, + { + "epoch": 0.08666, + "grad_norm": 1.0248023151858396, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 8666 + }, + { + "epoch": 0.08667, + "grad_norm": 1.1148548362397608, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 8667 + }, + { + "epoch": 0.08668, + "grad_norm": 1.0412822855875483, + "learning_rate": 0.003, + "loss": 4.032, + "step": 8668 + }, + { + "epoch": 0.08669, + "grad_norm": 1.2096435963768866, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 8669 + }, + { + "epoch": 0.0867, + "grad_norm": 1.123266840921136, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 8670 + }, + { + "epoch": 0.08671, + "grad_norm": 1.0877613190872542, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 8671 + }, + { + "epoch": 0.08672, + "grad_norm": 1.1406500907143475, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 8672 + }, + { + "epoch": 0.08673, + "grad_norm": 0.9398367617096632, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 8673 + }, + { + "epoch": 0.08674, + "grad_norm": 1.0253292803580725, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 8674 + }, + { + "epoch": 0.08675, + "grad_norm": 1.1953506016711317, + "learning_rate": 0.003, + "loss": 4.056, + "step": 8675 + }, + { + "epoch": 0.08676, + "grad_norm": 0.9488718707211112, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 8676 + }, + { + "epoch": 0.08677, + "grad_norm": 1.0431682629276453, + "learning_rate": 0.003, + "loss": 4.027, + "step": 8677 + }, + { + "epoch": 0.08678, + "grad_norm": 1.193005747951154, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 8678 + }, + { + "epoch": 0.08679, + "grad_norm": 1.0144417828651375, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8679 + }, + { + "epoch": 0.0868, + "grad_norm": 1.1348128884099915, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 8680 + }, + { + "epoch": 0.08681, + "grad_norm": 1.1097911864091226, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 8681 + }, + { + "epoch": 0.08682, + "grad_norm": 1.0697846104835937, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8682 + }, + { + "epoch": 0.08683, + "grad_norm": 1.0607468846889823, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 8683 + }, + { + "epoch": 0.08684, + "grad_norm": 1.0862632076397618, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 8684 + }, + { + "epoch": 0.08685, + "grad_norm": 1.1148108606990261, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8685 + }, + { + "epoch": 0.08686, + "grad_norm": 1.0892861530880062, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 8686 + }, + { + "epoch": 0.08687, + "grad_norm": 1.2033955582498337, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 8687 + }, + { + "epoch": 0.08688, + "grad_norm": 0.9860469350967052, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8688 + }, + { + "epoch": 0.08689, + "grad_norm": 0.887845078826922, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 8689 + }, + { + "epoch": 0.0869, + "grad_norm": 0.981947563992108, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 8690 + }, + { + "epoch": 0.08691, + "grad_norm": 1.2386584406913188, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 8691 + }, + { + "epoch": 0.08692, + "grad_norm": 1.1214274326159783, + "learning_rate": 0.003, + "loss": 4.047, + "step": 8692 + }, + { + "epoch": 0.08693, + "grad_norm": 0.9759935969342685, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 8693 + }, + { + "epoch": 0.08694, + "grad_norm": 0.9952282248413553, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 8694 + }, + { + "epoch": 0.08695, + "grad_norm": 1.2054815926609244, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 8695 + }, + { + "epoch": 0.08696, + "grad_norm": 0.9636503750723039, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 8696 + }, + { + "epoch": 0.08697, + "grad_norm": 1.0925520713595345, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 8697 + }, + { + "epoch": 0.08698, + "grad_norm": 0.8854366350443418, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 8698 + }, + { + "epoch": 0.08699, + "grad_norm": 0.9842767035973804, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 8699 + }, + { + "epoch": 0.087, + "grad_norm": 1.1497030489175517, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 8700 + }, + { + "epoch": 0.08701, + "grad_norm": 1.220812201035816, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 8701 + }, + { + "epoch": 0.08702, + "grad_norm": 1.0247846783178878, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8702 + }, + { + "epoch": 0.08703, + "grad_norm": 1.0010386929193278, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 8703 + }, + { + "epoch": 0.08704, + "grad_norm": 1.0616711530427767, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 8704 + }, + { + "epoch": 0.08705, + "grad_norm": 0.9976320350325579, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 8705 + }, + { + "epoch": 0.08706, + "grad_norm": 1.1942693214140914, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 8706 + }, + { + "epoch": 0.08707, + "grad_norm": 1.073245473169283, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 8707 + }, + { + "epoch": 0.08708, + "grad_norm": 1.1941251771612393, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 8708 + }, + { + "epoch": 0.08709, + "grad_norm": 1.054313122339438, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 8709 + }, + { + "epoch": 0.0871, + "grad_norm": 1.1735283159957155, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 8710 + }, + { + "epoch": 0.08711, + "grad_norm": 1.224731249555308, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 8711 + }, + { + "epoch": 0.08712, + "grad_norm": 1.1071733779330812, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 8712 + }, + { + "epoch": 0.08713, + "grad_norm": 1.207587077778471, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 8713 + }, + { + "epoch": 0.08714, + "grad_norm": 1.035590781407661, + "learning_rate": 0.003, + "loss": 4.042, + "step": 8714 + }, + { + "epoch": 0.08715, + "grad_norm": 0.9210397372280619, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 8715 + }, + { + "epoch": 0.08716, + "grad_norm": 1.0736853335371308, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 8716 + }, + { + "epoch": 0.08717, + "grad_norm": 1.2449278196726106, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 8717 + }, + { + "epoch": 0.08718, + "grad_norm": 0.9356165132727563, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 8718 + }, + { + "epoch": 0.08719, + "grad_norm": 1.1410701425015397, + "learning_rate": 0.003, + "loss": 4.041, + "step": 8719 + }, + { + "epoch": 0.0872, + "grad_norm": 1.0473904852864253, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 8720 + }, + { + "epoch": 0.08721, + "grad_norm": 1.1346851924291141, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 8721 + }, + { + "epoch": 0.08722, + "grad_norm": 0.9947376087216057, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8722 + }, + { + "epoch": 0.08723, + "grad_norm": 1.2565032752641185, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 8723 + }, + { + "epoch": 0.08724, + "grad_norm": 1.013947154344129, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8724 + }, + { + "epoch": 0.08725, + "grad_norm": 1.2550507340568873, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8725 + }, + { + "epoch": 0.08726, + "grad_norm": 1.1552353519239422, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 8726 + }, + { + "epoch": 0.08727, + "grad_norm": 1.0836695734061212, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 8727 + }, + { + "epoch": 0.08728, + "grad_norm": 0.9999247167010374, + "learning_rate": 0.003, + "loss": 4.011, + "step": 8728 + }, + { + "epoch": 0.08729, + "grad_norm": 1.142902216241174, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 8729 + }, + { + "epoch": 0.0873, + "grad_norm": 1.091022164759464, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 8730 + }, + { + "epoch": 0.08731, + "grad_norm": 1.0210290074405686, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 8731 + }, + { + "epoch": 0.08732, + "grad_norm": 1.1449497076583306, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 8732 + }, + { + "epoch": 0.08733, + "grad_norm": 1.0036816386493732, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 8733 + }, + { + "epoch": 0.08734, + "grad_norm": 1.3034518911251176, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 8734 + }, + { + "epoch": 0.08735, + "grad_norm": 1.035050110835489, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 8735 + }, + { + "epoch": 0.08736, + "grad_norm": 1.059239037825648, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8736 + }, + { + "epoch": 0.08737, + "grad_norm": 1.1209703870238803, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 8737 + }, + { + "epoch": 0.08738, + "grad_norm": 0.897095918930477, + "learning_rate": 0.003, + "loss": 4.02, + "step": 8738 + }, + { + "epoch": 0.08739, + "grad_norm": 0.9994054796041234, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 8739 + }, + { + "epoch": 0.0874, + "grad_norm": 1.3452908356127442, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 8740 + }, + { + "epoch": 0.08741, + "grad_norm": 1.1093603042004494, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 8741 + }, + { + "epoch": 0.08742, + "grad_norm": 0.9596481129611945, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 8742 + }, + { + "epoch": 0.08743, + "grad_norm": 1.0733682110935197, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 8743 + }, + { + "epoch": 0.08744, + "grad_norm": 1.212621060059208, + "learning_rate": 0.003, + "loss": 4.066, + "step": 8744 + }, + { + "epoch": 0.08745, + "grad_norm": 1.0255959780393786, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 8745 + }, + { + "epoch": 0.08746, + "grad_norm": 1.111182154226899, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 8746 + }, + { + "epoch": 0.08747, + "grad_norm": 1.0813261845403195, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 8747 + }, + { + "epoch": 0.08748, + "grad_norm": 1.2389683997355077, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 8748 + }, + { + "epoch": 0.08749, + "grad_norm": 1.1646233630246605, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 8749 + }, + { + "epoch": 0.0875, + "grad_norm": 0.9882478074471066, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 8750 + }, + { + "epoch": 0.08751, + "grad_norm": 1.0628828379350737, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 8751 + }, + { + "epoch": 0.08752, + "grad_norm": 1.0984158397543857, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 8752 + }, + { + "epoch": 0.08753, + "grad_norm": 1.2193128582910855, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 8753 + }, + { + "epoch": 0.08754, + "grad_norm": 1.2385678850122033, + "learning_rate": 0.003, + "loss": 4.054, + "step": 8754 + }, + { + "epoch": 0.08755, + "grad_norm": 0.9211225646462056, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8755 + }, + { + "epoch": 0.08756, + "grad_norm": 1.0770635030360587, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 8756 + }, + { + "epoch": 0.08757, + "grad_norm": 1.1930560657485945, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 8757 + }, + { + "epoch": 0.08758, + "grad_norm": 1.1384112467498324, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 8758 + }, + { + "epoch": 0.08759, + "grad_norm": 1.1172445551058783, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 8759 + }, + { + "epoch": 0.0876, + "grad_norm": 1.1801597872088998, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 8760 + }, + { + "epoch": 0.08761, + "grad_norm": 0.9352294073790094, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 8761 + }, + { + "epoch": 0.08762, + "grad_norm": 0.8823537867998336, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 8762 + }, + { + "epoch": 0.08763, + "grad_norm": 1.0898587675324372, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 8763 + }, + { + "epoch": 0.08764, + "grad_norm": 0.9778108833971553, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 8764 + }, + { + "epoch": 0.08765, + "grad_norm": 1.1127307572804943, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 8765 + }, + { + "epoch": 0.08766, + "grad_norm": 1.064542658962637, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 8766 + }, + { + "epoch": 0.08767, + "grad_norm": 1.2769881975725046, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 8767 + }, + { + "epoch": 0.08768, + "grad_norm": 0.9716458694094685, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 8768 + }, + { + "epoch": 0.08769, + "grad_norm": 1.02263370249674, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 8769 + }, + { + "epoch": 0.0877, + "grad_norm": 1.1197891339803154, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 8770 + }, + { + "epoch": 0.08771, + "grad_norm": 1.0343547006779303, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 8771 + }, + { + "epoch": 0.08772, + "grad_norm": 1.1391338162247877, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 8772 + }, + { + "epoch": 0.08773, + "grad_norm": 1.133183973465544, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 8773 + }, + { + "epoch": 0.08774, + "grad_norm": 0.9767687479388363, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 8774 + }, + { + "epoch": 0.08775, + "grad_norm": 1.0614747177631898, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 8775 + }, + { + "epoch": 0.08776, + "grad_norm": 1.2783088947042238, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 8776 + }, + { + "epoch": 0.08777, + "grad_norm": 0.7758444140217153, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 8777 + }, + { + "epoch": 0.08778, + "grad_norm": 0.933691111864922, + "learning_rate": 0.003, + "loss": 4.038, + "step": 8778 + }, + { + "epoch": 0.08779, + "grad_norm": 1.1241941614649351, + "learning_rate": 0.003, + "loss": 4.057, + "step": 8779 + }, + { + "epoch": 0.0878, + "grad_norm": 1.2838395411494257, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 8780 + }, + { + "epoch": 0.08781, + "grad_norm": 1.0245994272171686, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8781 + }, + { + "epoch": 0.08782, + "grad_norm": 1.166015780846121, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 8782 + }, + { + "epoch": 0.08783, + "grad_norm": 0.94136343088169, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8783 + }, + { + "epoch": 0.08784, + "grad_norm": 1.0537644571277438, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8784 + }, + { + "epoch": 0.08785, + "grad_norm": 1.188275684770546, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 8785 + }, + { + "epoch": 0.08786, + "grad_norm": 0.9863606748261747, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 8786 + }, + { + "epoch": 0.08787, + "grad_norm": 1.128286694821681, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 8787 + }, + { + "epoch": 0.08788, + "grad_norm": 1.079026401112349, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 8788 + }, + { + "epoch": 0.08789, + "grad_norm": 1.1880750581034827, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 8789 + }, + { + "epoch": 0.0879, + "grad_norm": 1.0176004388763833, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 8790 + }, + { + "epoch": 0.08791, + "grad_norm": 1.2038507485153178, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 8791 + }, + { + "epoch": 0.08792, + "grad_norm": 0.9980217290009189, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 8792 + }, + { + "epoch": 0.08793, + "grad_norm": 1.1566691894267236, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 8793 + }, + { + "epoch": 0.08794, + "grad_norm": 1.0311955693118802, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 8794 + }, + { + "epoch": 0.08795, + "grad_norm": 1.0735551697097647, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 8795 + }, + { + "epoch": 0.08796, + "grad_norm": 1.2673748855568097, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 8796 + }, + { + "epoch": 0.08797, + "grad_norm": 1.0868436946789142, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 8797 + }, + { + "epoch": 0.08798, + "grad_norm": 1.042119281795597, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 8798 + }, + { + "epoch": 0.08799, + "grad_norm": 1.0602674706510509, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 8799 + }, + { + "epoch": 0.088, + "grad_norm": 1.008077319870118, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 8800 + }, + { + "epoch": 0.08801, + "grad_norm": 1.0969522818507715, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 8801 + }, + { + "epoch": 0.08802, + "grad_norm": 1.1376221068560126, + "learning_rate": 0.003, + "loss": 4.052, + "step": 8802 + }, + { + "epoch": 0.08803, + "grad_norm": 1.0879978004992539, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 8803 + }, + { + "epoch": 0.08804, + "grad_norm": 1.1606225965715604, + "learning_rate": 0.003, + "loss": 4.025, + "step": 8804 + }, + { + "epoch": 0.08805, + "grad_norm": 0.966413249750066, + "learning_rate": 0.003, + "loss": 4.015, + "step": 8805 + }, + { + "epoch": 0.08806, + "grad_norm": 1.178861743093662, + "learning_rate": 0.003, + "loss": 4.031, + "step": 8806 + }, + { + "epoch": 0.08807, + "grad_norm": 0.9970083203505135, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 8807 + }, + { + "epoch": 0.08808, + "grad_norm": 1.148232319300287, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 8808 + }, + { + "epoch": 0.08809, + "grad_norm": 1.02006120805353, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8809 + }, + { + "epoch": 0.0881, + "grad_norm": 1.2541318447732, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 8810 + }, + { + "epoch": 0.08811, + "grad_norm": 0.9654813555045425, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 8811 + }, + { + "epoch": 0.08812, + "grad_norm": 1.0815748935370195, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 8812 + }, + { + "epoch": 0.08813, + "grad_norm": 1.2017717029030808, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 8813 + }, + { + "epoch": 0.08814, + "grad_norm": 1.061652224882425, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 8814 + }, + { + "epoch": 0.08815, + "grad_norm": 1.1941981502914418, + "learning_rate": 0.003, + "loss": 4.053, + "step": 8815 + }, + { + "epoch": 0.08816, + "grad_norm": 0.9722887512032639, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 8816 + }, + { + "epoch": 0.08817, + "grad_norm": 1.1137225093485594, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 8817 + }, + { + "epoch": 0.08818, + "grad_norm": 0.9632280469657819, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 8818 + }, + { + "epoch": 0.08819, + "grad_norm": 1.2912552481164463, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 8819 + }, + { + "epoch": 0.0882, + "grad_norm": 1.2656702522604, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 8820 + }, + { + "epoch": 0.08821, + "grad_norm": 0.8641185704105159, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 8821 + }, + { + "epoch": 0.08822, + "grad_norm": 0.8526858641972325, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 8822 + }, + { + "epoch": 0.08823, + "grad_norm": 0.9761138588500732, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 8823 + }, + { + "epoch": 0.08824, + "grad_norm": 0.9990171985622588, + "learning_rate": 0.003, + "loss": 4.032, + "step": 8824 + }, + { + "epoch": 0.08825, + "grad_norm": 0.9853141720864859, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 8825 + }, + { + "epoch": 0.08826, + "grad_norm": 1.216924716523876, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 8826 + }, + { + "epoch": 0.08827, + "grad_norm": 1.135457838710045, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8827 + }, + { + "epoch": 0.08828, + "grad_norm": 0.9724238442453864, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 8828 + }, + { + "epoch": 0.08829, + "grad_norm": 1.1262292498469912, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 8829 + }, + { + "epoch": 0.0883, + "grad_norm": 0.94574518398273, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 8830 + }, + { + "epoch": 0.08831, + "grad_norm": 1.1897119041270636, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 8831 + }, + { + "epoch": 0.08832, + "grad_norm": 1.2115168343495315, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 8832 + }, + { + "epoch": 0.08833, + "grad_norm": 0.996139762537626, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 8833 + }, + { + "epoch": 0.08834, + "grad_norm": 1.2368135937696196, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 8834 + }, + { + "epoch": 0.08835, + "grad_norm": 0.9570660017751258, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 8835 + }, + { + "epoch": 0.08836, + "grad_norm": 0.9995686259154789, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 8836 + }, + { + "epoch": 0.08837, + "grad_norm": 1.170370927533513, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 8837 + }, + { + "epoch": 0.08838, + "grad_norm": 1.0993199810585577, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 8838 + }, + { + "epoch": 0.08839, + "grad_norm": 1.258419211501553, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 8839 + }, + { + "epoch": 0.0884, + "grad_norm": 0.9274678150150035, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 8840 + }, + { + "epoch": 0.08841, + "grad_norm": 1.055913242044825, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 8841 + }, + { + "epoch": 0.08842, + "grad_norm": 1.1426746785347814, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 8842 + }, + { + "epoch": 0.08843, + "grad_norm": 1.1251979418413638, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 8843 + }, + { + "epoch": 0.08844, + "grad_norm": 1.0852468922230825, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 8844 + }, + { + "epoch": 0.08845, + "grad_norm": 1.1567649884610698, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 8845 + }, + { + "epoch": 0.08846, + "grad_norm": 1.082231327966706, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 8846 + }, + { + "epoch": 0.08847, + "grad_norm": 1.3745652698854023, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 8847 + }, + { + "epoch": 0.08848, + "grad_norm": 0.9014879193366324, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 8848 + }, + { + "epoch": 0.08849, + "grad_norm": 1.3017860095067897, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 8849 + }, + { + "epoch": 0.0885, + "grad_norm": 1.1398535241359604, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 8850 + }, + { + "epoch": 0.08851, + "grad_norm": 1.2652223839390064, + "learning_rate": 0.003, + "loss": 4.05, + "step": 8851 + }, + { + "epoch": 0.08852, + "grad_norm": 0.8569529658036033, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 8852 + }, + { + "epoch": 0.08853, + "grad_norm": 0.8417432899055782, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 8853 + }, + { + "epoch": 0.08854, + "grad_norm": 0.9651867070757068, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 8854 + }, + { + "epoch": 0.08855, + "grad_norm": 1.024221501922812, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 8855 + }, + { + "epoch": 0.08856, + "grad_norm": 1.1804109249837946, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 8856 + }, + { + "epoch": 0.08857, + "grad_norm": 0.955306519494686, + "learning_rate": 0.003, + "loss": 4.042, + "step": 8857 + }, + { + "epoch": 0.08858, + "grad_norm": 1.0294596526147064, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 8858 + }, + { + "epoch": 0.08859, + "grad_norm": 1.0845239322361329, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 8859 + }, + { + "epoch": 0.0886, + "grad_norm": 1.0515539677387107, + "learning_rate": 0.003, + "loss": 4.056, + "step": 8860 + }, + { + "epoch": 0.08861, + "grad_norm": 1.1312815390155653, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 8861 + }, + { + "epoch": 0.08862, + "grad_norm": 1.2225356916127106, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 8862 + }, + { + "epoch": 0.08863, + "grad_norm": 1.1065958969223766, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 8863 + }, + { + "epoch": 0.08864, + "grad_norm": 0.8686988551731697, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 8864 + }, + { + "epoch": 0.08865, + "grad_norm": 0.9406066773821836, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 8865 + }, + { + "epoch": 0.08866, + "grad_norm": 1.1068806392407193, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 8866 + }, + { + "epoch": 0.08867, + "grad_norm": 1.0109355114141556, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 8867 + }, + { + "epoch": 0.08868, + "grad_norm": 1.0297444159445588, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 8868 + }, + { + "epoch": 0.08869, + "grad_norm": 1.1859657375260197, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 8869 + }, + { + "epoch": 0.0887, + "grad_norm": 1.0367364996267392, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 8870 + }, + { + "epoch": 0.08871, + "grad_norm": 1.1770481681438267, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 8871 + }, + { + "epoch": 0.08872, + "grad_norm": 1.055437903261702, + "learning_rate": 0.003, + "loss": 4.073, + "step": 8872 + }, + { + "epoch": 0.08873, + "grad_norm": 1.1436383402758115, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 8873 + }, + { + "epoch": 0.08874, + "grad_norm": 1.0270645045752425, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 8874 + }, + { + "epoch": 0.08875, + "grad_norm": 1.2691814590014312, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 8875 + }, + { + "epoch": 0.08876, + "grad_norm": 1.0685113087608666, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 8876 + }, + { + "epoch": 0.08877, + "grad_norm": 0.9066843918350342, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 8877 + }, + { + "epoch": 0.08878, + "grad_norm": 1.1136428586139093, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 8878 + }, + { + "epoch": 0.08879, + "grad_norm": 1.1426232349637628, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 8879 + }, + { + "epoch": 0.0888, + "grad_norm": 0.9763661504928037, + "learning_rate": 0.003, + "loss": 4.022, + "step": 8880 + }, + { + "epoch": 0.08881, + "grad_norm": 0.9760789964859367, + "learning_rate": 0.003, + "loss": 4.01, + "step": 8881 + }, + { + "epoch": 0.08882, + "grad_norm": 0.993562866244837, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 8882 + }, + { + "epoch": 0.08883, + "grad_norm": 1.1608396003529295, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 8883 + }, + { + "epoch": 0.08884, + "grad_norm": 1.1323551569394956, + "learning_rate": 0.003, + "loss": 4.042, + "step": 8884 + }, + { + "epoch": 0.08885, + "grad_norm": 1.2349009939188973, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 8885 + }, + { + "epoch": 0.08886, + "grad_norm": 1.1914990238513254, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8886 + }, + { + "epoch": 0.08887, + "grad_norm": 1.0598046198071127, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 8887 + }, + { + "epoch": 0.08888, + "grad_norm": 1.1625466358037462, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8888 + }, + { + "epoch": 0.08889, + "grad_norm": 0.9052379335151111, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 8889 + }, + { + "epoch": 0.0889, + "grad_norm": 0.9819634809947653, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 8890 + }, + { + "epoch": 0.08891, + "grad_norm": 1.093718586313721, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 8891 + }, + { + "epoch": 0.08892, + "grad_norm": 1.213161423531245, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 8892 + }, + { + "epoch": 0.08893, + "grad_norm": 1.0485073867558885, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 8893 + }, + { + "epoch": 0.08894, + "grad_norm": 1.2824133272408857, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8894 + }, + { + "epoch": 0.08895, + "grad_norm": 0.8423439384851096, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 8895 + }, + { + "epoch": 0.08896, + "grad_norm": 1.0799448562495333, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 8896 + }, + { + "epoch": 0.08897, + "grad_norm": 1.4422293408188802, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 8897 + }, + { + "epoch": 0.08898, + "grad_norm": 1.032118344307113, + "learning_rate": 0.003, + "loss": 4.042, + "step": 8898 + }, + { + "epoch": 0.08899, + "grad_norm": 1.0522805801471267, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8899 + }, + { + "epoch": 0.089, + "grad_norm": 1.0679360633133137, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 8900 + }, + { + "epoch": 0.08901, + "grad_norm": 1.2000864304391685, + "learning_rate": 0.003, + "loss": 4.044, + "step": 8901 + }, + { + "epoch": 0.08902, + "grad_norm": 1.0038739754064978, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 8902 + }, + { + "epoch": 0.08903, + "grad_norm": 1.4580752514981392, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 8903 + }, + { + "epoch": 0.08904, + "grad_norm": 0.9501849947581935, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 8904 + }, + { + "epoch": 0.08905, + "grad_norm": 1.0865285261488837, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 8905 + }, + { + "epoch": 0.08906, + "grad_norm": 1.192560315216147, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 8906 + }, + { + "epoch": 0.08907, + "grad_norm": 0.86974732874746, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 8907 + }, + { + "epoch": 0.08908, + "grad_norm": 1.1193535519892115, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 8908 + }, + { + "epoch": 0.08909, + "grad_norm": 1.053123300357203, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8909 + }, + { + "epoch": 0.0891, + "grad_norm": 1.1339836887698285, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 8910 + }, + { + "epoch": 0.08911, + "grad_norm": 0.9277496841752985, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8911 + }, + { + "epoch": 0.08912, + "grad_norm": 0.9315539212514233, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8912 + }, + { + "epoch": 0.08913, + "grad_norm": 1.0576553940280595, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 8913 + }, + { + "epoch": 0.08914, + "grad_norm": 1.1640237850058512, + "learning_rate": 0.003, + "loss": 4.056, + "step": 8914 + }, + { + "epoch": 0.08915, + "grad_norm": 1.0188304936594577, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 8915 + }, + { + "epoch": 0.08916, + "grad_norm": 1.426269112438244, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 8916 + }, + { + "epoch": 0.08917, + "grad_norm": 1.0546818705564507, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 8917 + }, + { + "epoch": 0.08918, + "grad_norm": 1.2200135436078994, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 8918 + }, + { + "epoch": 0.08919, + "grad_norm": 0.8818184266787755, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 8919 + }, + { + "epoch": 0.0892, + "grad_norm": 0.9755990420691693, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 8920 + }, + { + "epoch": 0.08921, + "grad_norm": 1.2471123516060463, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 8921 + }, + { + "epoch": 0.08922, + "grad_norm": 0.8822858234291155, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 8922 + }, + { + "epoch": 0.08923, + "grad_norm": 0.9810038940078862, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 8923 + }, + { + "epoch": 0.08924, + "grad_norm": 1.2492114317483891, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 8924 + }, + { + "epoch": 0.08925, + "grad_norm": 0.9702532902239348, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8925 + }, + { + "epoch": 0.08926, + "grad_norm": 1.2387626568378134, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 8926 + }, + { + "epoch": 0.08927, + "grad_norm": 1.0513850982219397, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 8927 + }, + { + "epoch": 0.08928, + "grad_norm": 1.1575708697727565, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 8928 + }, + { + "epoch": 0.08929, + "grad_norm": 1.1942394751331307, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 8929 + }, + { + "epoch": 0.0893, + "grad_norm": 0.9099352819617809, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 8930 + }, + { + "epoch": 0.08931, + "grad_norm": 1.1574700090391592, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 8931 + }, + { + "epoch": 0.08932, + "grad_norm": 1.1712040522318599, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 8932 + }, + { + "epoch": 0.08933, + "grad_norm": 1.1298101748632317, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 8933 + }, + { + "epoch": 0.08934, + "grad_norm": 0.8845864177793066, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 8934 + }, + { + "epoch": 0.08935, + "grad_norm": 0.9728746053736606, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 8935 + }, + { + "epoch": 0.08936, + "grad_norm": 1.1064146202001022, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 8936 + }, + { + "epoch": 0.08937, + "grad_norm": 1.1728943500806064, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 8937 + }, + { + "epoch": 0.08938, + "grad_norm": 1.1024217274450334, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 8938 + }, + { + "epoch": 0.08939, + "grad_norm": 1.0804309602863862, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 8939 + }, + { + "epoch": 0.0894, + "grad_norm": 1.1974313951117248, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 8940 + }, + { + "epoch": 0.08941, + "grad_norm": 1.0472060105935805, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 8941 + }, + { + "epoch": 0.08942, + "grad_norm": 1.1849247191126862, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 8942 + }, + { + "epoch": 0.08943, + "grad_norm": 1.066833548679808, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8943 + }, + { + "epoch": 0.08944, + "grad_norm": 1.1269339808223942, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 8944 + }, + { + "epoch": 0.08945, + "grad_norm": 1.163904186867617, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 8945 + }, + { + "epoch": 0.08946, + "grad_norm": 1.1313006895031748, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 8946 + }, + { + "epoch": 0.08947, + "grad_norm": 1.2553855338780335, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 8947 + }, + { + "epoch": 0.08948, + "grad_norm": 1.1756270052552653, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 8948 + }, + { + "epoch": 0.08949, + "grad_norm": 0.9764941068571328, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 8949 + }, + { + "epoch": 0.0895, + "grad_norm": 1.0240642198990701, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 8950 + }, + { + "epoch": 0.08951, + "grad_norm": 1.2604446368244644, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 8951 + }, + { + "epoch": 0.08952, + "grad_norm": 1.069105761223158, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 8952 + }, + { + "epoch": 0.08953, + "grad_norm": 1.2897654616156728, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 8953 + }, + { + "epoch": 0.08954, + "grad_norm": 0.8927523209357381, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 8954 + }, + { + "epoch": 0.08955, + "grad_norm": 0.7994347598547118, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 8955 + }, + { + "epoch": 0.08956, + "grad_norm": 0.7721118918818628, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 8956 + }, + { + "epoch": 0.08957, + "grad_norm": 0.888735574661208, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 8957 + }, + { + "epoch": 0.08958, + "grad_norm": 1.011303363408731, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 8958 + }, + { + "epoch": 0.08959, + "grad_norm": 1.0580277710381667, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 8959 + }, + { + "epoch": 0.0896, + "grad_norm": 1.1815138015719915, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 8960 + }, + { + "epoch": 0.08961, + "grad_norm": 1.0763278082170449, + "learning_rate": 0.003, + "loss": 4.029, + "step": 8961 + }, + { + "epoch": 0.08962, + "grad_norm": 1.1547543247106453, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 8962 + }, + { + "epoch": 0.08963, + "grad_norm": 0.9474596852663282, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 8963 + }, + { + "epoch": 0.08964, + "grad_norm": 0.9994796332550057, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8964 + }, + { + "epoch": 0.08965, + "grad_norm": 1.0971047941927174, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 8965 + }, + { + "epoch": 0.08966, + "grad_norm": 0.9978749462040863, + "learning_rate": 0.003, + "loss": 4.018, + "step": 8966 + }, + { + "epoch": 0.08967, + "grad_norm": 1.091462772934639, + "learning_rate": 0.003, + "loss": 4.059, + "step": 8967 + }, + { + "epoch": 0.08968, + "grad_norm": 0.9928173167320807, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 8968 + }, + { + "epoch": 0.08969, + "grad_norm": 1.2786532555094048, + "learning_rate": 0.003, + "loss": 4.035, + "step": 8969 + }, + { + "epoch": 0.0897, + "grad_norm": 0.9304812924425294, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8970 + }, + { + "epoch": 0.08971, + "grad_norm": 1.029273303406794, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 8971 + }, + { + "epoch": 0.08972, + "grad_norm": 1.1442621209111814, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 8972 + }, + { + "epoch": 0.08973, + "grad_norm": 1.0166706811028725, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 8973 + }, + { + "epoch": 0.08974, + "grad_norm": 1.3160111354791268, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 8974 + }, + { + "epoch": 0.08975, + "grad_norm": 0.9964973979910691, + "learning_rate": 0.003, + "loss": 4.034, + "step": 8975 + }, + { + "epoch": 0.08976, + "grad_norm": 1.1506428299487088, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 8976 + }, + { + "epoch": 0.08977, + "grad_norm": 1.0330018963705825, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 8977 + }, + { + "epoch": 0.08978, + "grad_norm": 1.0887492374364582, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 8978 + }, + { + "epoch": 0.08979, + "grad_norm": 1.1346166358786325, + "learning_rate": 0.003, + "loss": 4.024, + "step": 8979 + }, + { + "epoch": 0.0898, + "grad_norm": 1.21619016375306, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8980 + }, + { + "epoch": 0.08981, + "grad_norm": 0.9851538869152199, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 8981 + }, + { + "epoch": 0.08982, + "grad_norm": 1.2126645439081598, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 8982 + }, + { + "epoch": 0.08983, + "grad_norm": 1.0447731466887387, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 8983 + }, + { + "epoch": 0.08984, + "grad_norm": 1.1497594183247177, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 8984 + }, + { + "epoch": 0.08985, + "grad_norm": 1.2483832316844437, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8985 + }, + { + "epoch": 0.08986, + "grad_norm": 1.2191689592626789, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 8986 + }, + { + "epoch": 0.08987, + "grad_norm": 1.076658986801492, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 8987 + }, + { + "epoch": 0.08988, + "grad_norm": 1.2479820133910688, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 8988 + }, + { + "epoch": 0.08989, + "grad_norm": 0.9034809767704942, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 8989 + }, + { + "epoch": 0.0899, + "grad_norm": 1.1404867965095085, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 8990 + }, + { + "epoch": 0.08991, + "grad_norm": 1.073000683119538, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 8991 + }, + { + "epoch": 0.08992, + "grad_norm": 1.2207306198770145, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 8992 + }, + { + "epoch": 0.08993, + "grad_norm": 1.0842166692368216, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 8993 + }, + { + "epoch": 0.08994, + "grad_norm": 1.0045211383021964, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 8994 + }, + { + "epoch": 0.08995, + "grad_norm": 1.316656487044687, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 8995 + }, + { + "epoch": 0.08996, + "grad_norm": 0.9870905235403588, + "learning_rate": 0.003, + "loss": 4.063, + "step": 8996 + }, + { + "epoch": 0.08997, + "grad_norm": 1.1131944813902874, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 8997 + }, + { + "epoch": 0.08998, + "grad_norm": 0.9575251947968014, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 8998 + }, + { + "epoch": 0.08999, + "grad_norm": 1.079155634238572, + "learning_rate": 0.003, + "loss": 4.048, + "step": 8999 + }, + { + "epoch": 0.09, + "grad_norm": 1.3073657014522342, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9000 + }, + { + "epoch": 0.09001, + "grad_norm": 0.9878617942632868, + "learning_rate": 0.003, + "loss": 4.036, + "step": 9001 + }, + { + "epoch": 0.09002, + "grad_norm": 1.4475702739335359, + "learning_rate": 0.003, + "loss": 4.036, + "step": 9002 + }, + { + "epoch": 0.09003, + "grad_norm": 1.0308588085625798, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 9003 + }, + { + "epoch": 0.09004, + "grad_norm": 1.221089293392804, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 9004 + }, + { + "epoch": 0.09005, + "grad_norm": 1.0231796835565534, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 9005 + }, + { + "epoch": 0.09006, + "grad_norm": 1.0417871456423533, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 9006 + }, + { + "epoch": 0.09007, + "grad_norm": 1.1191814093275658, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 9007 + }, + { + "epoch": 0.09008, + "grad_norm": 1.1348119020549952, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 9008 + }, + { + "epoch": 0.09009, + "grad_norm": 1.1633559345033404, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 9009 + }, + { + "epoch": 0.0901, + "grad_norm": 1.1458799262789814, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9010 + }, + { + "epoch": 0.09011, + "grad_norm": 1.0636561316185342, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 9011 + }, + { + "epoch": 0.09012, + "grad_norm": 1.1477528884261912, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9012 + }, + { + "epoch": 0.09013, + "grad_norm": 1.0045238113278767, + "learning_rate": 0.003, + "loss": 4.027, + "step": 9013 + }, + { + "epoch": 0.09014, + "grad_norm": 1.1254428234025435, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 9014 + }, + { + "epoch": 0.09015, + "grad_norm": 1.1052529228915646, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 9015 + }, + { + "epoch": 0.09016, + "grad_norm": 1.0143709342989635, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 9016 + }, + { + "epoch": 0.09017, + "grad_norm": 1.0995658886429716, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 9017 + }, + { + "epoch": 0.09018, + "grad_norm": 1.0029225082537132, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 9018 + }, + { + "epoch": 0.09019, + "grad_norm": 1.3935530723752598, + "learning_rate": 0.003, + "loss": 4.035, + "step": 9019 + }, + { + "epoch": 0.0902, + "grad_norm": 0.9681733776693029, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9020 + }, + { + "epoch": 0.09021, + "grad_norm": 1.3687701542155535, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 9021 + }, + { + "epoch": 0.09022, + "grad_norm": 0.9332102242731483, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 9022 + }, + { + "epoch": 0.09023, + "grad_norm": 1.0483328137620929, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 9023 + }, + { + "epoch": 0.09024, + "grad_norm": 1.1150169635102571, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 9024 + }, + { + "epoch": 0.09025, + "grad_norm": 1.2002769998310732, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 9025 + }, + { + "epoch": 0.09026, + "grad_norm": 1.2011034969162588, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 9026 + }, + { + "epoch": 0.09027, + "grad_norm": 1.146578370959906, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 9027 + }, + { + "epoch": 0.09028, + "grad_norm": 1.2228846480536117, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 9028 + }, + { + "epoch": 0.09029, + "grad_norm": 1.037224821411794, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 9029 + }, + { + "epoch": 0.0903, + "grad_norm": 0.9844125119573323, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 9030 + }, + { + "epoch": 0.09031, + "grad_norm": 1.1495286013814523, + "learning_rate": 0.003, + "loss": 4.022, + "step": 9031 + }, + { + "epoch": 0.09032, + "grad_norm": 1.1742914278318812, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 9032 + }, + { + "epoch": 0.09033, + "grad_norm": 1.3904421332901014, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 9033 + }, + { + "epoch": 0.09034, + "grad_norm": 1.0278814470733673, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 9034 + }, + { + "epoch": 0.09035, + "grad_norm": 1.1229673655711763, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 9035 + }, + { + "epoch": 0.09036, + "grad_norm": 0.9056875520752794, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 9036 + }, + { + "epoch": 0.09037, + "grad_norm": 1.1006054177092766, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 9037 + }, + { + "epoch": 0.09038, + "grad_norm": 1.1348621040993052, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 9038 + }, + { + "epoch": 0.09039, + "grad_norm": 1.05381794931122, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 9039 + }, + { + "epoch": 0.0904, + "grad_norm": 1.0109885109102685, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 9040 + }, + { + "epoch": 0.09041, + "grad_norm": 1.123512949286327, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 9041 + }, + { + "epoch": 0.09042, + "grad_norm": 1.0917985051699848, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 9042 + }, + { + "epoch": 0.09043, + "grad_norm": 1.084941390137943, + "learning_rate": 0.003, + "loss": 4.027, + "step": 9043 + }, + { + "epoch": 0.09044, + "grad_norm": 1.12950639714153, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 9044 + }, + { + "epoch": 0.09045, + "grad_norm": 0.9435189761788078, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 9045 + }, + { + "epoch": 0.09046, + "grad_norm": 1.1205216961300428, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 9046 + }, + { + "epoch": 0.09047, + "grad_norm": 1.1101023596152269, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 9047 + }, + { + "epoch": 0.09048, + "grad_norm": 1.4711338068553876, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 9048 + }, + { + "epoch": 0.09049, + "grad_norm": 0.7958927509366164, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 9049 + }, + { + "epoch": 0.0905, + "grad_norm": 0.8137554584809007, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 9050 + }, + { + "epoch": 0.09051, + "grad_norm": 0.9025372956999882, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 9051 + }, + { + "epoch": 0.09052, + "grad_norm": 0.8871827348097391, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 9052 + }, + { + "epoch": 0.09053, + "grad_norm": 1.0593382418378432, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 9053 + }, + { + "epoch": 0.09054, + "grad_norm": 1.185043130945484, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 9054 + }, + { + "epoch": 0.09055, + "grad_norm": 1.123987691026757, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 9055 + }, + { + "epoch": 0.09056, + "grad_norm": 1.0852908797471614, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 9056 + }, + { + "epoch": 0.09057, + "grad_norm": 0.9184881520876066, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 9057 + }, + { + "epoch": 0.09058, + "grad_norm": 0.8821076958966486, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 9058 + }, + { + "epoch": 0.09059, + "grad_norm": 0.9653080945803115, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 9059 + }, + { + "epoch": 0.0906, + "grad_norm": 1.157246811985134, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 9060 + }, + { + "epoch": 0.09061, + "grad_norm": 1.106358395927521, + "learning_rate": 0.003, + "loss": 4.042, + "step": 9061 + }, + { + "epoch": 0.09062, + "grad_norm": 1.1308906707844582, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 9062 + }, + { + "epoch": 0.09063, + "grad_norm": 1.3615134157248625, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 9063 + }, + { + "epoch": 0.09064, + "grad_norm": 0.9632769176971355, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 9064 + }, + { + "epoch": 0.09065, + "grad_norm": 1.193055845335768, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 9065 + }, + { + "epoch": 0.09066, + "grad_norm": 1.0217847841679015, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 9066 + }, + { + "epoch": 0.09067, + "grad_norm": 1.2708025422078024, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 9067 + }, + { + "epoch": 0.09068, + "grad_norm": 0.999091796379082, + "learning_rate": 0.003, + "loss": 4.046, + "step": 9068 + }, + { + "epoch": 0.09069, + "grad_norm": 1.1937745015052195, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 9069 + }, + { + "epoch": 0.0907, + "grad_norm": 1.2267119910601545, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 9070 + }, + { + "epoch": 0.09071, + "grad_norm": 0.9307012025024503, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 9071 + }, + { + "epoch": 0.09072, + "grad_norm": 1.0537322841979997, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 9072 + }, + { + "epoch": 0.09073, + "grad_norm": 1.2016861411462987, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 9073 + }, + { + "epoch": 0.09074, + "grad_norm": 0.8496237355918259, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 9074 + }, + { + "epoch": 0.09075, + "grad_norm": 0.9908939634552881, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 9075 + }, + { + "epoch": 0.09076, + "grad_norm": 1.1968746368032812, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 9076 + }, + { + "epoch": 0.09077, + "grad_norm": 0.970351786121277, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 9077 + }, + { + "epoch": 0.09078, + "grad_norm": 1.1497763383817403, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 9078 + }, + { + "epoch": 0.09079, + "grad_norm": 1.0830961717463246, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 9079 + }, + { + "epoch": 0.0908, + "grad_norm": 0.9609418328913767, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 9080 + }, + { + "epoch": 0.09081, + "grad_norm": 1.1000854820573318, + "learning_rate": 0.003, + "loss": 4.051, + "step": 9081 + }, + { + "epoch": 0.09082, + "grad_norm": 1.1636952655382402, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9082 + }, + { + "epoch": 0.09083, + "grad_norm": 1.1420961957835405, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 9083 + }, + { + "epoch": 0.09084, + "grad_norm": 1.0553826184749373, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 9084 + }, + { + "epoch": 0.09085, + "grad_norm": 1.0255413615584303, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 9085 + }, + { + "epoch": 0.09086, + "grad_norm": 1.215452301963506, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 9086 + }, + { + "epoch": 0.09087, + "grad_norm": 0.9583487655084203, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 9087 + }, + { + "epoch": 0.09088, + "grad_norm": 1.232091628428844, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 9088 + }, + { + "epoch": 0.09089, + "grad_norm": 1.1137319600932658, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 9089 + }, + { + "epoch": 0.0909, + "grad_norm": 1.144755645285606, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 9090 + }, + { + "epoch": 0.09091, + "grad_norm": 1.0300971496763878, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 9091 + }, + { + "epoch": 0.09092, + "grad_norm": 1.0100476409907833, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 9092 + }, + { + "epoch": 0.09093, + "grad_norm": 1.1494952695266714, + "learning_rate": 0.003, + "loss": 4.032, + "step": 9093 + }, + { + "epoch": 0.09094, + "grad_norm": 1.0095676827058278, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9094 + }, + { + "epoch": 0.09095, + "grad_norm": 1.2093302711410523, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 9095 + }, + { + "epoch": 0.09096, + "grad_norm": 0.976858163705672, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 9096 + }, + { + "epoch": 0.09097, + "grad_norm": 1.3853264618148677, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 9097 + }, + { + "epoch": 0.09098, + "grad_norm": 1.0766280332726217, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 9098 + }, + { + "epoch": 0.09099, + "grad_norm": 1.1136120589991296, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 9099 + }, + { + "epoch": 0.091, + "grad_norm": 1.1601021677903895, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 9100 + }, + { + "epoch": 0.09101, + "grad_norm": 1.3417195527068442, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 9101 + }, + { + "epoch": 0.09102, + "grad_norm": 0.9070301102163859, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9102 + }, + { + "epoch": 0.09103, + "grad_norm": 0.9573335281707542, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 9103 + }, + { + "epoch": 0.09104, + "grad_norm": 1.0556076158321603, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 9104 + }, + { + "epoch": 0.09105, + "grad_norm": 1.1979973249595275, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 9105 + }, + { + "epoch": 0.09106, + "grad_norm": 1.1589318950983944, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 9106 + }, + { + "epoch": 0.09107, + "grad_norm": 1.1814885851810335, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 9107 + }, + { + "epoch": 0.09108, + "grad_norm": 1.019986448191847, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 9108 + }, + { + "epoch": 0.09109, + "grad_norm": 1.057663996654355, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 9109 + }, + { + "epoch": 0.0911, + "grad_norm": 1.2256223665855766, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 9110 + }, + { + "epoch": 0.09111, + "grad_norm": 0.9461029364575244, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 9111 + }, + { + "epoch": 0.09112, + "grad_norm": 1.049429275804046, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 9112 + }, + { + "epoch": 0.09113, + "grad_norm": 1.031956399076363, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 9113 + }, + { + "epoch": 0.09114, + "grad_norm": 1.0174397153586041, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 9114 + }, + { + "epoch": 0.09115, + "grad_norm": 1.2745379192778026, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 9115 + }, + { + "epoch": 0.09116, + "grad_norm": 0.9837529257881423, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 9116 + }, + { + "epoch": 0.09117, + "grad_norm": 1.1786562942934684, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 9117 + }, + { + "epoch": 0.09118, + "grad_norm": 1.0560853969394606, + "learning_rate": 0.003, + "loss": 4.037, + "step": 9118 + }, + { + "epoch": 0.09119, + "grad_norm": 1.0377188929161931, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9119 + }, + { + "epoch": 0.0912, + "grad_norm": 1.1207049150632644, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 9120 + }, + { + "epoch": 0.09121, + "grad_norm": 1.0193731745757568, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 9121 + }, + { + "epoch": 0.09122, + "grad_norm": 1.2624197726748643, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 9122 + }, + { + "epoch": 0.09123, + "grad_norm": 1.072056196669728, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 9123 + }, + { + "epoch": 0.09124, + "grad_norm": 1.239441100721789, + "learning_rate": 0.003, + "loss": 4.039, + "step": 9124 + }, + { + "epoch": 0.09125, + "grad_norm": 1.083356005410815, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 9125 + }, + { + "epoch": 0.09126, + "grad_norm": 1.266014219278627, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 9126 + }, + { + "epoch": 0.09127, + "grad_norm": 0.8598933039930851, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 9127 + }, + { + "epoch": 0.09128, + "grad_norm": 0.89093357601972, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 9128 + }, + { + "epoch": 0.09129, + "grad_norm": 1.11080170345324, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9129 + }, + { + "epoch": 0.0913, + "grad_norm": 1.1921225096267718, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 9130 + }, + { + "epoch": 0.09131, + "grad_norm": 0.9911325172513188, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 9131 + }, + { + "epoch": 0.09132, + "grad_norm": 1.210452388671294, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 9132 + }, + { + "epoch": 0.09133, + "grad_norm": 1.1006247067475978, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 9133 + }, + { + "epoch": 0.09134, + "grad_norm": 1.192482017694382, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 9134 + }, + { + "epoch": 0.09135, + "grad_norm": 1.050325724178467, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 9135 + }, + { + "epoch": 0.09136, + "grad_norm": 1.2143867555226864, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 9136 + }, + { + "epoch": 0.09137, + "grad_norm": 0.9760810890066407, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 9137 + }, + { + "epoch": 0.09138, + "grad_norm": 1.1888855045602282, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 9138 + }, + { + "epoch": 0.09139, + "grad_norm": 0.9681065989709288, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 9139 + }, + { + "epoch": 0.0914, + "grad_norm": 1.3038699850066278, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9140 + }, + { + "epoch": 0.09141, + "grad_norm": 0.9119873610357526, + "learning_rate": 0.003, + "loss": 4.069, + "step": 9141 + }, + { + "epoch": 0.09142, + "grad_norm": 0.97753456917582, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 9142 + }, + { + "epoch": 0.09143, + "grad_norm": 1.2011099485139207, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 9143 + }, + { + "epoch": 0.09144, + "grad_norm": 0.8653559838348887, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 9144 + }, + { + "epoch": 0.09145, + "grad_norm": 0.6804719677612727, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 9145 + }, + { + "epoch": 0.09146, + "grad_norm": 0.8281509343825025, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 9146 + }, + { + "epoch": 0.09147, + "grad_norm": 1.046932924628029, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 9147 + }, + { + "epoch": 0.09148, + "grad_norm": 0.9481937690438011, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 9148 + }, + { + "epoch": 0.09149, + "grad_norm": 1.1467040624244416, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 9149 + }, + { + "epoch": 0.0915, + "grad_norm": 1.2688544713513863, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 9150 + }, + { + "epoch": 0.09151, + "grad_norm": 0.9410854425311685, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 9151 + }, + { + "epoch": 0.09152, + "grad_norm": 1.0715331038388722, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 9152 + }, + { + "epoch": 0.09153, + "grad_norm": 1.011570665232525, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 9153 + }, + { + "epoch": 0.09154, + "grad_norm": 1.0783280541516371, + "learning_rate": 0.003, + "loss": 4.048, + "step": 9154 + }, + { + "epoch": 0.09155, + "grad_norm": 1.0855343965090731, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 9155 + }, + { + "epoch": 0.09156, + "grad_norm": 1.0564142389909517, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 9156 + }, + { + "epoch": 0.09157, + "grad_norm": 1.0945895233993794, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 9157 + }, + { + "epoch": 0.09158, + "grad_norm": 1.336449148127849, + "learning_rate": 0.003, + "loss": 4.035, + "step": 9158 + }, + { + "epoch": 0.09159, + "grad_norm": 0.991375084782573, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 9159 + }, + { + "epoch": 0.0916, + "grad_norm": 1.0591224251327636, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 9160 + }, + { + "epoch": 0.09161, + "grad_norm": 1.164704576534139, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 9161 + }, + { + "epoch": 0.09162, + "grad_norm": 0.9413438539519919, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 9162 + }, + { + "epoch": 0.09163, + "grad_norm": 0.9714944791493136, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 9163 + }, + { + "epoch": 0.09164, + "grad_norm": 1.0287059172603712, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 9164 + }, + { + "epoch": 0.09165, + "grad_norm": 1.1861303886211818, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 9165 + }, + { + "epoch": 0.09166, + "grad_norm": 1.124122240608982, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 9166 + }, + { + "epoch": 0.09167, + "grad_norm": 1.1459093885614982, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 9167 + }, + { + "epoch": 0.09168, + "grad_norm": 1.119778165627705, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 9168 + }, + { + "epoch": 0.09169, + "grad_norm": 1.1111751785567634, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 9169 + }, + { + "epoch": 0.0917, + "grad_norm": 1.2226524744645328, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 9170 + }, + { + "epoch": 0.09171, + "grad_norm": 0.8656526632807879, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 9171 + }, + { + "epoch": 0.09172, + "grad_norm": 1.099267531571083, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 9172 + }, + { + "epoch": 0.09173, + "grad_norm": 1.1162934002071323, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 9173 + }, + { + "epoch": 0.09174, + "grad_norm": 1.1000518721341896, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 9174 + }, + { + "epoch": 0.09175, + "grad_norm": 1.267658430539927, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 9175 + }, + { + "epoch": 0.09176, + "grad_norm": 1.0571564301320642, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 9176 + }, + { + "epoch": 0.09177, + "grad_norm": 1.1069807499260333, + "learning_rate": 0.003, + "loss": 4.002, + "step": 9177 + }, + { + "epoch": 0.09178, + "grad_norm": 1.0241191541435508, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 9178 + }, + { + "epoch": 0.09179, + "grad_norm": 1.2215349660778567, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 9179 + }, + { + "epoch": 0.0918, + "grad_norm": 0.9063283480014228, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 9180 + }, + { + "epoch": 0.09181, + "grad_norm": 0.8608588303075517, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 9181 + }, + { + "epoch": 0.09182, + "grad_norm": 1.080015607354298, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 9182 + }, + { + "epoch": 0.09183, + "grad_norm": 1.187987053928514, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 9183 + }, + { + "epoch": 0.09184, + "grad_norm": 1.0370353123734928, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 9184 + }, + { + "epoch": 0.09185, + "grad_norm": 1.1348969491194556, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 9185 + }, + { + "epoch": 0.09186, + "grad_norm": 1.2036726155070432, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 9186 + }, + { + "epoch": 0.09187, + "grad_norm": 1.0871963616696243, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 9187 + }, + { + "epoch": 0.09188, + "grad_norm": 1.1723509075908214, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9188 + }, + { + "epoch": 0.09189, + "grad_norm": 1.1132456681131127, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 9189 + }, + { + "epoch": 0.0919, + "grad_norm": 1.044542911903216, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 9190 + }, + { + "epoch": 0.09191, + "grad_norm": 1.2056609382294916, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 9191 + }, + { + "epoch": 0.09192, + "grad_norm": 1.0206520860594175, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 9192 + }, + { + "epoch": 0.09193, + "grad_norm": 1.1522627595760395, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 9193 + }, + { + "epoch": 0.09194, + "grad_norm": 1.036571230879693, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 9194 + }, + { + "epoch": 0.09195, + "grad_norm": 1.156868074376868, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 9195 + }, + { + "epoch": 0.09196, + "grad_norm": 0.9632802843738769, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 9196 + }, + { + "epoch": 0.09197, + "grad_norm": 1.2366512084187664, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9197 + }, + { + "epoch": 0.09198, + "grad_norm": 0.9807141780820184, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 9198 + }, + { + "epoch": 0.09199, + "grad_norm": 1.0704002219425641, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 9199 + }, + { + "epoch": 0.092, + "grad_norm": 1.1110145741263948, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 9200 + }, + { + "epoch": 0.09201, + "grad_norm": 1.090835930160533, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 9201 + }, + { + "epoch": 0.09202, + "grad_norm": 1.0317362468150382, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 9202 + }, + { + "epoch": 0.09203, + "grad_norm": 1.2267741986981442, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9203 + }, + { + "epoch": 0.09204, + "grad_norm": 1.1325105553589676, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 9204 + }, + { + "epoch": 0.09205, + "grad_norm": 1.1638105186655754, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 9205 + }, + { + "epoch": 0.09206, + "grad_norm": 1.0692805133374446, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 9206 + }, + { + "epoch": 0.09207, + "grad_norm": 1.175551262927089, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 9207 + }, + { + "epoch": 0.09208, + "grad_norm": 1.0644580288738281, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 9208 + }, + { + "epoch": 0.09209, + "grad_norm": 0.9664436871225908, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 9209 + }, + { + "epoch": 0.0921, + "grad_norm": 1.1254680317199708, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 9210 + }, + { + "epoch": 0.09211, + "grad_norm": 1.200712594293563, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 9211 + }, + { + "epoch": 0.09212, + "grad_norm": 0.9683431414690981, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 9212 + }, + { + "epoch": 0.09213, + "grad_norm": 0.944968780879836, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 9213 + }, + { + "epoch": 0.09214, + "grad_norm": 1.0413526215293398, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 9214 + }, + { + "epoch": 0.09215, + "grad_norm": 1.1772555410236571, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 9215 + }, + { + "epoch": 0.09216, + "grad_norm": 0.9624083137396001, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 9216 + }, + { + "epoch": 0.09217, + "grad_norm": 1.0511031168385416, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 9217 + }, + { + "epoch": 0.09218, + "grad_norm": 1.3269317543115233, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 9218 + }, + { + "epoch": 0.09219, + "grad_norm": 0.9923317369891772, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 9219 + }, + { + "epoch": 0.0922, + "grad_norm": 1.073030298407734, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 9220 + }, + { + "epoch": 0.09221, + "grad_norm": 1.0446285156707587, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 9221 + }, + { + "epoch": 0.09222, + "grad_norm": 1.1434810207183548, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9222 + }, + { + "epoch": 0.09223, + "grad_norm": 1.1025201940993825, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 9223 + }, + { + "epoch": 0.09224, + "grad_norm": 0.9566021361855169, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 9224 + }, + { + "epoch": 0.09225, + "grad_norm": 1.2323077112756355, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 9225 + }, + { + "epoch": 0.09226, + "grad_norm": 0.899900247722213, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 9226 + }, + { + "epoch": 0.09227, + "grad_norm": 1.0849893705736053, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 9227 + }, + { + "epoch": 0.09228, + "grad_norm": 1.2542611304862052, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 9228 + }, + { + "epoch": 0.09229, + "grad_norm": 0.8407437326641111, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 9229 + }, + { + "epoch": 0.0923, + "grad_norm": 1.158954671915461, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 9230 + }, + { + "epoch": 0.09231, + "grad_norm": 1.1704792802830555, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 9231 + }, + { + "epoch": 0.09232, + "grad_norm": 1.1029452862218423, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 9232 + }, + { + "epoch": 0.09233, + "grad_norm": 1.2002856333009415, + "learning_rate": 0.003, + "loss": 4.06, + "step": 9233 + }, + { + "epoch": 0.09234, + "grad_norm": 0.8750251942831692, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 9234 + }, + { + "epoch": 0.09235, + "grad_norm": 1.0716615659635735, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 9235 + }, + { + "epoch": 0.09236, + "grad_norm": 1.5813090203643494, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 9236 + }, + { + "epoch": 0.09237, + "grad_norm": 1.0338774462879339, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 9237 + }, + { + "epoch": 0.09238, + "grad_norm": 1.135346462252927, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 9238 + }, + { + "epoch": 0.09239, + "grad_norm": 1.1394764093548577, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 9239 + }, + { + "epoch": 0.0924, + "grad_norm": 1.1428354439905772, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 9240 + }, + { + "epoch": 0.09241, + "grad_norm": 1.0203493481299601, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 9241 + }, + { + "epoch": 0.09242, + "grad_norm": 0.9451602651492341, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 9242 + }, + { + "epoch": 0.09243, + "grad_norm": 1.0754642967463526, + "learning_rate": 0.003, + "loss": 4.053, + "step": 9243 + }, + { + "epoch": 0.09244, + "grad_norm": 1.2105994420477935, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 9244 + }, + { + "epoch": 0.09245, + "grad_norm": 1.1177071678503647, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 9245 + }, + { + "epoch": 0.09246, + "grad_norm": 1.1262767459012595, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 9246 + }, + { + "epoch": 0.09247, + "grad_norm": 1.1101169385517686, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 9247 + }, + { + "epoch": 0.09248, + "grad_norm": 1.039107535835496, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 9248 + }, + { + "epoch": 0.09249, + "grad_norm": 1.051375380829021, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 9249 + }, + { + "epoch": 0.0925, + "grad_norm": 1.0818027535404036, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 9250 + }, + { + "epoch": 0.09251, + "grad_norm": 1.21441854307002, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 9251 + }, + { + "epoch": 0.09252, + "grad_norm": 0.9446590679589214, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 9252 + }, + { + "epoch": 0.09253, + "grad_norm": 0.9688539640875207, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 9253 + }, + { + "epoch": 0.09254, + "grad_norm": 1.1291024219182868, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 9254 + }, + { + "epoch": 0.09255, + "grad_norm": 1.211862634432545, + "learning_rate": 0.003, + "loss": 4.016, + "step": 9255 + }, + { + "epoch": 0.09256, + "grad_norm": 1.2218560976542767, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 9256 + }, + { + "epoch": 0.09257, + "grad_norm": 1.0643221369589915, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 9257 + }, + { + "epoch": 0.09258, + "grad_norm": 1.1527043491418376, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 9258 + }, + { + "epoch": 0.09259, + "grad_norm": 0.9926788558086477, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 9259 + }, + { + "epoch": 0.0926, + "grad_norm": 1.0433542223940273, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 9260 + }, + { + "epoch": 0.09261, + "grad_norm": 1.207804917706118, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 9261 + }, + { + "epoch": 0.09262, + "grad_norm": 1.1324209587635743, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 9262 + }, + { + "epoch": 0.09263, + "grad_norm": 1.172369074851875, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 9263 + }, + { + "epoch": 0.09264, + "grad_norm": 1.1682667843729466, + "learning_rate": 0.003, + "loss": 4.073, + "step": 9264 + }, + { + "epoch": 0.09265, + "grad_norm": 1.253090916375967, + "learning_rate": 0.003, + "loss": 4.045, + "step": 9265 + }, + { + "epoch": 0.09266, + "grad_norm": 1.1781887620358225, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 9266 + }, + { + "epoch": 0.09267, + "grad_norm": 1.0284374738702509, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 9267 + }, + { + "epoch": 0.09268, + "grad_norm": 1.1665787828723402, + "learning_rate": 0.003, + "loss": 4.027, + "step": 9268 + }, + { + "epoch": 0.09269, + "grad_norm": 1.1264162171369585, + "learning_rate": 0.003, + "loss": 4.034, + "step": 9269 + }, + { + "epoch": 0.0927, + "grad_norm": 1.1949132258262425, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 9270 + }, + { + "epoch": 0.09271, + "grad_norm": 0.9696066482483353, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 9271 + }, + { + "epoch": 0.09272, + "grad_norm": 1.0301396773994387, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 9272 + }, + { + "epoch": 0.09273, + "grad_norm": 1.0893059409180028, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 9273 + }, + { + "epoch": 0.09274, + "grad_norm": 1.0947192726893542, + "learning_rate": 0.003, + "loss": 4.037, + "step": 9274 + }, + { + "epoch": 0.09275, + "grad_norm": 1.0723410183602133, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 9275 + }, + { + "epoch": 0.09276, + "grad_norm": 1.1691696906210551, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9276 + }, + { + "epoch": 0.09277, + "grad_norm": 1.0647822469371042, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 9277 + }, + { + "epoch": 0.09278, + "grad_norm": 1.2167469789912233, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 9278 + }, + { + "epoch": 0.09279, + "grad_norm": 0.9896695614661197, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 9279 + }, + { + "epoch": 0.0928, + "grad_norm": 1.3634170937343293, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9280 + }, + { + "epoch": 0.09281, + "grad_norm": 0.8620980137769152, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 9281 + }, + { + "epoch": 0.09282, + "grad_norm": 1.0427330391887681, + "learning_rate": 0.003, + "loss": 4.069, + "step": 9282 + }, + { + "epoch": 0.09283, + "grad_norm": 1.1710140561731026, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 9283 + }, + { + "epoch": 0.09284, + "grad_norm": 1.4120083965462074, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 9284 + }, + { + "epoch": 0.09285, + "grad_norm": 1.0005243949124853, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9285 + }, + { + "epoch": 0.09286, + "grad_norm": 1.2076261862310649, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 9286 + }, + { + "epoch": 0.09287, + "grad_norm": 0.9611404585275731, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9287 + }, + { + "epoch": 0.09288, + "grad_norm": 0.9896962192199752, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 9288 + }, + { + "epoch": 0.09289, + "grad_norm": 1.1918897636488415, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 9289 + }, + { + "epoch": 0.0929, + "grad_norm": 1.100511638945172, + "learning_rate": 0.003, + "loss": 4.045, + "step": 9290 + }, + { + "epoch": 0.09291, + "grad_norm": 0.9953764094766955, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 9291 + }, + { + "epoch": 0.09292, + "grad_norm": 0.9663852236515597, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 9292 + }, + { + "epoch": 0.09293, + "grad_norm": 1.0635777370604826, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 9293 + }, + { + "epoch": 0.09294, + "grad_norm": 1.2456008375763008, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 9294 + }, + { + "epoch": 0.09295, + "grad_norm": 0.8653819364739406, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 9295 + }, + { + "epoch": 0.09296, + "grad_norm": 1.1680264346725173, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 9296 + }, + { + "epoch": 0.09297, + "grad_norm": 1.1901508293685317, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 9297 + }, + { + "epoch": 0.09298, + "grad_norm": 1.0915606863010314, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 9298 + }, + { + "epoch": 0.09299, + "grad_norm": 1.1592108148961227, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 9299 + }, + { + "epoch": 0.093, + "grad_norm": 1.0708244671033804, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 9300 + }, + { + "epoch": 0.09301, + "grad_norm": 1.1033798817954517, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 9301 + }, + { + "epoch": 0.09302, + "grad_norm": 1.0088796578839236, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 9302 + }, + { + "epoch": 0.09303, + "grad_norm": 1.2858796878215408, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 9303 + }, + { + "epoch": 0.09304, + "grad_norm": 1.2506744639450122, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 9304 + }, + { + "epoch": 0.09305, + "grad_norm": 0.9961386724217615, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 9305 + }, + { + "epoch": 0.09306, + "grad_norm": 1.2149212174491897, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 9306 + }, + { + "epoch": 0.09307, + "grad_norm": 0.9136578619084738, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 9307 + }, + { + "epoch": 0.09308, + "grad_norm": 0.9588826151722423, + "learning_rate": 0.003, + "loss": 4.018, + "step": 9308 + }, + { + "epoch": 0.09309, + "grad_norm": 0.957182512469483, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 9309 + }, + { + "epoch": 0.0931, + "grad_norm": 1.2935022678548451, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 9310 + }, + { + "epoch": 0.09311, + "grad_norm": 0.9790928444088314, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 9311 + }, + { + "epoch": 0.09312, + "grad_norm": 1.150857565224749, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 9312 + }, + { + "epoch": 0.09313, + "grad_norm": 1.075107605301709, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 9313 + }, + { + "epoch": 0.09314, + "grad_norm": 1.1182712741118235, + "learning_rate": 0.003, + "loss": 4.036, + "step": 9314 + }, + { + "epoch": 0.09315, + "grad_norm": 1.2529463174245064, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 9315 + }, + { + "epoch": 0.09316, + "grad_norm": 0.9771774475983117, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 9316 + }, + { + "epoch": 0.09317, + "grad_norm": 1.3760618473803496, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 9317 + }, + { + "epoch": 0.09318, + "grad_norm": 0.7429498285526682, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 9318 + }, + { + "epoch": 0.09319, + "grad_norm": 0.9792252701782946, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 9319 + }, + { + "epoch": 0.0932, + "grad_norm": 1.4975639708650046, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 9320 + }, + { + "epoch": 0.09321, + "grad_norm": 0.8902587907079172, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 9321 + }, + { + "epoch": 0.09322, + "grad_norm": 0.9093500134430933, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 9322 + }, + { + "epoch": 0.09323, + "grad_norm": 1.0993796211404194, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 9323 + }, + { + "epoch": 0.09324, + "grad_norm": 1.160680899752396, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 9324 + }, + { + "epoch": 0.09325, + "grad_norm": 0.9385567636610579, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 9325 + }, + { + "epoch": 0.09326, + "grad_norm": 0.9621843602737219, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 9326 + }, + { + "epoch": 0.09327, + "grad_norm": 1.2283368544933888, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 9327 + }, + { + "epoch": 0.09328, + "grad_norm": 1.138118155085037, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 9328 + }, + { + "epoch": 0.09329, + "grad_norm": 1.3360636810202806, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 9329 + }, + { + "epoch": 0.0933, + "grad_norm": 0.997995204861989, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 9330 + }, + { + "epoch": 0.09331, + "grad_norm": 1.1201516817051664, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 9331 + }, + { + "epoch": 0.09332, + "grad_norm": 1.204525418131745, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 9332 + }, + { + "epoch": 0.09333, + "grad_norm": 1.0575312400096248, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 9333 + }, + { + "epoch": 0.09334, + "grad_norm": 1.2701572053232315, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 9334 + }, + { + "epoch": 0.09335, + "grad_norm": 1.1585842605178802, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 9335 + }, + { + "epoch": 0.09336, + "grad_norm": 1.0311564658307473, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 9336 + }, + { + "epoch": 0.09337, + "grad_norm": 0.9367345134695763, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 9337 + }, + { + "epoch": 0.09338, + "grad_norm": 0.9632251089955324, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 9338 + }, + { + "epoch": 0.09339, + "grad_norm": 1.0383186326297353, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 9339 + }, + { + "epoch": 0.0934, + "grad_norm": 1.086317533403553, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 9340 + }, + { + "epoch": 0.09341, + "grad_norm": 1.066313152390867, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 9341 + }, + { + "epoch": 0.09342, + "grad_norm": 1.2845490767261778, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 9342 + }, + { + "epoch": 0.09343, + "grad_norm": 1.200116094561525, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 9343 + }, + { + "epoch": 0.09344, + "grad_norm": 1.1152409513420605, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 9344 + }, + { + "epoch": 0.09345, + "grad_norm": 1.1107749483966356, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 9345 + }, + { + "epoch": 0.09346, + "grad_norm": 1.1283130554525544, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9346 + }, + { + "epoch": 0.09347, + "grad_norm": 1.0736955106456132, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 9347 + }, + { + "epoch": 0.09348, + "grad_norm": 1.0736643024406574, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9348 + }, + { + "epoch": 0.09349, + "grad_norm": 1.1989810783719919, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 9349 + }, + { + "epoch": 0.0935, + "grad_norm": 1.0740480563955976, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 9350 + }, + { + "epoch": 0.09351, + "grad_norm": 0.9996894673701714, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 9351 + }, + { + "epoch": 0.09352, + "grad_norm": 1.2492290188517632, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 9352 + }, + { + "epoch": 0.09353, + "grad_norm": 1.0984353858675837, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 9353 + }, + { + "epoch": 0.09354, + "grad_norm": 1.0494227900731494, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 9354 + }, + { + "epoch": 0.09355, + "grad_norm": 1.0887626536039114, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 9355 + }, + { + "epoch": 0.09356, + "grad_norm": 1.1063458069116743, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 9356 + }, + { + "epoch": 0.09357, + "grad_norm": 1.1772620726746457, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 9357 + }, + { + "epoch": 0.09358, + "grad_norm": 0.8349208876816591, + "learning_rate": 0.003, + "loss": 4.035, + "step": 9358 + }, + { + "epoch": 0.09359, + "grad_norm": 0.8649691894837728, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 9359 + }, + { + "epoch": 0.0936, + "grad_norm": 0.9462358195573604, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 9360 + }, + { + "epoch": 0.09361, + "grad_norm": 1.233886976766547, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 9361 + }, + { + "epoch": 0.09362, + "grad_norm": 1.2087100800252286, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 9362 + }, + { + "epoch": 0.09363, + "grad_norm": 0.9584661416221599, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 9363 + }, + { + "epoch": 0.09364, + "grad_norm": 1.3630115362152988, + "learning_rate": 0.003, + "loss": 4.037, + "step": 9364 + }, + { + "epoch": 0.09365, + "grad_norm": 1.0115250367617443, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 9365 + }, + { + "epoch": 0.09366, + "grad_norm": 1.410143423428131, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 9366 + }, + { + "epoch": 0.09367, + "grad_norm": 1.0026644890567558, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9367 + }, + { + "epoch": 0.09368, + "grad_norm": 1.2102005850211326, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 9368 + }, + { + "epoch": 0.09369, + "grad_norm": 1.1309040815939808, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 9369 + }, + { + "epoch": 0.0937, + "grad_norm": 0.9973045348674824, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 9370 + }, + { + "epoch": 0.09371, + "grad_norm": 1.231370608565574, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 9371 + }, + { + "epoch": 0.09372, + "grad_norm": 1.0924523718241692, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 9372 + }, + { + "epoch": 0.09373, + "grad_norm": 1.0504664280616463, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 9373 + }, + { + "epoch": 0.09374, + "grad_norm": 1.1272188425084024, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 9374 + }, + { + "epoch": 0.09375, + "grad_norm": 1.1348460508047482, + "learning_rate": 0.003, + "loss": 4.018, + "step": 9375 + }, + { + "epoch": 0.09376, + "grad_norm": 1.1298959210406103, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 9376 + }, + { + "epoch": 0.09377, + "grad_norm": 1.186524901763047, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9377 + }, + { + "epoch": 0.09378, + "grad_norm": 1.0314318127210171, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9378 + }, + { + "epoch": 0.09379, + "grad_norm": 1.5680911633070103, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 9379 + }, + { + "epoch": 0.0938, + "grad_norm": 0.9013424550963559, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 9380 + }, + { + "epoch": 0.09381, + "grad_norm": 0.9574099783748187, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 9381 + }, + { + "epoch": 0.09382, + "grad_norm": 1.2086515825142305, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 9382 + }, + { + "epoch": 0.09383, + "grad_norm": 1.0350405673887666, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9383 + }, + { + "epoch": 0.09384, + "grad_norm": 1.0680807378772312, + "learning_rate": 0.003, + "loss": 4.061, + "step": 9384 + }, + { + "epoch": 0.09385, + "grad_norm": 1.0961146064007061, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 9385 + }, + { + "epoch": 0.09386, + "grad_norm": 0.9545455608465853, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 9386 + }, + { + "epoch": 0.09387, + "grad_norm": 0.9765112894095569, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 9387 + }, + { + "epoch": 0.09388, + "grad_norm": 1.0782991672647486, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 9388 + }, + { + "epoch": 0.09389, + "grad_norm": 1.0731997920493421, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 9389 + }, + { + "epoch": 0.0939, + "grad_norm": 1.2364650097930605, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 9390 + }, + { + "epoch": 0.09391, + "grad_norm": 1.0663640185817809, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 9391 + }, + { + "epoch": 0.09392, + "grad_norm": 1.1877737611342285, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 9392 + }, + { + "epoch": 0.09393, + "grad_norm": 1.0955254115743633, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 9393 + }, + { + "epoch": 0.09394, + "grad_norm": 1.4092921791527784, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 9394 + }, + { + "epoch": 0.09395, + "grad_norm": 0.9358321234864352, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 9395 + }, + { + "epoch": 0.09396, + "grad_norm": 1.2594661199138193, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 9396 + }, + { + "epoch": 0.09397, + "grad_norm": 0.9883524181089541, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 9397 + }, + { + "epoch": 0.09398, + "grad_norm": 1.0673063130537923, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 9398 + }, + { + "epoch": 0.09399, + "grad_norm": 1.1875700898134038, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 9399 + }, + { + "epoch": 0.094, + "grad_norm": 1.0418263510990502, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 9400 + }, + { + "epoch": 0.09401, + "grad_norm": 1.1642465528112664, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 9401 + }, + { + "epoch": 0.09402, + "grad_norm": 1.2786047972483314, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 9402 + }, + { + "epoch": 0.09403, + "grad_norm": 0.8618399669597347, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 9403 + }, + { + "epoch": 0.09404, + "grad_norm": 1.102847094348384, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 9404 + }, + { + "epoch": 0.09405, + "grad_norm": 1.2834464451783154, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 9405 + }, + { + "epoch": 0.09406, + "grad_norm": 0.9871554834956486, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 9406 + }, + { + "epoch": 0.09407, + "grad_norm": 1.1610654126184679, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9407 + }, + { + "epoch": 0.09408, + "grad_norm": 1.170917230941906, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 9408 + }, + { + "epoch": 0.09409, + "grad_norm": 1.1318228712154021, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 9409 + }, + { + "epoch": 0.0941, + "grad_norm": 1.1234390873721771, + "learning_rate": 0.003, + "loss": 4.058, + "step": 9410 + }, + { + "epoch": 0.09411, + "grad_norm": 0.8647601508974965, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 9411 + }, + { + "epoch": 0.09412, + "grad_norm": 1.0471840870012292, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 9412 + }, + { + "epoch": 0.09413, + "grad_norm": 1.2190310827593032, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 9413 + }, + { + "epoch": 0.09414, + "grad_norm": 0.8477403736289396, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 9414 + }, + { + "epoch": 0.09415, + "grad_norm": 0.9371919042158918, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 9415 + }, + { + "epoch": 0.09416, + "grad_norm": 1.1490720658755587, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 9416 + }, + { + "epoch": 0.09417, + "grad_norm": 1.1018851741635174, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 9417 + }, + { + "epoch": 0.09418, + "grad_norm": 1.2892671070061477, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 9418 + }, + { + "epoch": 0.09419, + "grad_norm": 1.1101253491606295, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 9419 + }, + { + "epoch": 0.0942, + "grad_norm": 1.3071470583049012, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 9420 + }, + { + "epoch": 0.09421, + "grad_norm": 0.863117611253664, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 9421 + }, + { + "epoch": 0.09422, + "grad_norm": 1.0738994904421413, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 9422 + }, + { + "epoch": 0.09423, + "grad_norm": 1.1740757656465912, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 9423 + }, + { + "epoch": 0.09424, + "grad_norm": 0.9999480734746848, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9424 + }, + { + "epoch": 0.09425, + "grad_norm": 1.1947521810325221, + "learning_rate": 0.003, + "loss": 4.059, + "step": 9425 + }, + { + "epoch": 0.09426, + "grad_norm": 1.076053138054749, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 9426 + }, + { + "epoch": 0.09427, + "grad_norm": 1.2539611359358045, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 9427 + }, + { + "epoch": 0.09428, + "grad_norm": 1.2042161531904478, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 9428 + }, + { + "epoch": 0.09429, + "grad_norm": 1.2520918442905509, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 9429 + }, + { + "epoch": 0.0943, + "grad_norm": 0.8090989376412021, + "learning_rate": 0.003, + "loss": 4.034, + "step": 9430 + }, + { + "epoch": 0.09431, + "grad_norm": 0.9501739667787066, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 9431 + }, + { + "epoch": 0.09432, + "grad_norm": 1.0564545187269092, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 9432 + }, + { + "epoch": 0.09433, + "grad_norm": 1.1105124478676938, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 9433 + }, + { + "epoch": 0.09434, + "grad_norm": 1.1365861267709172, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 9434 + }, + { + "epoch": 0.09435, + "grad_norm": 1.0626647857526774, + "learning_rate": 0.003, + "loss": 4.031, + "step": 9435 + }, + { + "epoch": 0.09436, + "grad_norm": 1.3048334209489638, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 9436 + }, + { + "epoch": 0.09437, + "grad_norm": 1.192258709260609, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 9437 + }, + { + "epoch": 0.09438, + "grad_norm": 1.141173762966999, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 9438 + }, + { + "epoch": 0.09439, + "grad_norm": 0.9959914893100231, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 9439 + }, + { + "epoch": 0.0944, + "grad_norm": 1.1812893575099583, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 9440 + }, + { + "epoch": 0.09441, + "grad_norm": 0.9472244329470745, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 9441 + }, + { + "epoch": 0.09442, + "grad_norm": 1.1629963337515048, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 9442 + }, + { + "epoch": 0.09443, + "grad_norm": 0.8487426150195475, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 9443 + }, + { + "epoch": 0.09444, + "grad_norm": 0.9207154747911219, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 9444 + }, + { + "epoch": 0.09445, + "grad_norm": 1.115602354686498, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 9445 + }, + { + "epoch": 0.09446, + "grad_norm": 1.3343852809989445, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 9446 + }, + { + "epoch": 0.09447, + "grad_norm": 0.9863278032640667, + "learning_rate": 0.003, + "loss": 4.046, + "step": 9447 + }, + { + "epoch": 0.09448, + "grad_norm": 1.1763365235364014, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 9448 + }, + { + "epoch": 0.09449, + "grad_norm": 1.105703389589132, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 9449 + }, + { + "epoch": 0.0945, + "grad_norm": 1.01551147072131, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 9450 + }, + { + "epoch": 0.09451, + "grad_norm": 1.0383763855846733, + "learning_rate": 0.003, + "loss": 4.045, + "step": 9451 + }, + { + "epoch": 0.09452, + "grad_norm": 1.0456254521002109, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 9452 + }, + { + "epoch": 0.09453, + "grad_norm": 1.3691070796802907, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 9453 + }, + { + "epoch": 0.09454, + "grad_norm": 0.9155459999489047, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 9454 + }, + { + "epoch": 0.09455, + "grad_norm": 1.0062151085426871, + "learning_rate": 0.003, + "loss": 4.032, + "step": 9455 + }, + { + "epoch": 0.09456, + "grad_norm": 1.775196393766159, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 9456 + }, + { + "epoch": 0.09457, + "grad_norm": 0.8268537533403297, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9457 + }, + { + "epoch": 0.09458, + "grad_norm": 0.9446756093027355, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 9458 + }, + { + "epoch": 0.09459, + "grad_norm": 1.2872893558697978, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 9459 + }, + { + "epoch": 0.0946, + "grad_norm": 1.0414735406519482, + "learning_rate": 0.003, + "loss": 4.064, + "step": 9460 + }, + { + "epoch": 0.09461, + "grad_norm": 1.2923952872869107, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 9461 + }, + { + "epoch": 0.09462, + "grad_norm": 1.0982135647456706, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 9462 + }, + { + "epoch": 0.09463, + "grad_norm": 1.195985441134548, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9463 + }, + { + "epoch": 0.09464, + "grad_norm": 1.094216310472828, + "learning_rate": 0.003, + "loss": 4.032, + "step": 9464 + }, + { + "epoch": 0.09465, + "grad_norm": 1.1344022479380613, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 9465 + }, + { + "epoch": 0.09466, + "grad_norm": 1.1649272025527642, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 9466 + }, + { + "epoch": 0.09467, + "grad_norm": 1.1437197614588612, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 9467 + }, + { + "epoch": 0.09468, + "grad_norm": 1.134510156394556, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 9468 + }, + { + "epoch": 0.09469, + "grad_norm": 1.1456897293699813, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 9469 + }, + { + "epoch": 0.0947, + "grad_norm": 0.9370646196560669, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 9470 + }, + { + "epoch": 0.09471, + "grad_norm": 1.1605962963311105, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 9471 + }, + { + "epoch": 0.09472, + "grad_norm": 1.0220332905341656, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 9472 + }, + { + "epoch": 0.09473, + "grad_norm": 1.3119106941290712, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 9473 + }, + { + "epoch": 0.09474, + "grad_norm": 1.0696197454349894, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 9474 + }, + { + "epoch": 0.09475, + "grad_norm": 1.183557234231492, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 9475 + }, + { + "epoch": 0.09476, + "grad_norm": 0.936355154816999, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 9476 + }, + { + "epoch": 0.09477, + "grad_norm": 1.0826345363697292, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 9477 + }, + { + "epoch": 0.09478, + "grad_norm": 1.2268375814601145, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 9478 + }, + { + "epoch": 0.09479, + "grad_norm": 0.9077481557382702, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9479 + }, + { + "epoch": 0.0948, + "grad_norm": 1.1577127173889232, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 9480 + }, + { + "epoch": 0.09481, + "grad_norm": 1.164940507783553, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 9481 + }, + { + "epoch": 0.09482, + "grad_norm": 1.0376186580785653, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 9482 + }, + { + "epoch": 0.09483, + "grad_norm": 1.0668111334949657, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 9483 + }, + { + "epoch": 0.09484, + "grad_norm": 0.9544000956467396, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 9484 + }, + { + "epoch": 0.09485, + "grad_norm": 1.3563985618454755, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 9485 + }, + { + "epoch": 0.09486, + "grad_norm": 0.9907132939829558, + "learning_rate": 0.003, + "loss": 4.046, + "step": 9486 + }, + { + "epoch": 0.09487, + "grad_norm": 1.140513002856383, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 9487 + }, + { + "epoch": 0.09488, + "grad_norm": 1.2992557947553103, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 9488 + }, + { + "epoch": 0.09489, + "grad_norm": 0.9910288087212725, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 9489 + }, + { + "epoch": 0.0949, + "grad_norm": 1.281341022406897, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 9490 + }, + { + "epoch": 0.09491, + "grad_norm": 1.0178254732217256, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 9491 + }, + { + "epoch": 0.09492, + "grad_norm": 1.0599592630101493, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 9492 + }, + { + "epoch": 0.09493, + "grad_norm": 1.0707052087473112, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 9493 + }, + { + "epoch": 0.09494, + "grad_norm": 1.184811169391914, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 9494 + }, + { + "epoch": 0.09495, + "grad_norm": 1.1050306365464493, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 9495 + }, + { + "epoch": 0.09496, + "grad_norm": 1.060159425305823, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 9496 + }, + { + "epoch": 0.09497, + "grad_norm": 1.0825941688780625, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 9497 + }, + { + "epoch": 0.09498, + "grad_norm": 0.9601571426627881, + "learning_rate": 0.003, + "loss": 4.011, + "step": 9498 + }, + { + "epoch": 0.09499, + "grad_norm": 1.0590750166016245, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 9499 + }, + { + "epoch": 0.095, + "grad_norm": 1.2786584082845458, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 9500 + }, + { + "epoch": 0.09501, + "grad_norm": 1.1290326342601904, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 9501 + }, + { + "epoch": 0.09502, + "grad_norm": 0.9912457750329704, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9502 + }, + { + "epoch": 0.09503, + "grad_norm": 1.152658098138242, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 9503 + }, + { + "epoch": 0.09504, + "grad_norm": 1.0919690729103646, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 9504 + }, + { + "epoch": 0.09505, + "grad_norm": 1.1519487718748778, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 9505 + }, + { + "epoch": 0.09506, + "grad_norm": 1.262006574899147, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 9506 + }, + { + "epoch": 0.09507, + "grad_norm": 1.0711373212381228, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 9507 + }, + { + "epoch": 0.09508, + "grad_norm": 1.1239192938184404, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 9508 + }, + { + "epoch": 0.09509, + "grad_norm": 1.3389515718167913, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 9509 + }, + { + "epoch": 0.0951, + "grad_norm": 0.8958060198263552, + "learning_rate": 0.003, + "loss": 4.036, + "step": 9510 + }, + { + "epoch": 0.09511, + "grad_norm": 1.0719956101779906, + "learning_rate": 0.003, + "loss": 4.026, + "step": 9511 + }, + { + "epoch": 0.09512, + "grad_norm": 1.305626481744099, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 9512 + }, + { + "epoch": 0.09513, + "grad_norm": 0.9176719274352512, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 9513 + }, + { + "epoch": 0.09514, + "grad_norm": 0.9499238409052401, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 9514 + }, + { + "epoch": 0.09515, + "grad_norm": 1.063622459645422, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9515 + }, + { + "epoch": 0.09516, + "grad_norm": 1.1652384446120476, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 9516 + }, + { + "epoch": 0.09517, + "grad_norm": 1.178061476075233, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 9517 + }, + { + "epoch": 0.09518, + "grad_norm": 1.3678980044323894, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 9518 + }, + { + "epoch": 0.09519, + "grad_norm": 1.046217095028117, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 9519 + }, + { + "epoch": 0.0952, + "grad_norm": 1.1818557546526918, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 9520 + }, + { + "epoch": 0.09521, + "grad_norm": 1.1304356171833927, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 9521 + }, + { + "epoch": 0.09522, + "grad_norm": 1.120019603471895, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 9522 + }, + { + "epoch": 0.09523, + "grad_norm": 1.104879448193623, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9523 + }, + { + "epoch": 0.09524, + "grad_norm": 0.873368086596089, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 9524 + }, + { + "epoch": 0.09525, + "grad_norm": 0.9961985148368994, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 9525 + }, + { + "epoch": 0.09526, + "grad_norm": 1.2676375654165073, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 9526 + }, + { + "epoch": 0.09527, + "grad_norm": 1.1529943279598172, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 9527 + }, + { + "epoch": 0.09528, + "grad_norm": 1.131678617939542, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 9528 + }, + { + "epoch": 0.09529, + "grad_norm": 1.0147106663564123, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 9529 + }, + { + "epoch": 0.0953, + "grad_norm": 1.2878762896528653, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 9530 + }, + { + "epoch": 0.09531, + "grad_norm": 1.0096287930269905, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 9531 + }, + { + "epoch": 0.09532, + "grad_norm": 1.278489534695974, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 9532 + }, + { + "epoch": 0.09533, + "grad_norm": 1.068314907618256, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 9533 + }, + { + "epoch": 0.09534, + "grad_norm": 1.0541105462925286, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 9534 + }, + { + "epoch": 0.09535, + "grad_norm": 1.0542416896874893, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 9535 + }, + { + "epoch": 0.09536, + "grad_norm": 1.203990932473719, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 9536 + }, + { + "epoch": 0.09537, + "grad_norm": 1.0609650874854466, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 9537 + }, + { + "epoch": 0.09538, + "grad_norm": 1.0566099485690075, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 9538 + }, + { + "epoch": 0.09539, + "grad_norm": 1.1711495828610736, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 9539 + }, + { + "epoch": 0.0954, + "grad_norm": 1.0370146448146371, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 9540 + }, + { + "epoch": 0.09541, + "grad_norm": 1.0159923428799353, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 9541 + }, + { + "epoch": 0.09542, + "grad_norm": 1.1293172537885205, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 9542 + }, + { + "epoch": 0.09543, + "grad_norm": 1.0450117640491212, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 9543 + }, + { + "epoch": 0.09544, + "grad_norm": 1.1603566723882397, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 9544 + }, + { + "epoch": 0.09545, + "grad_norm": 1.0447933808065626, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 9545 + }, + { + "epoch": 0.09546, + "grad_norm": 0.9458484110007981, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 9546 + }, + { + "epoch": 0.09547, + "grad_norm": 1.0876570915875363, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 9547 + }, + { + "epoch": 0.09548, + "grad_norm": 1.1875507966969954, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9548 + }, + { + "epoch": 0.09549, + "grad_norm": 0.8814604626466929, + "learning_rate": 0.003, + "loss": 4.063, + "step": 9549 + }, + { + "epoch": 0.0955, + "grad_norm": 1.0246586128840154, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 9550 + }, + { + "epoch": 0.09551, + "grad_norm": 1.2387660010338577, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 9551 + }, + { + "epoch": 0.09552, + "grad_norm": 0.9754198125538579, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 9552 + }, + { + "epoch": 0.09553, + "grad_norm": 1.1577414898591938, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 9553 + }, + { + "epoch": 0.09554, + "grad_norm": 1.0570088924704193, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 9554 + }, + { + "epoch": 0.09555, + "grad_norm": 1.2106521594790243, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 9555 + }, + { + "epoch": 0.09556, + "grad_norm": 1.0733450373625477, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 9556 + }, + { + "epoch": 0.09557, + "grad_norm": 1.1919871564102347, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 9557 + }, + { + "epoch": 0.09558, + "grad_norm": 1.146997350654266, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 9558 + }, + { + "epoch": 0.09559, + "grad_norm": 1.168927128415256, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 9559 + }, + { + "epoch": 0.0956, + "grad_norm": 1.11230556582203, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 9560 + }, + { + "epoch": 0.09561, + "grad_norm": 1.0881575661118905, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 9561 + }, + { + "epoch": 0.09562, + "grad_norm": 1.3389741513893663, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 9562 + }, + { + "epoch": 0.09563, + "grad_norm": 1.1779078269635281, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 9563 + }, + { + "epoch": 0.09564, + "grad_norm": 1.2633874376703453, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 9564 + }, + { + "epoch": 0.09565, + "grad_norm": 0.8461819044779642, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 9565 + }, + { + "epoch": 0.09566, + "grad_norm": 0.9430171254625663, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 9566 + }, + { + "epoch": 0.09567, + "grad_norm": 1.0543925802682974, + "learning_rate": 0.003, + "loss": 4.033, + "step": 9567 + }, + { + "epoch": 0.09568, + "grad_norm": 1.2147511761418746, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 9568 + }, + { + "epoch": 0.09569, + "grad_norm": 1.139812511049965, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 9569 + }, + { + "epoch": 0.0957, + "grad_norm": 1.2066618266931726, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 9570 + }, + { + "epoch": 0.09571, + "grad_norm": 0.9969088728201041, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 9571 + }, + { + "epoch": 0.09572, + "grad_norm": 1.521564608038327, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 9572 + }, + { + "epoch": 0.09573, + "grad_norm": 0.767016893061818, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 9573 + }, + { + "epoch": 0.09574, + "grad_norm": 0.9077964583623702, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 9574 + }, + { + "epoch": 0.09575, + "grad_norm": 0.9789511548929908, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 9575 + }, + { + "epoch": 0.09576, + "grad_norm": 1.1272087126973815, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 9576 + }, + { + "epoch": 0.09577, + "grad_norm": 1.2010384957358553, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 9577 + }, + { + "epoch": 0.09578, + "grad_norm": 1.1758915609790765, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 9578 + }, + { + "epoch": 0.09579, + "grad_norm": 1.1677535297765045, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 9579 + }, + { + "epoch": 0.0958, + "grad_norm": 1.1752106864410679, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 9580 + }, + { + "epoch": 0.09581, + "grad_norm": 1.0390685668953878, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 9581 + }, + { + "epoch": 0.09582, + "grad_norm": 0.9448991333037483, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 9582 + }, + { + "epoch": 0.09583, + "grad_norm": 1.0699012101296939, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 9583 + }, + { + "epoch": 0.09584, + "grad_norm": 1.1114200647456693, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9584 + }, + { + "epoch": 0.09585, + "grad_norm": 1.1110522535404246, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 9585 + }, + { + "epoch": 0.09586, + "grad_norm": 1.0693682178529131, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 9586 + }, + { + "epoch": 0.09587, + "grad_norm": 1.0907022200921574, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 9587 + }, + { + "epoch": 0.09588, + "grad_norm": 1.3052673040023657, + "learning_rate": 0.003, + "loss": 4.037, + "step": 9588 + }, + { + "epoch": 0.09589, + "grad_norm": 1.1273974285859623, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 9589 + }, + { + "epoch": 0.0959, + "grad_norm": 0.9575377458395997, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 9590 + }, + { + "epoch": 0.09591, + "grad_norm": 1.2408320836823072, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 9591 + }, + { + "epoch": 0.09592, + "grad_norm": 1.0230179861054507, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 9592 + }, + { + "epoch": 0.09593, + "grad_norm": 1.2007397484572089, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 9593 + }, + { + "epoch": 0.09594, + "grad_norm": 0.9589259749600783, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 9594 + }, + { + "epoch": 0.09595, + "grad_norm": 1.0812271670511917, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 9595 + }, + { + "epoch": 0.09596, + "grad_norm": 1.215276423755771, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 9596 + }, + { + "epoch": 0.09597, + "grad_norm": 1.055514060891504, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 9597 + }, + { + "epoch": 0.09598, + "grad_norm": 1.0847458195628048, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 9598 + }, + { + "epoch": 0.09599, + "grad_norm": 1.3389087793122878, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 9599 + }, + { + "epoch": 0.096, + "grad_norm": 0.9055383982261854, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 9600 + }, + { + "epoch": 0.09601, + "grad_norm": 0.9811940695731026, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 9601 + }, + { + "epoch": 0.09602, + "grad_norm": 1.0789527786902395, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 9602 + }, + { + "epoch": 0.09603, + "grad_norm": 1.1727478141730425, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 9603 + }, + { + "epoch": 0.09604, + "grad_norm": 1.023524788242411, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 9604 + }, + { + "epoch": 0.09605, + "grad_norm": 1.2858580097981922, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 9605 + }, + { + "epoch": 0.09606, + "grad_norm": 0.8845738717710074, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 9606 + }, + { + "epoch": 0.09607, + "grad_norm": 0.9417694564288732, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 9607 + }, + { + "epoch": 0.09608, + "grad_norm": 1.328516456493561, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 9608 + }, + { + "epoch": 0.09609, + "grad_norm": 1.2629011203498186, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 9609 + }, + { + "epoch": 0.0961, + "grad_norm": 1.1729883232988318, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 9610 + }, + { + "epoch": 0.09611, + "grad_norm": 1.393118460283264, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 9611 + }, + { + "epoch": 0.09612, + "grad_norm": 1.0698096910097765, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 9612 + }, + { + "epoch": 0.09613, + "grad_norm": 1.007696738311522, + "learning_rate": 0.003, + "loss": 4.051, + "step": 9613 + }, + { + "epoch": 0.09614, + "grad_norm": 1.4803986138018956, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9614 + }, + { + "epoch": 0.09615, + "grad_norm": 1.0559026629823391, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 9615 + }, + { + "epoch": 0.09616, + "grad_norm": 1.1782897391468559, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9616 + }, + { + "epoch": 0.09617, + "grad_norm": 1.0127744559386622, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 9617 + }, + { + "epoch": 0.09618, + "grad_norm": 1.2681610401673449, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 9618 + }, + { + "epoch": 0.09619, + "grad_norm": 0.9294594817912581, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 9619 + }, + { + "epoch": 0.0962, + "grad_norm": 0.9905671912694861, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 9620 + }, + { + "epoch": 0.09621, + "grad_norm": 1.3373852398607728, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 9621 + }, + { + "epoch": 0.09622, + "grad_norm": 1.1541351242021758, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 9622 + }, + { + "epoch": 0.09623, + "grad_norm": 1.112301616615713, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 9623 + }, + { + "epoch": 0.09624, + "grad_norm": 1.3430937174739355, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 9624 + }, + { + "epoch": 0.09625, + "grad_norm": 1.080047687143114, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 9625 + }, + { + "epoch": 0.09626, + "grad_norm": 1.0694079912518233, + "learning_rate": 0.003, + "loss": 4.03, + "step": 9626 + }, + { + "epoch": 0.09627, + "grad_norm": 1.139432118757624, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 9627 + }, + { + "epoch": 0.09628, + "grad_norm": 0.958627843592087, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 9628 + }, + { + "epoch": 0.09629, + "grad_norm": 1.0662067967300446, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 9629 + }, + { + "epoch": 0.0963, + "grad_norm": 1.0412238973539465, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 9630 + }, + { + "epoch": 0.09631, + "grad_norm": 1.0125046957428312, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 9631 + }, + { + "epoch": 0.09632, + "grad_norm": 1.3555424752709917, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 9632 + }, + { + "epoch": 0.09633, + "grad_norm": 1.063757740872899, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 9633 + }, + { + "epoch": 0.09634, + "grad_norm": 1.20136927846276, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 9634 + }, + { + "epoch": 0.09635, + "grad_norm": 1.1423655529057926, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 9635 + }, + { + "epoch": 0.09636, + "grad_norm": 1.3277021817787018, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 9636 + }, + { + "epoch": 0.09637, + "grad_norm": 0.916236476287278, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 9637 + }, + { + "epoch": 0.09638, + "grad_norm": 1.1712812625089448, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 9638 + }, + { + "epoch": 0.09639, + "grad_norm": 1.1016378675492433, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 9639 + }, + { + "epoch": 0.0964, + "grad_norm": 1.0341626312817283, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 9640 + }, + { + "epoch": 0.09641, + "grad_norm": 1.1660677439533391, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 9641 + }, + { + "epoch": 0.09642, + "grad_norm": 0.9807879097387026, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 9642 + }, + { + "epoch": 0.09643, + "grad_norm": 1.2769541125682597, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 9643 + }, + { + "epoch": 0.09644, + "grad_norm": 0.9284721817809056, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 9644 + }, + { + "epoch": 0.09645, + "grad_norm": 1.2692913801981347, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9645 + }, + { + "epoch": 0.09646, + "grad_norm": 0.9226038318267903, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 9646 + }, + { + "epoch": 0.09647, + "grad_norm": 1.1716846999519674, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 9647 + }, + { + "epoch": 0.09648, + "grad_norm": 1.0984624781628638, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 9648 + }, + { + "epoch": 0.09649, + "grad_norm": 1.032713451324886, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 9649 + }, + { + "epoch": 0.0965, + "grad_norm": 1.2893129243430874, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 9650 + }, + { + "epoch": 0.09651, + "grad_norm": 1.2914240519466096, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 9651 + }, + { + "epoch": 0.09652, + "grad_norm": 1.0304038397431154, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 9652 + }, + { + "epoch": 0.09653, + "grad_norm": 1.1558865874046644, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 9653 + }, + { + "epoch": 0.09654, + "grad_norm": 1.0769712394558766, + "learning_rate": 0.003, + "loss": 4.013, + "step": 9654 + }, + { + "epoch": 0.09655, + "grad_norm": 1.0046687966997174, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 9655 + }, + { + "epoch": 0.09656, + "grad_norm": 1.208629121263327, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 9656 + }, + { + "epoch": 0.09657, + "grad_norm": 1.0524663213041456, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 9657 + }, + { + "epoch": 0.09658, + "grad_norm": 1.1446885277300836, + "learning_rate": 0.003, + "loss": 4.031, + "step": 9658 + }, + { + "epoch": 0.09659, + "grad_norm": 1.1799030583312893, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 9659 + }, + { + "epoch": 0.0966, + "grad_norm": 1.1396175467308287, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9660 + }, + { + "epoch": 0.09661, + "grad_norm": 1.2031365613861558, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 9661 + }, + { + "epoch": 0.09662, + "grad_norm": 1.0450126941262252, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 9662 + }, + { + "epoch": 0.09663, + "grad_norm": 1.1872140736780354, + "learning_rate": 0.003, + "loss": 4.059, + "step": 9663 + }, + { + "epoch": 0.09664, + "grad_norm": 1.10816766795799, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 9664 + }, + { + "epoch": 0.09665, + "grad_norm": 1.0920304192442827, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 9665 + }, + { + "epoch": 0.09666, + "grad_norm": 1.0614448386527346, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 9666 + }, + { + "epoch": 0.09667, + "grad_norm": 1.222612568204848, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 9667 + }, + { + "epoch": 0.09668, + "grad_norm": 1.0545430433180965, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 9668 + }, + { + "epoch": 0.09669, + "grad_norm": 1.0956783811426631, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 9669 + }, + { + "epoch": 0.0967, + "grad_norm": 0.9610892180277101, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 9670 + }, + { + "epoch": 0.09671, + "grad_norm": 1.36766207674131, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 9671 + }, + { + "epoch": 0.09672, + "grad_norm": 0.9365680632039836, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 9672 + }, + { + "epoch": 0.09673, + "grad_norm": 1.0029650441296576, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 9673 + }, + { + "epoch": 0.09674, + "grad_norm": 1.1463824116993444, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 9674 + }, + { + "epoch": 0.09675, + "grad_norm": 1.058735123681513, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 9675 + }, + { + "epoch": 0.09676, + "grad_norm": 1.0766125342365733, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 9676 + }, + { + "epoch": 0.09677, + "grad_norm": 1.1203720989393562, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 9677 + }, + { + "epoch": 0.09678, + "grad_norm": 1.0951111296319187, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 9678 + }, + { + "epoch": 0.09679, + "grad_norm": 1.0249113062787711, + "learning_rate": 0.003, + "loss": 4.051, + "step": 9679 + }, + { + "epoch": 0.0968, + "grad_norm": 1.057027117497527, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 9680 + }, + { + "epoch": 0.09681, + "grad_norm": 1.2583816959197518, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 9681 + }, + { + "epoch": 0.09682, + "grad_norm": 1.2002373803491195, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9682 + }, + { + "epoch": 0.09683, + "grad_norm": 0.9521536845012109, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 9683 + }, + { + "epoch": 0.09684, + "grad_norm": 1.176715675692431, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 9684 + }, + { + "epoch": 0.09685, + "grad_norm": 1.2536000297265582, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 9685 + }, + { + "epoch": 0.09686, + "grad_norm": 1.2657578935031644, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 9686 + }, + { + "epoch": 0.09687, + "grad_norm": 1.045426925577382, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 9687 + }, + { + "epoch": 0.09688, + "grad_norm": 1.1541282081642033, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 9688 + }, + { + "epoch": 0.09689, + "grad_norm": 1.0905131544606357, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 9689 + }, + { + "epoch": 0.0969, + "grad_norm": 1.2174044997009796, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 9690 + }, + { + "epoch": 0.09691, + "grad_norm": 0.9988922116041921, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 9691 + }, + { + "epoch": 0.09692, + "grad_norm": 1.425363077395206, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 9692 + }, + { + "epoch": 0.09693, + "grad_norm": 0.9614291217845489, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 9693 + }, + { + "epoch": 0.09694, + "grad_norm": 1.1263744616665738, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 9694 + }, + { + "epoch": 0.09695, + "grad_norm": 0.9927369724604576, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 9695 + }, + { + "epoch": 0.09696, + "grad_norm": 1.204799634186295, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 9696 + }, + { + "epoch": 0.09697, + "grad_norm": 1.1019920635375835, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 9697 + }, + { + "epoch": 0.09698, + "grad_norm": 1.3083890446725224, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9698 + }, + { + "epoch": 0.09699, + "grad_norm": 0.8673207887854278, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 9699 + }, + { + "epoch": 0.097, + "grad_norm": 1.0937709937276758, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 9700 + }, + { + "epoch": 0.09701, + "grad_norm": 1.1096124083859713, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 9701 + }, + { + "epoch": 0.09702, + "grad_norm": 1.0470332399648397, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 9702 + }, + { + "epoch": 0.09703, + "grad_norm": 1.116033751774405, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 9703 + }, + { + "epoch": 0.09704, + "grad_norm": 1.322938797815828, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 9704 + }, + { + "epoch": 0.09705, + "grad_norm": 1.1674145052754212, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 9705 + }, + { + "epoch": 0.09706, + "grad_norm": 1.080731872467136, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 9706 + }, + { + "epoch": 0.09707, + "grad_norm": 1.0930211810900559, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 9707 + }, + { + "epoch": 0.09708, + "grad_norm": 1.1779520488029591, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 9708 + }, + { + "epoch": 0.09709, + "grad_norm": 1.1874834288052714, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9709 + }, + { + "epoch": 0.0971, + "grad_norm": 1.070110334968628, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 9710 + }, + { + "epoch": 0.09711, + "grad_norm": 0.9584275425436, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 9711 + }, + { + "epoch": 0.09712, + "grad_norm": 1.2305623246651165, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 9712 + }, + { + "epoch": 0.09713, + "grad_norm": 1.0142149218985719, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 9713 + }, + { + "epoch": 0.09714, + "grad_norm": 1.0721750280658653, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 9714 + }, + { + "epoch": 0.09715, + "grad_norm": 1.0682520259093737, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 9715 + }, + { + "epoch": 0.09716, + "grad_norm": 1.0242254335123158, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 9716 + }, + { + "epoch": 0.09717, + "grad_norm": 1.1149187761150097, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 9717 + }, + { + "epoch": 0.09718, + "grad_norm": 0.9890961402875876, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 9718 + }, + { + "epoch": 0.09719, + "grad_norm": 1.0299139105443698, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 9719 + }, + { + "epoch": 0.0972, + "grad_norm": 1.1558417735093496, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 9720 + }, + { + "epoch": 0.09721, + "grad_norm": 0.9555209453551541, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 9721 + }, + { + "epoch": 0.09722, + "grad_norm": 1.1785724250641767, + "learning_rate": 0.003, + "loss": 4.038, + "step": 9722 + }, + { + "epoch": 0.09723, + "grad_norm": 0.9949897584530741, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 9723 + }, + { + "epoch": 0.09724, + "grad_norm": 1.2568210621473228, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 9724 + }, + { + "epoch": 0.09725, + "grad_norm": 1.332241847275785, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 9725 + }, + { + "epoch": 0.09726, + "grad_norm": 1.2144800862045018, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 9726 + }, + { + "epoch": 0.09727, + "grad_norm": 1.011363287246841, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 9727 + }, + { + "epoch": 0.09728, + "grad_norm": 1.2120085901786424, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 9728 + }, + { + "epoch": 0.09729, + "grad_norm": 0.867741516356202, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 9729 + }, + { + "epoch": 0.0973, + "grad_norm": 0.8878505496669983, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 9730 + }, + { + "epoch": 0.09731, + "grad_norm": 1.0837727849054648, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 9731 + }, + { + "epoch": 0.09732, + "grad_norm": 1.2212446559927213, + "learning_rate": 0.003, + "loss": 4.043, + "step": 9732 + }, + { + "epoch": 0.09733, + "grad_norm": 1.2393726295211063, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 9733 + }, + { + "epoch": 0.09734, + "grad_norm": 1.188214463071944, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 9734 + }, + { + "epoch": 0.09735, + "grad_norm": 0.9872370881140842, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 9735 + }, + { + "epoch": 0.09736, + "grad_norm": 1.2580153206408196, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 9736 + }, + { + "epoch": 0.09737, + "grad_norm": 0.8841844781969095, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 9737 + }, + { + "epoch": 0.09738, + "grad_norm": 0.976812400102225, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 9738 + }, + { + "epoch": 0.09739, + "grad_norm": 1.1508306667326322, + "learning_rate": 0.003, + "loss": 4.048, + "step": 9739 + }, + { + "epoch": 0.0974, + "grad_norm": 1.196271566840404, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 9740 + }, + { + "epoch": 0.09741, + "grad_norm": 1.3082102267631772, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 9741 + }, + { + "epoch": 0.09742, + "grad_norm": 0.8994765241690306, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 9742 + }, + { + "epoch": 0.09743, + "grad_norm": 1.1682198162883546, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 9743 + }, + { + "epoch": 0.09744, + "grad_norm": 1.1469753871425274, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 9744 + }, + { + "epoch": 0.09745, + "grad_norm": 1.0422496922123192, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 9745 + }, + { + "epoch": 0.09746, + "grad_norm": 1.1368341501666346, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 9746 + }, + { + "epoch": 0.09747, + "grad_norm": 1.0608575439585237, + "learning_rate": 0.003, + "loss": 4.036, + "step": 9747 + }, + { + "epoch": 0.09748, + "grad_norm": 1.1013061298997167, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 9748 + }, + { + "epoch": 0.09749, + "grad_norm": 1.0678436420733588, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 9749 + }, + { + "epoch": 0.0975, + "grad_norm": 1.2786918164008214, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 9750 + }, + { + "epoch": 0.09751, + "grad_norm": 1.28463107870462, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 9751 + }, + { + "epoch": 0.09752, + "grad_norm": 1.0428808856437923, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 9752 + }, + { + "epoch": 0.09753, + "grad_norm": 1.192666194400153, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 9753 + }, + { + "epoch": 0.09754, + "grad_norm": 0.9605862543799322, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 9754 + }, + { + "epoch": 0.09755, + "grad_norm": 0.9697232496137559, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 9755 + }, + { + "epoch": 0.09756, + "grad_norm": 1.224040574050268, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 9756 + }, + { + "epoch": 0.09757, + "grad_norm": 1.1631102520521532, + "learning_rate": 0.003, + "loss": 4.036, + "step": 9757 + }, + { + "epoch": 0.09758, + "grad_norm": 1.2505629557746405, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 9758 + }, + { + "epoch": 0.09759, + "grad_norm": 0.9191423900203707, + "learning_rate": 0.003, + "loss": 4.028, + "step": 9759 + }, + { + "epoch": 0.0976, + "grad_norm": 1.1424671614519222, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 9760 + }, + { + "epoch": 0.09761, + "grad_norm": 1.102658611997917, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9761 + }, + { + "epoch": 0.09762, + "grad_norm": 1.1444051134380946, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 9762 + }, + { + "epoch": 0.09763, + "grad_norm": 0.9264161068395702, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 9763 + }, + { + "epoch": 0.09764, + "grad_norm": 1.1546408453955734, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 9764 + }, + { + "epoch": 0.09765, + "grad_norm": 1.0603790276325908, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 9765 + }, + { + "epoch": 0.09766, + "grad_norm": 1.1751196941995863, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 9766 + }, + { + "epoch": 0.09767, + "grad_norm": 1.0177083794650912, + "learning_rate": 0.003, + "loss": 4.046, + "step": 9767 + }, + { + "epoch": 0.09768, + "grad_norm": 1.273731379095236, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 9768 + }, + { + "epoch": 0.09769, + "grad_norm": 1.0996220193558155, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9769 + }, + { + "epoch": 0.0977, + "grad_norm": 1.0840911902806654, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 9770 + }, + { + "epoch": 0.09771, + "grad_norm": 1.2555935788729693, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 9771 + }, + { + "epoch": 0.09772, + "grad_norm": 1.0357209296089733, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 9772 + }, + { + "epoch": 0.09773, + "grad_norm": 1.1082243796424802, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 9773 + }, + { + "epoch": 0.09774, + "grad_norm": 1.1536368515267816, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 9774 + }, + { + "epoch": 0.09775, + "grad_norm": 1.1867161847229521, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 9775 + }, + { + "epoch": 0.09776, + "grad_norm": 0.9452459057379259, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 9776 + }, + { + "epoch": 0.09777, + "grad_norm": 1.302622907890426, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 9777 + }, + { + "epoch": 0.09778, + "grad_norm": 0.9532508643293588, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 9778 + }, + { + "epoch": 0.09779, + "grad_norm": 1.1242482678635257, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 9779 + }, + { + "epoch": 0.0978, + "grad_norm": 0.9839145612462165, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 9780 + }, + { + "epoch": 0.09781, + "grad_norm": 1.1498232001451654, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 9781 + }, + { + "epoch": 0.09782, + "grad_norm": 1.271609091807339, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 9782 + }, + { + "epoch": 0.09783, + "grad_norm": 1.058335316998838, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 9783 + }, + { + "epoch": 0.09784, + "grad_norm": 1.37604134703034, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 9784 + }, + { + "epoch": 0.09785, + "grad_norm": 1.1945900295219143, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 9785 + }, + { + "epoch": 0.09786, + "grad_norm": 1.1892035335129034, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 9786 + }, + { + "epoch": 0.09787, + "grad_norm": 0.9198710416706617, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 9787 + }, + { + "epoch": 0.09788, + "grad_norm": 1.0963105756254876, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 9788 + }, + { + "epoch": 0.09789, + "grad_norm": 1.0434621221623521, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 9789 + }, + { + "epoch": 0.0979, + "grad_norm": 1.1550849940577819, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 9790 + }, + { + "epoch": 0.09791, + "grad_norm": 1.357987097806882, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 9791 + }, + { + "epoch": 0.09792, + "grad_norm": 1.042044201245543, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 9792 + }, + { + "epoch": 0.09793, + "grad_norm": 1.1919512358344675, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 9793 + }, + { + "epoch": 0.09794, + "grad_norm": 1.2241595549926594, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 9794 + }, + { + "epoch": 0.09795, + "grad_norm": 1.0586631623737053, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 9795 + }, + { + "epoch": 0.09796, + "grad_norm": 1.1914802085441085, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 9796 + }, + { + "epoch": 0.09797, + "grad_norm": 1.1115787515349747, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9797 + }, + { + "epoch": 0.09798, + "grad_norm": 0.9816749943466317, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 9798 + }, + { + "epoch": 0.09799, + "grad_norm": 1.1660406158959165, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 9799 + }, + { + "epoch": 0.098, + "grad_norm": 1.0709459229028373, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 9800 + }, + { + "epoch": 0.09801, + "grad_norm": 1.1990212956387598, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 9801 + }, + { + "epoch": 0.09802, + "grad_norm": 1.0567665044317882, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 9802 + }, + { + "epoch": 0.09803, + "grad_norm": 1.0778757992404968, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 9803 + }, + { + "epoch": 0.09804, + "grad_norm": 1.0525175964682807, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 9804 + }, + { + "epoch": 0.09805, + "grad_norm": 1.241338187112226, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 9805 + }, + { + "epoch": 0.09806, + "grad_norm": 1.0640857294444321, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 9806 + }, + { + "epoch": 0.09807, + "grad_norm": 1.1456311229706166, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 9807 + }, + { + "epoch": 0.09808, + "grad_norm": 1.1124715383770616, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 9808 + }, + { + "epoch": 0.09809, + "grad_norm": 1.0282435668770187, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 9809 + }, + { + "epoch": 0.0981, + "grad_norm": 1.0398795512010792, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 9810 + }, + { + "epoch": 0.09811, + "grad_norm": 0.9279075100819835, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 9811 + }, + { + "epoch": 0.09812, + "grad_norm": 1.0782186992787437, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 9812 + }, + { + "epoch": 0.09813, + "grad_norm": 1.3085300578739432, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9813 + }, + { + "epoch": 0.09814, + "grad_norm": 1.1101588548323373, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 9814 + }, + { + "epoch": 0.09815, + "grad_norm": 1.1431391579353671, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 9815 + }, + { + "epoch": 0.09816, + "grad_norm": 1.164627516774537, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 9816 + }, + { + "epoch": 0.09817, + "grad_norm": 1.1241257621100056, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 9817 + }, + { + "epoch": 0.09818, + "grad_norm": 1.1105822418561284, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 9818 + }, + { + "epoch": 0.09819, + "grad_norm": 1.2260148961373039, + "learning_rate": 0.003, + "loss": 4.046, + "step": 9819 + }, + { + "epoch": 0.0982, + "grad_norm": 1.0867720780444274, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 9820 + }, + { + "epoch": 0.09821, + "grad_norm": 0.925875521249806, + "learning_rate": 0.003, + "loss": 4.017, + "step": 9821 + }, + { + "epoch": 0.09822, + "grad_norm": 1.4487529076721064, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 9822 + }, + { + "epoch": 0.09823, + "grad_norm": 0.8294595735585626, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 9823 + }, + { + "epoch": 0.09824, + "grad_norm": 1.073309157510469, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 9824 + }, + { + "epoch": 0.09825, + "grad_norm": 1.279144134720854, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 9825 + }, + { + "epoch": 0.09826, + "grad_norm": 1.1476433885186508, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 9826 + }, + { + "epoch": 0.09827, + "grad_norm": 1.043838880236461, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 9827 + }, + { + "epoch": 0.09828, + "grad_norm": 1.1968693614717298, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 9828 + }, + { + "epoch": 0.09829, + "grad_norm": 1.0457138129997414, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 9829 + }, + { + "epoch": 0.0983, + "grad_norm": 1.5123833188618667, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 9830 + }, + { + "epoch": 0.09831, + "grad_norm": 0.8711665475933555, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 9831 + }, + { + "epoch": 0.09832, + "grad_norm": 1.144386517076489, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 9832 + }, + { + "epoch": 0.09833, + "grad_norm": 1.246511769111947, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 9833 + }, + { + "epoch": 0.09834, + "grad_norm": 1.13349377739888, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 9834 + }, + { + "epoch": 0.09835, + "grad_norm": 1.0894251468551766, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 9835 + }, + { + "epoch": 0.09836, + "grad_norm": 1.1354298533553584, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 9836 + }, + { + "epoch": 0.09837, + "grad_norm": 1.1000173143853216, + "learning_rate": 0.003, + "loss": 3.996, + "step": 9837 + }, + { + "epoch": 0.09838, + "grad_norm": 1.119210791465233, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 9838 + }, + { + "epoch": 0.09839, + "grad_norm": 0.9524635366093233, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 9839 + }, + { + "epoch": 0.0984, + "grad_norm": 1.0797554846674096, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 9840 + }, + { + "epoch": 0.09841, + "grad_norm": 1.1411790679575085, + "learning_rate": 0.003, + "loss": 4.045, + "step": 9841 + }, + { + "epoch": 0.09842, + "grad_norm": 1.1281791126318816, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 9842 + }, + { + "epoch": 0.09843, + "grad_norm": 0.9134744520495551, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 9843 + }, + { + "epoch": 0.09844, + "grad_norm": 1.0160022299473974, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 9844 + }, + { + "epoch": 0.09845, + "grad_norm": 1.3428743136543202, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 9845 + }, + { + "epoch": 0.09846, + "grad_norm": 1.0189991251253638, + "learning_rate": 0.003, + "loss": 4.011, + "step": 9846 + }, + { + "epoch": 0.09847, + "grad_norm": 1.1704243872927704, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 9847 + }, + { + "epoch": 0.09848, + "grad_norm": 0.9081125793690473, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 9848 + }, + { + "epoch": 0.09849, + "grad_norm": 1.0336695774661442, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 9849 + }, + { + "epoch": 0.0985, + "grad_norm": 1.185375263926777, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 9850 + }, + { + "epoch": 0.09851, + "grad_norm": 1.1305392937955534, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 9851 + }, + { + "epoch": 0.09852, + "grad_norm": 0.9581592333628471, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 9852 + }, + { + "epoch": 0.09853, + "grad_norm": 1.2114861648250734, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 9853 + }, + { + "epoch": 0.09854, + "grad_norm": 1.0284507718101719, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 9854 + }, + { + "epoch": 0.09855, + "grad_norm": 1.2232319748941225, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 9855 + }, + { + "epoch": 0.09856, + "grad_norm": 1.0855515860790759, + "learning_rate": 0.003, + "loss": 4.062, + "step": 9856 + }, + { + "epoch": 0.09857, + "grad_norm": 1.0753853243294782, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 9857 + }, + { + "epoch": 0.09858, + "grad_norm": 1.206054994842722, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 9858 + }, + { + "epoch": 0.09859, + "grad_norm": 1.098563875256706, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 9859 + }, + { + "epoch": 0.0986, + "grad_norm": 1.1139961945407197, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 9860 + }, + { + "epoch": 0.09861, + "grad_norm": 1.1965232138904764, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 9861 + }, + { + "epoch": 0.09862, + "grad_norm": 1.1112915398055838, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 9862 + }, + { + "epoch": 0.09863, + "grad_norm": 1.1765592866838956, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 9863 + }, + { + "epoch": 0.09864, + "grad_norm": 1.281604965271072, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 9864 + }, + { + "epoch": 0.09865, + "grad_norm": 0.9474747599256009, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 9865 + }, + { + "epoch": 0.09866, + "grad_norm": 1.2526292758671924, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 9866 + }, + { + "epoch": 0.09867, + "grad_norm": 0.9782340542182837, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 9867 + }, + { + "epoch": 0.09868, + "grad_norm": 1.2384674527105715, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 9868 + }, + { + "epoch": 0.09869, + "grad_norm": 1.1272829215817646, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 9869 + }, + { + "epoch": 0.0987, + "grad_norm": 1.0396668913063039, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 9870 + }, + { + "epoch": 0.09871, + "grad_norm": 1.037702753803307, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 9871 + }, + { + "epoch": 0.09872, + "grad_norm": 1.1864414611980871, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 9872 + }, + { + "epoch": 0.09873, + "grad_norm": 1.0076047434777218, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 9873 + }, + { + "epoch": 0.09874, + "grad_norm": 1.1197613528576318, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 9874 + }, + { + "epoch": 0.09875, + "grad_norm": 0.9686098749546355, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 9875 + }, + { + "epoch": 0.09876, + "grad_norm": 1.04278415670397, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 9876 + }, + { + "epoch": 0.09877, + "grad_norm": 1.402750496107436, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9877 + }, + { + "epoch": 0.09878, + "grad_norm": 1.2330935372023277, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 9878 + }, + { + "epoch": 0.09879, + "grad_norm": 1.025634526463866, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 9879 + }, + { + "epoch": 0.0988, + "grad_norm": 1.0537640601114184, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 9880 + }, + { + "epoch": 0.09881, + "grad_norm": 1.2290027208808174, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 9881 + }, + { + "epoch": 0.09882, + "grad_norm": 0.8930767525332688, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 9882 + }, + { + "epoch": 0.09883, + "grad_norm": 1.0530103820851344, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 9883 + }, + { + "epoch": 0.09884, + "grad_norm": 1.2191490701919565, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 9884 + }, + { + "epoch": 0.09885, + "grad_norm": 1.0218706297933426, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 9885 + }, + { + "epoch": 0.09886, + "grad_norm": 1.150967177807416, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 9886 + }, + { + "epoch": 0.09887, + "grad_norm": 1.129373196803706, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9887 + }, + { + "epoch": 0.09888, + "grad_norm": 1.149004087106683, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 9888 + }, + { + "epoch": 0.09889, + "grad_norm": 1.2702764689102495, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 9889 + }, + { + "epoch": 0.0989, + "grad_norm": 1.2676550107001645, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 9890 + }, + { + "epoch": 0.09891, + "grad_norm": 1.2615471554776427, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 9891 + }, + { + "epoch": 0.09892, + "grad_norm": 1.1167316738746, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 9892 + }, + { + "epoch": 0.09893, + "grad_norm": 0.9054671662434778, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 9893 + }, + { + "epoch": 0.09894, + "grad_norm": 0.9236484052010723, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9894 + }, + { + "epoch": 0.09895, + "grad_norm": 1.208985347287342, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 9895 + }, + { + "epoch": 0.09896, + "grad_norm": 1.061049368371955, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 9896 + }, + { + "epoch": 0.09897, + "grad_norm": 1.329238894639678, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 9897 + }, + { + "epoch": 0.09898, + "grad_norm": 0.9855278608052852, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 9898 + }, + { + "epoch": 0.09899, + "grad_norm": 1.1077024624762888, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 9899 + }, + { + "epoch": 0.099, + "grad_norm": 1.206281945713825, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 9900 + }, + { + "epoch": 0.09901, + "grad_norm": 1.1299933707769678, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 9901 + }, + { + "epoch": 0.09902, + "grad_norm": 1.195651589757622, + "learning_rate": 0.003, + "loss": 4.038, + "step": 9902 + }, + { + "epoch": 0.09903, + "grad_norm": 0.9114394972258246, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 9903 + }, + { + "epoch": 0.09904, + "grad_norm": 0.8585107487801131, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 9904 + }, + { + "epoch": 0.09905, + "grad_norm": 0.8579026965268124, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 9905 + }, + { + "epoch": 0.09906, + "grad_norm": 0.9985171119339783, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 9906 + }, + { + "epoch": 0.09907, + "grad_norm": 1.1354189296859807, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 9907 + }, + { + "epoch": 0.09908, + "grad_norm": 1.0976033717968405, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 9908 + }, + { + "epoch": 0.09909, + "grad_norm": 1.1043513109244145, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 9909 + }, + { + "epoch": 0.0991, + "grad_norm": 1.2913134989756772, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 9910 + }, + { + "epoch": 0.09911, + "grad_norm": 1.0885873929556769, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 9911 + }, + { + "epoch": 0.09912, + "grad_norm": 1.439636649962891, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 9912 + }, + { + "epoch": 0.09913, + "grad_norm": 0.9927728387511651, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9913 + }, + { + "epoch": 0.09914, + "grad_norm": 1.224689926683272, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 9914 + }, + { + "epoch": 0.09915, + "grad_norm": 1.0185052815824782, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 9915 + }, + { + "epoch": 0.09916, + "grad_norm": 1.1039015304713733, + "learning_rate": 0.003, + "loss": 4.047, + "step": 9916 + }, + { + "epoch": 0.09917, + "grad_norm": 0.9560283744143093, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 9917 + }, + { + "epoch": 0.09918, + "grad_norm": 1.249009304246015, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 9918 + }, + { + "epoch": 0.09919, + "grad_norm": 1.0276598807835757, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 9919 + }, + { + "epoch": 0.0992, + "grad_norm": 1.4384067882969795, + "learning_rate": 0.003, + "loss": 4.053, + "step": 9920 + }, + { + "epoch": 0.09921, + "grad_norm": 1.0855607508489593, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 9921 + }, + { + "epoch": 0.09922, + "grad_norm": 1.3281753668853509, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 9922 + }, + { + "epoch": 0.09923, + "grad_norm": 1.3210796402306015, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 9923 + }, + { + "epoch": 0.09924, + "grad_norm": 0.9624480778672281, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 9924 + }, + { + "epoch": 0.09925, + "grad_norm": 1.0010419815015703, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 9925 + }, + { + "epoch": 0.09926, + "grad_norm": 1.481720653662113, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 9926 + }, + { + "epoch": 0.09927, + "grad_norm": 0.9737956838613526, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 9927 + }, + { + "epoch": 0.09928, + "grad_norm": 1.0743465025663415, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 9928 + }, + { + "epoch": 0.09929, + "grad_norm": 1.0978928933680199, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 9929 + }, + { + "epoch": 0.0993, + "grad_norm": 1.154918711721139, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 9930 + }, + { + "epoch": 0.09931, + "grad_norm": 1.292932588277131, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 9931 + }, + { + "epoch": 0.09932, + "grad_norm": 1.068387070754382, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 9932 + }, + { + "epoch": 0.09933, + "grad_norm": 1.0193039042669239, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 9933 + }, + { + "epoch": 0.09934, + "grad_norm": 1.0548178768539755, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 9934 + }, + { + "epoch": 0.09935, + "grad_norm": 1.1606839398396906, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 9935 + }, + { + "epoch": 0.09936, + "grad_norm": 1.2692975261466353, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9936 + }, + { + "epoch": 0.09937, + "grad_norm": 1.0782042799196752, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 9937 + }, + { + "epoch": 0.09938, + "grad_norm": 1.127980051105687, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 9938 + }, + { + "epoch": 0.09939, + "grad_norm": 1.0287441783269569, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 9939 + }, + { + "epoch": 0.0994, + "grad_norm": 1.2132424703891527, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 9940 + }, + { + "epoch": 0.09941, + "grad_norm": 0.9732597157956879, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 9941 + }, + { + "epoch": 0.09942, + "grad_norm": 1.1389068598359773, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 9942 + }, + { + "epoch": 0.09943, + "grad_norm": 1.102208282726071, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 9943 + }, + { + "epoch": 0.09944, + "grad_norm": 1.1127584490971916, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 9944 + }, + { + "epoch": 0.09945, + "grad_norm": 1.1635207514619523, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 9945 + }, + { + "epoch": 0.09946, + "grad_norm": 1.0595853146121823, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 9946 + }, + { + "epoch": 0.09947, + "grad_norm": 1.2408973624096764, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 9947 + }, + { + "epoch": 0.09948, + "grad_norm": 1.2344813666227283, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9948 + }, + { + "epoch": 0.09949, + "grad_norm": 1.1766030065853175, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 9949 + }, + { + "epoch": 0.0995, + "grad_norm": 1.0973536641516977, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 9950 + }, + { + "epoch": 0.09951, + "grad_norm": 1.0236399693187321, + "learning_rate": 0.003, + "loss": 4.008, + "step": 9951 + }, + { + "epoch": 0.09952, + "grad_norm": 0.9915712705437711, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 9952 + }, + { + "epoch": 0.09953, + "grad_norm": 1.1963753411506206, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 9953 + }, + { + "epoch": 0.09954, + "grad_norm": 1.104880815250887, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 9954 + }, + { + "epoch": 0.09955, + "grad_norm": 1.1708956908608728, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 9955 + }, + { + "epoch": 0.09956, + "grad_norm": 0.9866423739533496, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 9956 + }, + { + "epoch": 0.09957, + "grad_norm": 1.3165503592770573, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 9957 + }, + { + "epoch": 0.09958, + "grad_norm": 0.9536837415813932, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9958 + }, + { + "epoch": 0.09959, + "grad_norm": 1.3409977002707099, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 9959 + }, + { + "epoch": 0.0996, + "grad_norm": 0.9814138686721887, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 9960 + }, + { + "epoch": 0.09961, + "grad_norm": 1.3803176304264348, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 9961 + }, + { + "epoch": 0.09962, + "grad_norm": 1.1780995573672939, + "learning_rate": 0.003, + "loss": 4.053, + "step": 9962 + }, + { + "epoch": 0.09963, + "grad_norm": 1.0505789057253156, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 9963 + }, + { + "epoch": 0.09964, + "grad_norm": 1.2577481781152686, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 9964 + }, + { + "epoch": 0.09965, + "grad_norm": 0.8750383187391458, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 9965 + }, + { + "epoch": 0.09966, + "grad_norm": 0.9324526260136522, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 9966 + }, + { + "epoch": 0.09967, + "grad_norm": 1.3064826835022834, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 9967 + }, + { + "epoch": 0.09968, + "grad_norm": 1.0488845951972847, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 9968 + }, + { + "epoch": 0.09969, + "grad_norm": 1.2753549754302325, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 9969 + }, + { + "epoch": 0.0997, + "grad_norm": 1.039234075567994, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 9970 + }, + { + "epoch": 0.09971, + "grad_norm": 1.2214075051221316, + "learning_rate": 0.003, + "loss": 4.035, + "step": 9971 + }, + { + "epoch": 0.09972, + "grad_norm": 1.1623111884041941, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 9972 + }, + { + "epoch": 0.09973, + "grad_norm": 1.1088110650963057, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 9973 + }, + { + "epoch": 0.09974, + "grad_norm": 1.0426875423604298, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 9974 + }, + { + "epoch": 0.09975, + "grad_norm": 1.259426553612719, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 9975 + }, + { + "epoch": 0.09976, + "grad_norm": 1.0298690402873982, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 9976 + }, + { + "epoch": 0.09977, + "grad_norm": 1.3610931313558177, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 9977 + }, + { + "epoch": 0.09978, + "grad_norm": 0.9132158967205964, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 9978 + }, + { + "epoch": 0.09979, + "grad_norm": 1.219498068275462, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 9979 + }, + { + "epoch": 0.0998, + "grad_norm": 1.1274470393928808, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 9980 + }, + { + "epoch": 0.09981, + "grad_norm": 1.1231505756772355, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 9981 + }, + { + "epoch": 0.09982, + "grad_norm": 1.1169231509650528, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 9982 + }, + { + "epoch": 0.09983, + "grad_norm": 1.1018331890490325, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 9983 + }, + { + "epoch": 0.09984, + "grad_norm": 1.0714964487278968, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 9984 + }, + { + "epoch": 0.09985, + "grad_norm": 1.1863401109530898, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9985 + }, + { + "epoch": 0.09986, + "grad_norm": 1.042279602365793, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 9986 + }, + { + "epoch": 0.09987, + "grad_norm": 1.2665835602618778, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 9987 + }, + { + "epoch": 0.09988, + "grad_norm": 0.89345425739561, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 9988 + }, + { + "epoch": 0.09989, + "grad_norm": 1.1833757769860112, + "learning_rate": 0.003, + "loss": 4.027, + "step": 9989 + }, + { + "epoch": 0.0999, + "grad_norm": 1.2356695097233596, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 9990 + }, + { + "epoch": 0.09991, + "grad_norm": 1.1575667617820786, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 9991 + }, + { + "epoch": 0.09992, + "grad_norm": 1.0547400553533788, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 9992 + }, + { + "epoch": 0.09993, + "grad_norm": 1.1235377413439607, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 9993 + }, + { + "epoch": 0.09994, + "grad_norm": 1.0785414820906185, + "learning_rate": 0.003, + "loss": 4.044, + "step": 9994 + }, + { + "epoch": 0.09995, + "grad_norm": 1.2688086844276698, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 9995 + }, + { + "epoch": 0.09996, + "grad_norm": 0.9896774509795319, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 9996 + }, + { + "epoch": 0.09997, + "grad_norm": 1.4769420952220298, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 9997 + }, + { + "epoch": 0.09998, + "grad_norm": 1.1514423885404736, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 9998 + }, + { + "epoch": 0.09999, + "grad_norm": 0.9975467451325027, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 9999 + }, + { + "epoch": 0.1, + "grad_norm": 1.0688722481962825, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 10000 + }, + { + "epoch": 0.10001, + "grad_norm": 1.0376217486708665, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 10001 + }, + { + "epoch": 0.10002, + "grad_norm": 1.1705774677906358, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 10002 + }, + { + "epoch": 0.10003, + "grad_norm": 0.9861302809355035, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 10003 + }, + { + "epoch": 0.10004, + "grad_norm": 1.1789334001801655, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10004 + }, + { + "epoch": 0.10005, + "grad_norm": 1.1989313194786173, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 10005 + }, + { + "epoch": 0.10006, + "grad_norm": 0.9685441038364853, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 10006 + }, + { + "epoch": 0.10007, + "grad_norm": 1.1406363380895599, + "learning_rate": 0.003, + "loss": 4.013, + "step": 10007 + }, + { + "epoch": 0.10008, + "grad_norm": 1.1401310833512326, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 10008 + }, + { + "epoch": 0.10009, + "grad_norm": 1.2874194973352402, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 10009 + }, + { + "epoch": 0.1001, + "grad_norm": 0.9395473569554981, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 10010 + }, + { + "epoch": 0.10011, + "grad_norm": 1.1533792238419305, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 10011 + }, + { + "epoch": 0.10012, + "grad_norm": 1.1011717780365395, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10012 + }, + { + "epoch": 0.10013, + "grad_norm": 1.1016950112639643, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 10013 + }, + { + "epoch": 0.10014, + "grad_norm": 1.1527773791821763, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 10014 + }, + { + "epoch": 0.10015, + "grad_norm": 1.1942046889073201, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 10015 + }, + { + "epoch": 0.10016, + "grad_norm": 0.957925342866834, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 10016 + }, + { + "epoch": 0.10017, + "grad_norm": 1.15252875423962, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 10017 + }, + { + "epoch": 0.10018, + "grad_norm": 1.0374379123877617, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 10018 + }, + { + "epoch": 0.10019, + "grad_norm": 1.1794252200816104, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 10019 + }, + { + "epoch": 0.1002, + "grad_norm": 1.5203356193802966, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 10020 + }, + { + "epoch": 0.10021, + "grad_norm": 1.0029056619089192, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 10021 + }, + { + "epoch": 0.10022, + "grad_norm": 1.3936100699293918, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 10022 + }, + { + "epoch": 0.10023, + "grad_norm": 0.9821509836904537, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 10023 + }, + { + "epoch": 0.10024, + "grad_norm": 1.0077875906773899, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 10024 + }, + { + "epoch": 0.10025, + "grad_norm": 1.2413891542340003, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 10025 + }, + { + "epoch": 0.10026, + "grad_norm": 0.9367455732537656, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 10026 + }, + { + "epoch": 0.10027, + "grad_norm": 1.0593921888120654, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 10027 + }, + { + "epoch": 0.10028, + "grad_norm": 1.167695592578088, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 10028 + }, + { + "epoch": 0.10029, + "grad_norm": 1.1890588826725337, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 10029 + }, + { + "epoch": 0.1003, + "grad_norm": 1.1200344528615116, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 10030 + }, + { + "epoch": 0.10031, + "grad_norm": 1.1417105447998874, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 10031 + }, + { + "epoch": 0.10032, + "grad_norm": 1.0185250099652587, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 10032 + }, + { + "epoch": 0.10033, + "grad_norm": 1.370599313867179, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 10033 + }, + { + "epoch": 0.10034, + "grad_norm": 1.1853235683781365, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 10034 + }, + { + "epoch": 0.10035, + "grad_norm": 1.063599029412153, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 10035 + }, + { + "epoch": 0.10036, + "grad_norm": 1.3175134033811957, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 10036 + }, + { + "epoch": 0.10037, + "grad_norm": 1.0960419587396677, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 10037 + }, + { + "epoch": 0.10038, + "grad_norm": 1.0317353100182982, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 10038 + }, + { + "epoch": 0.10039, + "grad_norm": 1.0757103090572837, + "learning_rate": 0.003, + "loss": 4.029, + "step": 10039 + }, + { + "epoch": 0.1004, + "grad_norm": 1.087521733469096, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 10040 + }, + { + "epoch": 0.10041, + "grad_norm": 1.2084731171411833, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 10041 + }, + { + "epoch": 0.10042, + "grad_norm": 0.9946783035892525, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 10042 + }, + { + "epoch": 0.10043, + "grad_norm": 1.4298803409305711, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 10043 + }, + { + "epoch": 0.10044, + "grad_norm": 0.9427138046048616, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10044 + }, + { + "epoch": 0.10045, + "grad_norm": 1.1440424909635767, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 10045 + }, + { + "epoch": 0.10046, + "grad_norm": 1.2208645128728925, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 10046 + }, + { + "epoch": 0.10047, + "grad_norm": 1.1520779920124797, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 10047 + }, + { + "epoch": 0.10048, + "grad_norm": 1.1859885677535336, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 10048 + }, + { + "epoch": 0.10049, + "grad_norm": 1.1159990063667586, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 10049 + }, + { + "epoch": 0.1005, + "grad_norm": 1.1289846456525583, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 10050 + }, + { + "epoch": 0.10051, + "grad_norm": 1.113938198451074, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10051 + }, + { + "epoch": 0.10052, + "grad_norm": 1.108899705969844, + "learning_rate": 0.003, + "loss": 4.022, + "step": 10052 + }, + { + "epoch": 0.10053, + "grad_norm": 1.3756042460285796, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 10053 + }, + { + "epoch": 0.10054, + "grad_norm": 0.9418379581640134, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 10054 + }, + { + "epoch": 0.10055, + "grad_norm": 1.1739862813921886, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 10055 + }, + { + "epoch": 0.10056, + "grad_norm": 0.9555340144478537, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 10056 + }, + { + "epoch": 0.10057, + "grad_norm": 1.1683761393937124, + "learning_rate": 0.003, + "loss": 4.044, + "step": 10057 + }, + { + "epoch": 0.10058, + "grad_norm": 0.9254542092215158, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 10058 + }, + { + "epoch": 0.10059, + "grad_norm": 1.0185610648270182, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 10059 + }, + { + "epoch": 0.1006, + "grad_norm": 1.1877134215630414, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 10060 + }, + { + "epoch": 0.10061, + "grad_norm": 1.1455483826531043, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 10061 + }, + { + "epoch": 0.10062, + "grad_norm": 1.2513080657957403, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 10062 + }, + { + "epoch": 0.10063, + "grad_norm": 1.2124912360540205, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 10063 + }, + { + "epoch": 0.10064, + "grad_norm": 1.1274860749990432, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 10064 + }, + { + "epoch": 0.10065, + "grad_norm": 1.2298310787584599, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 10065 + }, + { + "epoch": 0.10066, + "grad_norm": 0.9103543254343075, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 10066 + }, + { + "epoch": 0.10067, + "grad_norm": 1.0030930500020137, + "learning_rate": 0.003, + "loss": 4.037, + "step": 10067 + }, + { + "epoch": 0.10068, + "grad_norm": 1.2779301791387994, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 10068 + }, + { + "epoch": 0.10069, + "grad_norm": 0.8818316506876382, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 10069 + }, + { + "epoch": 0.1007, + "grad_norm": 0.9710635888596929, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 10070 + }, + { + "epoch": 0.10071, + "grad_norm": 1.2723709892951096, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 10071 + }, + { + "epoch": 0.10072, + "grad_norm": 1.0605681039514947, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 10072 + }, + { + "epoch": 0.10073, + "grad_norm": 1.2166085354095868, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 10073 + }, + { + "epoch": 0.10074, + "grad_norm": 1.2473625383962492, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 10074 + }, + { + "epoch": 0.10075, + "grad_norm": 1.2226275619264804, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 10075 + }, + { + "epoch": 0.10076, + "grad_norm": 1.229825772234811, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 10076 + }, + { + "epoch": 0.10077, + "grad_norm": 0.9456387474821364, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 10077 + }, + { + "epoch": 0.10078, + "grad_norm": 1.1385609146582953, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 10078 + }, + { + "epoch": 0.10079, + "grad_norm": 1.2697352120922631, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 10079 + }, + { + "epoch": 0.1008, + "grad_norm": 1.4025860537648713, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 10080 + }, + { + "epoch": 0.10081, + "grad_norm": 1.0216197686884323, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 10081 + }, + { + "epoch": 0.10082, + "grad_norm": 1.132424109176119, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10082 + }, + { + "epoch": 0.10083, + "grad_norm": 1.021553444751305, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 10083 + }, + { + "epoch": 0.10084, + "grad_norm": 1.148442579664875, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 10084 + }, + { + "epoch": 0.10085, + "grad_norm": 1.0104285734886607, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 10085 + }, + { + "epoch": 0.10086, + "grad_norm": 1.1379868586708077, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 10086 + }, + { + "epoch": 0.10087, + "grad_norm": 1.186592736125496, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 10087 + }, + { + "epoch": 0.10088, + "grad_norm": 1.0060310853482595, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 10088 + }, + { + "epoch": 0.10089, + "grad_norm": 1.0640640849166343, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 10089 + }, + { + "epoch": 0.1009, + "grad_norm": 1.3560044661047932, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 10090 + }, + { + "epoch": 0.10091, + "grad_norm": 1.1078241142090288, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 10091 + }, + { + "epoch": 0.10092, + "grad_norm": 1.0068291505885563, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 10092 + }, + { + "epoch": 0.10093, + "grad_norm": 1.0599705107020405, + "learning_rate": 0.003, + "loss": 4.032, + "step": 10093 + }, + { + "epoch": 0.10094, + "grad_norm": 1.0876529083055502, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 10094 + }, + { + "epoch": 0.10095, + "grad_norm": 1.0961661282027626, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 10095 + }, + { + "epoch": 0.10096, + "grad_norm": 1.374941789482069, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 10096 + }, + { + "epoch": 0.10097, + "grad_norm": 0.9082696715295778, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 10097 + }, + { + "epoch": 0.10098, + "grad_norm": 1.0771725339779925, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 10098 + }, + { + "epoch": 0.10099, + "grad_norm": 1.0182992509469744, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 10099 + }, + { + "epoch": 0.101, + "grad_norm": 1.1428363685009697, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 10100 + }, + { + "epoch": 0.10101, + "grad_norm": 0.9694039996712733, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 10101 + }, + { + "epoch": 0.10102, + "grad_norm": 1.0472023173402798, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 10102 + }, + { + "epoch": 0.10103, + "grad_norm": 1.312075296718054, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 10103 + }, + { + "epoch": 0.10104, + "grad_norm": 1.069956052695916, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10104 + }, + { + "epoch": 0.10105, + "grad_norm": 1.2931704494456024, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 10105 + }, + { + "epoch": 0.10106, + "grad_norm": 1.1208723791794828, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 10106 + }, + { + "epoch": 0.10107, + "grad_norm": 1.1935123431071022, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 10107 + }, + { + "epoch": 0.10108, + "grad_norm": 1.1350366451944691, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 10108 + }, + { + "epoch": 0.10109, + "grad_norm": 1.3607305740575226, + "learning_rate": 0.003, + "loss": 4.019, + "step": 10109 + }, + { + "epoch": 0.1011, + "grad_norm": 0.9949435252832233, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 10110 + }, + { + "epoch": 0.10111, + "grad_norm": 1.2015497083429867, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 10111 + }, + { + "epoch": 0.10112, + "grad_norm": 1.0762531833294637, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 10112 + }, + { + "epoch": 0.10113, + "grad_norm": 1.3607307522223708, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 10113 + }, + { + "epoch": 0.10114, + "grad_norm": 0.8372866022166425, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 10114 + }, + { + "epoch": 0.10115, + "grad_norm": 1.169337588013949, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 10115 + }, + { + "epoch": 0.10116, + "grad_norm": 1.1177277939964916, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 10116 + }, + { + "epoch": 0.10117, + "grad_norm": 1.1778175104666457, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 10117 + }, + { + "epoch": 0.10118, + "grad_norm": 1.120506464897047, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 10118 + }, + { + "epoch": 0.10119, + "grad_norm": 1.1474574001683169, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 10119 + }, + { + "epoch": 0.1012, + "grad_norm": 0.8783084850811431, + "learning_rate": 0.003, + "loss": 4.015, + "step": 10120 + }, + { + "epoch": 0.10121, + "grad_norm": 0.9780743863814633, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10121 + }, + { + "epoch": 0.10122, + "grad_norm": 1.235259685072935, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 10122 + }, + { + "epoch": 0.10123, + "grad_norm": 0.8532144497138836, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 10123 + }, + { + "epoch": 0.10124, + "grad_norm": 1.0588961816914821, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 10124 + }, + { + "epoch": 0.10125, + "grad_norm": 1.0997888688267687, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 10125 + }, + { + "epoch": 0.10126, + "grad_norm": 1.012973351895914, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 10126 + }, + { + "epoch": 0.10127, + "grad_norm": 1.1601701911571694, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 10127 + }, + { + "epoch": 0.10128, + "grad_norm": 0.8661155150916877, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 10128 + }, + { + "epoch": 0.10129, + "grad_norm": 1.01187696037196, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 10129 + }, + { + "epoch": 0.1013, + "grad_norm": 1.377560868899849, + "learning_rate": 0.003, + "loss": 4.008, + "step": 10130 + }, + { + "epoch": 0.10131, + "grad_norm": 1.2225045415187876, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 10131 + }, + { + "epoch": 0.10132, + "grad_norm": 1.0805752380760862, + "learning_rate": 0.003, + "loss": 4.043, + "step": 10132 + }, + { + "epoch": 0.10133, + "grad_norm": 1.1689334091154489, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 10133 + }, + { + "epoch": 0.10134, + "grad_norm": 1.3581198813392088, + "learning_rate": 0.003, + "loss": 4.019, + "step": 10134 + }, + { + "epoch": 0.10135, + "grad_norm": 0.7676340885206456, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 10135 + }, + { + "epoch": 0.10136, + "grad_norm": 1.0234587289832506, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 10136 + }, + { + "epoch": 0.10137, + "grad_norm": 1.301339315671966, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 10137 + }, + { + "epoch": 0.10138, + "grad_norm": 0.9786141586414955, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 10138 + }, + { + "epoch": 0.10139, + "grad_norm": 1.227946401625274, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 10139 + }, + { + "epoch": 0.1014, + "grad_norm": 1.1398058430097253, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 10140 + }, + { + "epoch": 0.10141, + "grad_norm": 1.2511523411004795, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 10141 + }, + { + "epoch": 0.10142, + "grad_norm": 1.0530250525908387, + "learning_rate": 0.003, + "loss": 4.017, + "step": 10142 + }, + { + "epoch": 0.10143, + "grad_norm": 1.2429388861453106, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 10143 + }, + { + "epoch": 0.10144, + "grad_norm": 0.8865877007140897, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10144 + }, + { + "epoch": 0.10145, + "grad_norm": 1.0163852178819597, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 10145 + }, + { + "epoch": 0.10146, + "grad_norm": 1.3657614044980302, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 10146 + }, + { + "epoch": 0.10147, + "grad_norm": 0.9950226203446018, + "learning_rate": 0.003, + "loss": 4.036, + "step": 10147 + }, + { + "epoch": 0.10148, + "grad_norm": 1.1980345153961933, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 10148 + }, + { + "epoch": 0.10149, + "grad_norm": 1.019226564296152, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 10149 + }, + { + "epoch": 0.1015, + "grad_norm": 0.965993025404742, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 10150 + }, + { + "epoch": 0.10151, + "grad_norm": 1.3679875752682615, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 10151 + }, + { + "epoch": 0.10152, + "grad_norm": 1.0149437487586224, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10152 + }, + { + "epoch": 0.10153, + "grad_norm": 1.3162415560624259, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10153 + }, + { + "epoch": 0.10154, + "grad_norm": 1.0889175982737318, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 10154 + }, + { + "epoch": 0.10155, + "grad_norm": 1.108256381768691, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10155 + }, + { + "epoch": 0.10156, + "grad_norm": 1.0449850400339091, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 10156 + }, + { + "epoch": 0.10157, + "grad_norm": 1.2857069143842972, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 10157 + }, + { + "epoch": 0.10158, + "grad_norm": 0.96497871026518, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 10158 + }, + { + "epoch": 0.10159, + "grad_norm": 1.2412464612862566, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 10159 + }, + { + "epoch": 0.1016, + "grad_norm": 1.0649011421878327, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 10160 + }, + { + "epoch": 0.10161, + "grad_norm": 1.3828387996970612, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 10161 + }, + { + "epoch": 0.10162, + "grad_norm": 1.0268495952079426, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10162 + }, + { + "epoch": 0.10163, + "grad_norm": 1.2209399315758185, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10163 + }, + { + "epoch": 0.10164, + "grad_norm": 1.0382424289978558, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 10164 + }, + { + "epoch": 0.10165, + "grad_norm": 1.1431393692734733, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10165 + }, + { + "epoch": 0.10166, + "grad_norm": 1.025676794642058, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 10166 + }, + { + "epoch": 0.10167, + "grad_norm": 1.0773528056179977, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 10167 + }, + { + "epoch": 0.10168, + "grad_norm": 0.9799372640648593, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 10168 + }, + { + "epoch": 0.10169, + "grad_norm": 1.3573925168119307, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 10169 + }, + { + "epoch": 0.1017, + "grad_norm": 1.069207482050234, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 10170 + }, + { + "epoch": 0.10171, + "grad_norm": 1.1246476134224932, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 10171 + }, + { + "epoch": 0.10172, + "grad_norm": 1.0443602726524843, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 10172 + }, + { + "epoch": 0.10173, + "grad_norm": 1.2312653223111745, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10173 + }, + { + "epoch": 0.10174, + "grad_norm": 1.1122257381493827, + "learning_rate": 0.003, + "loss": 4.024, + "step": 10174 + }, + { + "epoch": 0.10175, + "grad_norm": 1.3749066864648898, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 10175 + }, + { + "epoch": 0.10176, + "grad_norm": 0.9339307181884096, + "learning_rate": 0.003, + "loss": 4.031, + "step": 10176 + }, + { + "epoch": 0.10177, + "grad_norm": 0.8986097272545992, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 10177 + }, + { + "epoch": 0.10178, + "grad_norm": 0.9946058914616095, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10178 + }, + { + "epoch": 0.10179, + "grad_norm": 1.499674078146875, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 10179 + }, + { + "epoch": 0.1018, + "grad_norm": 0.9496966180142791, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 10180 + }, + { + "epoch": 0.10181, + "grad_norm": 1.311618816110452, + "learning_rate": 0.003, + "loss": 4.041, + "step": 10181 + }, + { + "epoch": 0.10182, + "grad_norm": 1.0708634479960368, + "learning_rate": 0.003, + "loss": 4.049, + "step": 10182 + }, + { + "epoch": 0.10183, + "grad_norm": 1.1113870081016295, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 10183 + }, + { + "epoch": 0.10184, + "grad_norm": 1.1178353224836102, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 10184 + }, + { + "epoch": 0.10185, + "grad_norm": 1.1689044556545598, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 10185 + }, + { + "epoch": 0.10186, + "grad_norm": 1.419581610460463, + "learning_rate": 0.003, + "loss": 4.052, + "step": 10186 + }, + { + "epoch": 0.10187, + "grad_norm": 1.1733245395221399, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 10187 + }, + { + "epoch": 0.10188, + "grad_norm": 0.9580355128633431, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 10188 + }, + { + "epoch": 0.10189, + "grad_norm": 1.1722566739996738, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 10189 + }, + { + "epoch": 0.1019, + "grad_norm": 1.1612042034180012, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 10190 + }, + { + "epoch": 0.10191, + "grad_norm": 1.195329589772822, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 10191 + }, + { + "epoch": 0.10192, + "grad_norm": 1.0846595164302155, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 10192 + }, + { + "epoch": 0.10193, + "grad_norm": 1.1488860249777917, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10193 + }, + { + "epoch": 0.10194, + "grad_norm": 1.1895044545007407, + "learning_rate": 0.003, + "loss": 4.046, + "step": 10194 + }, + { + "epoch": 0.10195, + "grad_norm": 0.9769595430999226, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 10195 + }, + { + "epoch": 0.10196, + "grad_norm": 1.1017689182766344, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 10196 + }, + { + "epoch": 0.10197, + "grad_norm": 1.0793074132780305, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 10197 + }, + { + "epoch": 0.10198, + "grad_norm": 1.0506607464798967, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 10198 + }, + { + "epoch": 0.10199, + "grad_norm": 1.1837831773812761, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 10199 + }, + { + "epoch": 0.102, + "grad_norm": 1.076073379069326, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 10200 + }, + { + "epoch": 0.10201, + "grad_norm": 1.0825848802681943, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 10201 + }, + { + "epoch": 0.10202, + "grad_norm": 1.3241527514135416, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 10202 + }, + { + "epoch": 0.10203, + "grad_norm": 1.1724175197204518, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 10203 + }, + { + "epoch": 0.10204, + "grad_norm": 1.3185840871244239, + "learning_rate": 0.003, + "loss": 4.057, + "step": 10204 + }, + { + "epoch": 0.10205, + "grad_norm": 0.9734256552070326, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 10205 + }, + { + "epoch": 0.10206, + "grad_norm": 1.1022386726905078, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 10206 + }, + { + "epoch": 0.10207, + "grad_norm": 1.173133415900628, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 10207 + }, + { + "epoch": 0.10208, + "grad_norm": 1.1305215932951866, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 10208 + }, + { + "epoch": 0.10209, + "grad_norm": 1.190358205023772, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 10209 + }, + { + "epoch": 0.1021, + "grad_norm": 1.065896145776302, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 10210 + }, + { + "epoch": 0.10211, + "grad_norm": 1.3102703273182168, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 10211 + }, + { + "epoch": 0.10212, + "grad_norm": 0.9660207653575211, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10212 + }, + { + "epoch": 0.10213, + "grad_norm": 0.9543652414333238, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 10213 + }, + { + "epoch": 0.10214, + "grad_norm": 1.036805427205904, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 10214 + }, + { + "epoch": 0.10215, + "grad_norm": 1.057033327613356, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 10215 + }, + { + "epoch": 0.10216, + "grad_norm": 1.134350884528531, + "learning_rate": 0.003, + "loss": 4.043, + "step": 10216 + }, + { + "epoch": 0.10217, + "grad_norm": 1.1648146450494354, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 10217 + }, + { + "epoch": 0.10218, + "grad_norm": 1.4506285684005091, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 10218 + }, + { + "epoch": 0.10219, + "grad_norm": 0.9222728969038609, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 10219 + }, + { + "epoch": 0.1022, + "grad_norm": 1.0654406081445127, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 10220 + }, + { + "epoch": 0.10221, + "grad_norm": 1.252155041126181, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 10221 + }, + { + "epoch": 0.10222, + "grad_norm": 1.206778847995836, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 10222 + }, + { + "epoch": 0.10223, + "grad_norm": 1.2444900182503111, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 10223 + }, + { + "epoch": 0.10224, + "grad_norm": 1.040598030326006, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 10224 + }, + { + "epoch": 0.10225, + "grad_norm": 1.4878752264835577, + "learning_rate": 0.003, + "loss": 4.04, + "step": 10225 + }, + { + "epoch": 0.10226, + "grad_norm": 0.9402450092696958, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 10226 + }, + { + "epoch": 0.10227, + "grad_norm": 1.2884407987069295, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 10227 + }, + { + "epoch": 0.10228, + "grad_norm": 1.1522519965955926, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 10228 + }, + { + "epoch": 0.10229, + "grad_norm": 1.0468946409066122, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 10229 + }, + { + "epoch": 0.1023, + "grad_norm": 1.2964574881601119, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 10230 + }, + { + "epoch": 0.10231, + "grad_norm": 1.24484224526044, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 10231 + }, + { + "epoch": 0.10232, + "grad_norm": 1.0853521484203787, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 10232 + }, + { + "epoch": 0.10233, + "grad_norm": 1.065069643441235, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 10233 + }, + { + "epoch": 0.10234, + "grad_norm": 1.1844230860579215, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 10234 + }, + { + "epoch": 0.10235, + "grad_norm": 1.1858854955074398, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 10235 + }, + { + "epoch": 0.10236, + "grad_norm": 1.155823762579354, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 10236 + }, + { + "epoch": 0.10237, + "grad_norm": 1.259243563598782, + "learning_rate": 0.003, + "loss": 4.032, + "step": 10237 + }, + { + "epoch": 0.10238, + "grad_norm": 1.021888249686767, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 10238 + }, + { + "epoch": 0.10239, + "grad_norm": 1.2726229230148722, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 10239 + }, + { + "epoch": 0.1024, + "grad_norm": 0.8686221007981468, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 10240 + }, + { + "epoch": 0.10241, + "grad_norm": 1.0195922987283028, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 10241 + }, + { + "epoch": 0.10242, + "grad_norm": 1.4977495826980316, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 10242 + }, + { + "epoch": 0.10243, + "grad_norm": 0.8331453442599753, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 10243 + }, + { + "epoch": 0.10244, + "grad_norm": 0.9904099892605175, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 10244 + }, + { + "epoch": 0.10245, + "grad_norm": 1.1062252978239564, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 10245 + }, + { + "epoch": 0.10246, + "grad_norm": 1.2991428997983552, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 10246 + }, + { + "epoch": 0.10247, + "grad_norm": 1.017899165985093, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 10247 + }, + { + "epoch": 0.10248, + "grad_norm": 1.1832127454060877, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 10248 + }, + { + "epoch": 0.10249, + "grad_norm": 1.1079456008025137, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10249 + }, + { + "epoch": 0.1025, + "grad_norm": 1.3362106747505438, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 10250 + }, + { + "epoch": 0.10251, + "grad_norm": 1.1117749161341806, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 10251 + }, + { + "epoch": 0.10252, + "grad_norm": 1.2698367881846497, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 10252 + }, + { + "epoch": 0.10253, + "grad_norm": 1.149957054689331, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 10253 + }, + { + "epoch": 0.10254, + "grad_norm": 0.9917763597068653, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 10254 + }, + { + "epoch": 0.10255, + "grad_norm": 1.1692203345741474, + "learning_rate": 0.003, + "loss": 4.029, + "step": 10255 + }, + { + "epoch": 0.10256, + "grad_norm": 1.084143742321343, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 10256 + }, + { + "epoch": 0.10257, + "grad_norm": 1.2380367761444084, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 10257 + }, + { + "epoch": 0.10258, + "grad_norm": 1.1407215637889365, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10258 + }, + { + "epoch": 0.10259, + "grad_norm": 1.1517658724662878, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 10259 + }, + { + "epoch": 0.1026, + "grad_norm": 1.025390717700848, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 10260 + }, + { + "epoch": 0.10261, + "grad_norm": 1.2024437713037104, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 10261 + }, + { + "epoch": 0.10262, + "grad_norm": 1.0958713174379848, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10262 + }, + { + "epoch": 0.10263, + "grad_norm": 0.8765257399245001, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 10263 + }, + { + "epoch": 0.10264, + "grad_norm": 0.9633993597005445, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 10264 + }, + { + "epoch": 0.10265, + "grad_norm": 1.35821204523184, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 10265 + }, + { + "epoch": 0.10266, + "grad_norm": 1.0219625008871738, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 10266 + }, + { + "epoch": 0.10267, + "grad_norm": 1.2227874326162127, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 10267 + }, + { + "epoch": 0.10268, + "grad_norm": 0.978586011050389, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 10268 + }, + { + "epoch": 0.10269, + "grad_norm": 1.3366116601016778, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 10269 + }, + { + "epoch": 0.1027, + "grad_norm": 1.203267489591858, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 10270 + }, + { + "epoch": 0.10271, + "grad_norm": 1.098001417832544, + "learning_rate": 0.003, + "loss": 4.05, + "step": 10271 + }, + { + "epoch": 0.10272, + "grad_norm": 1.0307246792918792, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 10272 + }, + { + "epoch": 0.10273, + "grad_norm": 1.3053322880296834, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 10273 + }, + { + "epoch": 0.10274, + "grad_norm": 1.1385323131916913, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 10274 + }, + { + "epoch": 0.10275, + "grad_norm": 1.1261509616015806, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 10275 + }, + { + "epoch": 0.10276, + "grad_norm": 1.2868774897178286, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 10276 + }, + { + "epoch": 0.10277, + "grad_norm": 1.0360335060829537, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 10277 + }, + { + "epoch": 0.10278, + "grad_norm": 1.051397255526473, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 10278 + }, + { + "epoch": 0.10279, + "grad_norm": 1.1689181356787075, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 10279 + }, + { + "epoch": 0.1028, + "grad_norm": 1.0026627281522649, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 10280 + }, + { + "epoch": 0.10281, + "grad_norm": 1.2879919756618505, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 10281 + }, + { + "epoch": 0.10282, + "grad_norm": 1.244730230167987, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 10282 + }, + { + "epoch": 0.10283, + "grad_norm": 1.3076934827367117, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 10283 + }, + { + "epoch": 0.10284, + "grad_norm": 1.1046390773617638, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 10284 + }, + { + "epoch": 0.10285, + "grad_norm": 1.4413050093320747, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 10285 + }, + { + "epoch": 0.10286, + "grad_norm": 0.8441280250112908, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 10286 + }, + { + "epoch": 0.10287, + "grad_norm": 1.1644552941443256, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 10287 + }, + { + "epoch": 0.10288, + "grad_norm": 1.253613939194171, + "learning_rate": 0.003, + "loss": 4.071, + "step": 10288 + }, + { + "epoch": 0.10289, + "grad_norm": 1.1992464347629659, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 10289 + }, + { + "epoch": 0.1029, + "grad_norm": 1.1614080288077706, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 10290 + }, + { + "epoch": 0.10291, + "grad_norm": 1.1200552133267128, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 10291 + }, + { + "epoch": 0.10292, + "grad_norm": 1.1209107524052186, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 10292 + }, + { + "epoch": 0.10293, + "grad_norm": 1.1739986737744281, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 10293 + }, + { + "epoch": 0.10294, + "grad_norm": 0.9094006847879885, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 10294 + }, + { + "epoch": 0.10295, + "grad_norm": 1.0626685311138218, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 10295 + }, + { + "epoch": 0.10296, + "grad_norm": 1.3323240753459744, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 10296 + }, + { + "epoch": 0.10297, + "grad_norm": 0.9873549849936493, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 10297 + }, + { + "epoch": 0.10298, + "grad_norm": 1.1100743664279358, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 10298 + }, + { + "epoch": 0.10299, + "grad_norm": 1.2118577543293596, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10299 + }, + { + "epoch": 0.103, + "grad_norm": 1.1347265113461813, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 10300 + }, + { + "epoch": 0.10301, + "grad_norm": 1.0447143445167029, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 10301 + }, + { + "epoch": 0.10302, + "grad_norm": 1.051062325381112, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 10302 + }, + { + "epoch": 0.10303, + "grad_norm": 1.0323975069194364, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 10303 + }, + { + "epoch": 0.10304, + "grad_norm": 1.2789211780935934, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 10304 + }, + { + "epoch": 0.10305, + "grad_norm": 1.005336751800442, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 10305 + }, + { + "epoch": 0.10306, + "grad_norm": 1.1730045793816195, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 10306 + }, + { + "epoch": 0.10307, + "grad_norm": 1.1183555686080375, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 10307 + }, + { + "epoch": 0.10308, + "grad_norm": 1.0617744506344082, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 10308 + }, + { + "epoch": 0.10309, + "grad_norm": 1.22078736450434, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 10309 + }, + { + "epoch": 0.1031, + "grad_norm": 0.9980699183199506, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 10310 + }, + { + "epoch": 0.10311, + "grad_norm": 1.2462656361373519, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 10311 + }, + { + "epoch": 0.10312, + "grad_norm": 1.0351182027362364, + "learning_rate": 0.003, + "loss": 4.032, + "step": 10312 + }, + { + "epoch": 0.10313, + "grad_norm": 1.287680378354591, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 10313 + }, + { + "epoch": 0.10314, + "grad_norm": 0.9076021123086481, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10314 + }, + { + "epoch": 0.10315, + "grad_norm": 1.1481668973991306, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10315 + }, + { + "epoch": 0.10316, + "grad_norm": 1.0780477730461324, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 10316 + }, + { + "epoch": 0.10317, + "grad_norm": 1.1892911855004693, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 10317 + }, + { + "epoch": 0.10318, + "grad_norm": 1.0047821885404769, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 10318 + }, + { + "epoch": 0.10319, + "grad_norm": 1.28222936870825, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 10319 + }, + { + "epoch": 0.1032, + "grad_norm": 1.1296157925908548, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 10320 + }, + { + "epoch": 0.10321, + "grad_norm": 1.4853389937950883, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 10321 + }, + { + "epoch": 0.10322, + "grad_norm": 0.9787569170059056, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 10322 + }, + { + "epoch": 0.10323, + "grad_norm": 1.0906735449814466, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 10323 + }, + { + "epoch": 0.10324, + "grad_norm": 1.068837791641826, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 10324 + }, + { + "epoch": 0.10325, + "grad_norm": 1.2037337762025848, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10325 + }, + { + "epoch": 0.10326, + "grad_norm": 1.1133172543387875, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 10326 + }, + { + "epoch": 0.10327, + "grad_norm": 1.3167960269468966, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 10327 + }, + { + "epoch": 0.10328, + "grad_norm": 1.0696207926347543, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 10328 + }, + { + "epoch": 0.10329, + "grad_norm": 1.245399604421406, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 10329 + }, + { + "epoch": 0.1033, + "grad_norm": 1.0059697543492918, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 10330 + }, + { + "epoch": 0.10331, + "grad_norm": 1.5656251418654314, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 10331 + }, + { + "epoch": 0.10332, + "grad_norm": 0.8574115626605384, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 10332 + }, + { + "epoch": 0.10333, + "grad_norm": 0.9859350180122279, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 10333 + }, + { + "epoch": 0.10334, + "grad_norm": 1.2820418030806104, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 10334 + }, + { + "epoch": 0.10335, + "grad_norm": 0.8731451613756506, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 10335 + }, + { + "epoch": 0.10336, + "grad_norm": 1.014757675534356, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 10336 + }, + { + "epoch": 0.10337, + "grad_norm": 1.4909413920937094, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 10337 + }, + { + "epoch": 0.10338, + "grad_norm": 0.7816794866102409, + "learning_rate": 0.003, + "loss": 4.022, + "step": 10338 + }, + { + "epoch": 0.10339, + "grad_norm": 0.9243690877719841, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 10339 + }, + { + "epoch": 0.1034, + "grad_norm": 1.14694380987657, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 10340 + }, + { + "epoch": 0.10341, + "grad_norm": 1.3834936067183423, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 10341 + }, + { + "epoch": 0.10342, + "grad_norm": 1.1154396048927036, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 10342 + }, + { + "epoch": 0.10343, + "grad_norm": 1.1273639336085786, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 10343 + }, + { + "epoch": 0.10344, + "grad_norm": 1.1608493429486209, + "learning_rate": 0.003, + "loss": 4.04, + "step": 10344 + }, + { + "epoch": 0.10345, + "grad_norm": 1.06990959346778, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 10345 + }, + { + "epoch": 0.10346, + "grad_norm": 1.2479747749226988, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 10346 + }, + { + "epoch": 0.10347, + "grad_norm": 0.9737474653691929, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 10347 + }, + { + "epoch": 0.10348, + "grad_norm": 1.1819821090166085, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 10348 + }, + { + "epoch": 0.10349, + "grad_norm": 1.167019349942472, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 10349 + }, + { + "epoch": 0.1035, + "grad_norm": 1.3519029606445285, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 10350 + }, + { + "epoch": 0.10351, + "grad_norm": 0.8593715530349866, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 10351 + }, + { + "epoch": 0.10352, + "grad_norm": 0.9127235399409747, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10352 + }, + { + "epoch": 0.10353, + "grad_norm": 1.1579963956603465, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 10353 + }, + { + "epoch": 0.10354, + "grad_norm": 1.2797240045316958, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 10354 + }, + { + "epoch": 0.10355, + "grad_norm": 0.9612463917392908, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 10355 + }, + { + "epoch": 0.10356, + "grad_norm": 1.1043787877283193, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 10356 + }, + { + "epoch": 0.10357, + "grad_norm": 1.121926767893712, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 10357 + }, + { + "epoch": 0.10358, + "grad_norm": 1.0575829184316827, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 10358 + }, + { + "epoch": 0.10359, + "grad_norm": 1.1712412204356015, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 10359 + }, + { + "epoch": 0.1036, + "grad_norm": 1.1280743551856558, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 10360 + }, + { + "epoch": 0.10361, + "grad_norm": 1.100752064232454, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 10361 + }, + { + "epoch": 0.10362, + "grad_norm": 1.3611223839175068, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 10362 + }, + { + "epoch": 0.10363, + "grad_norm": 0.9602281325752356, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 10363 + }, + { + "epoch": 0.10364, + "grad_norm": 1.2553483424870702, + "learning_rate": 0.003, + "loss": 4.034, + "step": 10364 + }, + { + "epoch": 0.10365, + "grad_norm": 1.4024905356997521, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 10365 + }, + { + "epoch": 0.10366, + "grad_norm": 0.9242656529205991, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10366 + }, + { + "epoch": 0.10367, + "grad_norm": 1.1604823265405697, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 10367 + }, + { + "epoch": 0.10368, + "grad_norm": 1.1969970485888448, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 10368 + }, + { + "epoch": 0.10369, + "grad_norm": 1.0200300949712449, + "learning_rate": 0.003, + "loss": 4.021, + "step": 10369 + }, + { + "epoch": 0.1037, + "grad_norm": 1.2689639441368583, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 10370 + }, + { + "epoch": 0.10371, + "grad_norm": 1.0341059906639019, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 10371 + }, + { + "epoch": 0.10372, + "grad_norm": 1.2538857040073335, + "learning_rate": 0.003, + "loss": 4.018, + "step": 10372 + }, + { + "epoch": 0.10373, + "grad_norm": 1.1201024616892967, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 10373 + }, + { + "epoch": 0.10374, + "grad_norm": 1.2950845025933793, + "learning_rate": 0.003, + "loss": 4.027, + "step": 10374 + }, + { + "epoch": 0.10375, + "grad_norm": 0.9699551875266422, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 10375 + }, + { + "epoch": 0.10376, + "grad_norm": 1.1580083785809567, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 10376 + }, + { + "epoch": 0.10377, + "grad_norm": 1.1093131687693019, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10377 + }, + { + "epoch": 0.10378, + "grad_norm": 1.2624596205999834, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 10378 + }, + { + "epoch": 0.10379, + "grad_norm": 1.0914527517869759, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 10379 + }, + { + "epoch": 0.1038, + "grad_norm": 1.151305634171054, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 10380 + }, + { + "epoch": 0.10381, + "grad_norm": 1.18506187805289, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 10381 + }, + { + "epoch": 0.10382, + "grad_norm": 0.960814714372655, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 10382 + }, + { + "epoch": 0.10383, + "grad_norm": 1.1274336145652182, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 10383 + }, + { + "epoch": 0.10384, + "grad_norm": 1.0778718704979358, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 10384 + }, + { + "epoch": 0.10385, + "grad_norm": 1.3346207537542287, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 10385 + }, + { + "epoch": 0.10386, + "grad_norm": 1.1729682327745514, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 10386 + }, + { + "epoch": 0.10387, + "grad_norm": 1.036661733873201, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 10387 + }, + { + "epoch": 0.10388, + "grad_norm": 1.1267714195794842, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 10388 + }, + { + "epoch": 0.10389, + "grad_norm": 1.070601905354444, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 10389 + }, + { + "epoch": 0.1039, + "grad_norm": 1.1108942610835553, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 10390 + }, + { + "epoch": 0.10391, + "grad_norm": 1.0766563550369346, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 10391 + }, + { + "epoch": 0.10392, + "grad_norm": 1.2926774985748823, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 10392 + }, + { + "epoch": 0.10393, + "grad_norm": 1.021832754641526, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 10393 + }, + { + "epoch": 0.10394, + "grad_norm": 1.2672584449062123, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 10394 + }, + { + "epoch": 0.10395, + "grad_norm": 0.9905815246662731, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 10395 + }, + { + "epoch": 0.10396, + "grad_norm": 1.0082239163019409, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 10396 + }, + { + "epoch": 0.10397, + "grad_norm": 1.2087702334917223, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 10397 + }, + { + "epoch": 0.10398, + "grad_norm": 1.3670646129935469, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 10398 + }, + { + "epoch": 0.10399, + "grad_norm": 1.0071185422959221, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 10399 + }, + { + "epoch": 0.104, + "grad_norm": 1.201312509023756, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 10400 + }, + { + "epoch": 0.10401, + "grad_norm": 0.9420904682898148, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 10401 + }, + { + "epoch": 0.10402, + "grad_norm": 1.2050493388965062, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 10402 + }, + { + "epoch": 0.10403, + "grad_norm": 1.046563241466676, + "learning_rate": 0.003, + "loss": 4.03, + "step": 10403 + }, + { + "epoch": 0.10404, + "grad_norm": 1.1628697809891466, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 10404 + }, + { + "epoch": 0.10405, + "grad_norm": 1.3199812512263778, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10405 + }, + { + "epoch": 0.10406, + "grad_norm": 1.0594480118801703, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 10406 + }, + { + "epoch": 0.10407, + "grad_norm": 1.1233067919692772, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 10407 + }, + { + "epoch": 0.10408, + "grad_norm": 0.9910896481437917, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 10408 + }, + { + "epoch": 0.10409, + "grad_norm": 1.138918453922149, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 10409 + }, + { + "epoch": 0.1041, + "grad_norm": 1.1821655962343458, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 10410 + }, + { + "epoch": 0.10411, + "grad_norm": 1.214060735642106, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 10411 + }, + { + "epoch": 0.10412, + "grad_norm": 1.1271952204136202, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 10412 + }, + { + "epoch": 0.10413, + "grad_norm": 0.9720082652463884, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 10413 + }, + { + "epoch": 0.10414, + "grad_norm": 1.1999444084371422, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 10414 + }, + { + "epoch": 0.10415, + "grad_norm": 1.073604246138116, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 10415 + }, + { + "epoch": 0.10416, + "grad_norm": 1.108459057982964, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 10416 + }, + { + "epoch": 0.10417, + "grad_norm": 1.271581283838925, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 10417 + }, + { + "epoch": 0.10418, + "grad_norm": 1.079645522515878, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 10418 + }, + { + "epoch": 0.10419, + "grad_norm": 1.1152248774580158, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 10419 + }, + { + "epoch": 0.1042, + "grad_norm": 1.3480810386551398, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 10420 + }, + { + "epoch": 0.10421, + "grad_norm": 0.9290743153390996, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 10421 + }, + { + "epoch": 0.10422, + "grad_norm": 1.0777250714984647, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 10422 + }, + { + "epoch": 0.10423, + "grad_norm": 1.2450132081404977, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 10423 + }, + { + "epoch": 0.10424, + "grad_norm": 0.9622549883886152, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 10424 + }, + { + "epoch": 0.10425, + "grad_norm": 1.1529396436416368, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 10425 + }, + { + "epoch": 0.10426, + "grad_norm": 1.0658110282776228, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 10426 + }, + { + "epoch": 0.10427, + "grad_norm": 1.0117761795322662, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 10427 + }, + { + "epoch": 0.10428, + "grad_norm": 1.1685995916063228, + "learning_rate": 0.003, + "loss": 4.018, + "step": 10428 + }, + { + "epoch": 0.10429, + "grad_norm": 1.0923828450042794, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 10429 + }, + { + "epoch": 0.1043, + "grad_norm": 1.1655387142481999, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 10430 + }, + { + "epoch": 0.10431, + "grad_norm": 1.3531912916034343, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10431 + }, + { + "epoch": 0.10432, + "grad_norm": 0.9691209365956966, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 10432 + }, + { + "epoch": 0.10433, + "grad_norm": 1.0876767583765015, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 10433 + }, + { + "epoch": 0.10434, + "grad_norm": 1.0893989324457354, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 10434 + }, + { + "epoch": 0.10435, + "grad_norm": 1.1011090775263708, + "learning_rate": 0.003, + "loss": 4.016, + "step": 10435 + }, + { + "epoch": 0.10436, + "grad_norm": 1.2651035810376114, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 10436 + }, + { + "epoch": 0.10437, + "grad_norm": 1.077788583905916, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 10437 + }, + { + "epoch": 0.10438, + "grad_norm": 1.089658709482987, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 10438 + }, + { + "epoch": 0.10439, + "grad_norm": 1.3526434745275007, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 10439 + }, + { + "epoch": 0.1044, + "grad_norm": 1.072260827630628, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 10440 + }, + { + "epoch": 0.10441, + "grad_norm": 1.275665138867706, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10441 + }, + { + "epoch": 0.10442, + "grad_norm": 0.8780574008022751, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 10442 + }, + { + "epoch": 0.10443, + "grad_norm": 1.0997214258568853, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 10443 + }, + { + "epoch": 0.10444, + "grad_norm": 1.3431512939323784, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 10444 + }, + { + "epoch": 0.10445, + "grad_norm": 1.2074116232555316, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 10445 + }, + { + "epoch": 0.10446, + "grad_norm": 1.087105337997175, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 10446 + }, + { + "epoch": 0.10447, + "grad_norm": 1.0707463528471768, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10447 + }, + { + "epoch": 0.10448, + "grad_norm": 1.0692522472019002, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 10448 + }, + { + "epoch": 0.10449, + "grad_norm": 1.001943199192615, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 10449 + }, + { + "epoch": 0.1045, + "grad_norm": 1.259242230186131, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 10450 + }, + { + "epoch": 0.10451, + "grad_norm": 0.8956322006755586, + "learning_rate": 0.003, + "loss": 4.0, + "step": 10451 + }, + { + "epoch": 0.10452, + "grad_norm": 1.2342703982705026, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 10452 + }, + { + "epoch": 0.10453, + "grad_norm": 1.1659755217419676, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 10453 + }, + { + "epoch": 0.10454, + "grad_norm": 1.329730366250319, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10454 + }, + { + "epoch": 0.10455, + "grad_norm": 1.0525902475820774, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 10455 + }, + { + "epoch": 0.10456, + "grad_norm": 1.4730676027395804, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 10456 + }, + { + "epoch": 0.10457, + "grad_norm": 0.902647065287986, + "learning_rate": 0.003, + "loss": 4.039, + "step": 10457 + }, + { + "epoch": 0.10458, + "grad_norm": 0.9366078692778497, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 10458 + }, + { + "epoch": 0.10459, + "grad_norm": 1.185490872765169, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 10459 + }, + { + "epoch": 0.1046, + "grad_norm": 1.220542856466773, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 10460 + }, + { + "epoch": 0.10461, + "grad_norm": 1.0971020768947664, + "learning_rate": 0.003, + "loss": 4.008, + "step": 10461 + }, + { + "epoch": 0.10462, + "grad_norm": 1.0948259417406494, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 10462 + }, + { + "epoch": 0.10463, + "grad_norm": 1.0964751192354854, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 10463 + }, + { + "epoch": 0.10464, + "grad_norm": 1.1441448993227044, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 10464 + }, + { + "epoch": 0.10465, + "grad_norm": 1.1473924677302787, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 10465 + }, + { + "epoch": 0.10466, + "grad_norm": 1.1495513149677226, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 10466 + }, + { + "epoch": 0.10467, + "grad_norm": 1.0544675114536122, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10467 + }, + { + "epoch": 0.10468, + "grad_norm": 1.34298439731449, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 10468 + }, + { + "epoch": 0.10469, + "grad_norm": 1.2021399413747627, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 10469 + }, + { + "epoch": 0.1047, + "grad_norm": 1.1449386988739325, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 10470 + }, + { + "epoch": 0.10471, + "grad_norm": 1.0404658350920764, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 10471 + }, + { + "epoch": 0.10472, + "grad_norm": 1.2753272835439624, + "learning_rate": 0.003, + "loss": 4.019, + "step": 10472 + }, + { + "epoch": 0.10473, + "grad_norm": 1.1913855993707885, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 10473 + }, + { + "epoch": 0.10474, + "grad_norm": 1.1501830644015107, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 10474 + }, + { + "epoch": 0.10475, + "grad_norm": 1.243458316706119, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 10475 + }, + { + "epoch": 0.10476, + "grad_norm": 1.209956907576902, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 10476 + }, + { + "epoch": 0.10477, + "grad_norm": 1.0020350040143609, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 10477 + }, + { + "epoch": 0.10478, + "grad_norm": 1.0396708338378342, + "learning_rate": 0.003, + "loss": 4.018, + "step": 10478 + }, + { + "epoch": 0.10479, + "grad_norm": 1.1150470684477871, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10479 + }, + { + "epoch": 0.1048, + "grad_norm": 1.122637069690026, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 10480 + }, + { + "epoch": 0.10481, + "grad_norm": 1.4318413431939903, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 10481 + }, + { + "epoch": 0.10482, + "grad_norm": 1.0791992375976056, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 10482 + }, + { + "epoch": 0.10483, + "grad_norm": 1.3065570899996903, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 10483 + }, + { + "epoch": 0.10484, + "grad_norm": 1.1874032550959215, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 10484 + }, + { + "epoch": 0.10485, + "grad_norm": 1.1893565695289274, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 10485 + }, + { + "epoch": 0.10486, + "grad_norm": 1.097831179509903, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 10486 + }, + { + "epoch": 0.10487, + "grad_norm": 1.2993291047275668, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 10487 + }, + { + "epoch": 0.10488, + "grad_norm": 0.9496980239335198, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 10488 + }, + { + "epoch": 0.10489, + "grad_norm": 1.0629884631995594, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 10489 + }, + { + "epoch": 0.1049, + "grad_norm": 1.0603268965374313, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 10490 + }, + { + "epoch": 0.10491, + "grad_norm": 1.208967187962485, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 10491 + }, + { + "epoch": 0.10492, + "grad_norm": 1.2629214617983036, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 10492 + }, + { + "epoch": 0.10493, + "grad_norm": 1.2124437867421798, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 10493 + }, + { + "epoch": 0.10494, + "grad_norm": 0.9449400933558595, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 10494 + }, + { + "epoch": 0.10495, + "grad_norm": 1.1234515875951618, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 10495 + }, + { + "epoch": 0.10496, + "grad_norm": 1.0139506348014824, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 10496 + }, + { + "epoch": 0.10497, + "grad_norm": 1.3582522269851591, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 10497 + }, + { + "epoch": 0.10498, + "grad_norm": 0.8485161156664304, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 10498 + }, + { + "epoch": 0.10499, + "grad_norm": 0.9682827176949959, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 10499 + }, + { + "epoch": 0.105, + "grad_norm": 1.3767621333474114, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 10500 + }, + { + "epoch": 0.10501, + "grad_norm": 0.9911284707155722, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 10501 + }, + { + "epoch": 0.10502, + "grad_norm": 1.2098800890232249, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 10502 + }, + { + "epoch": 0.10503, + "grad_norm": 1.125482625749235, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 10503 + }, + { + "epoch": 0.10504, + "grad_norm": 1.049702069475873, + "learning_rate": 0.003, + "loss": 4.02, + "step": 10504 + }, + { + "epoch": 0.10505, + "grad_norm": 1.2213586640000345, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 10505 + }, + { + "epoch": 0.10506, + "grad_norm": 1.315035994618202, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10506 + }, + { + "epoch": 0.10507, + "grad_norm": 1.340531389565643, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 10507 + }, + { + "epoch": 0.10508, + "grad_norm": 0.9990031849773587, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 10508 + }, + { + "epoch": 0.10509, + "grad_norm": 1.1660130233668518, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 10509 + }, + { + "epoch": 0.1051, + "grad_norm": 1.039379676526276, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 10510 + }, + { + "epoch": 0.10511, + "grad_norm": 1.338586383397336, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 10511 + }, + { + "epoch": 0.10512, + "grad_norm": 1.0249548531611892, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 10512 + }, + { + "epoch": 0.10513, + "grad_norm": 1.2470715536322723, + "learning_rate": 0.003, + "loss": 4.003, + "step": 10513 + }, + { + "epoch": 0.10514, + "grad_norm": 0.9573916157628471, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 10514 + }, + { + "epoch": 0.10515, + "grad_norm": 1.299107259749984, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 10515 + }, + { + "epoch": 0.10516, + "grad_norm": 1.01524445765738, + "learning_rate": 0.003, + "loss": 4.0, + "step": 10516 + }, + { + "epoch": 0.10517, + "grad_norm": 1.1149804894227142, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 10517 + }, + { + "epoch": 0.10518, + "grad_norm": 1.1890411136167975, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 10518 + }, + { + "epoch": 0.10519, + "grad_norm": 0.9364490204330106, + "learning_rate": 0.003, + "loss": 4.005, + "step": 10519 + }, + { + "epoch": 0.1052, + "grad_norm": 1.1387642541625675, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 10520 + }, + { + "epoch": 0.10521, + "grad_norm": 1.1839482076334311, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 10521 + }, + { + "epoch": 0.10522, + "grad_norm": 1.0014230334791343, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 10522 + }, + { + "epoch": 0.10523, + "grad_norm": 1.1543281363304134, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 10523 + }, + { + "epoch": 0.10524, + "grad_norm": 1.0058642661701724, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 10524 + }, + { + "epoch": 0.10525, + "grad_norm": 1.2600433039619774, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 10525 + }, + { + "epoch": 0.10526, + "grad_norm": 1.0491904580994538, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10526 + }, + { + "epoch": 0.10527, + "grad_norm": 1.3835636637824966, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 10527 + }, + { + "epoch": 0.10528, + "grad_norm": 1.1190543265923238, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 10528 + }, + { + "epoch": 0.10529, + "grad_norm": 1.41063797874233, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 10529 + }, + { + "epoch": 0.1053, + "grad_norm": 1.0017135846243632, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 10530 + }, + { + "epoch": 0.10531, + "grad_norm": 1.1349481015715086, + "learning_rate": 0.003, + "loss": 4.028, + "step": 10531 + }, + { + "epoch": 0.10532, + "grad_norm": 1.0974147690879539, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 10532 + }, + { + "epoch": 0.10533, + "grad_norm": 1.1318583086940592, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 10533 + }, + { + "epoch": 0.10534, + "grad_norm": 0.9728943881000612, + "learning_rate": 0.003, + "loss": 4.034, + "step": 10534 + }, + { + "epoch": 0.10535, + "grad_norm": 1.4760229533414277, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 10535 + }, + { + "epoch": 0.10536, + "grad_norm": 1.0178312131559886, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 10536 + }, + { + "epoch": 0.10537, + "grad_norm": 1.326688326472958, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 10537 + }, + { + "epoch": 0.10538, + "grad_norm": 0.9540964220451233, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 10538 + }, + { + "epoch": 0.10539, + "grad_norm": 1.1867065524476788, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 10539 + }, + { + "epoch": 0.1054, + "grad_norm": 1.306268363150594, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 10540 + }, + { + "epoch": 0.10541, + "grad_norm": 1.1821939453890575, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 10541 + }, + { + "epoch": 0.10542, + "grad_norm": 1.340191153094188, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 10542 + }, + { + "epoch": 0.10543, + "grad_norm": 0.8866083438164943, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 10543 + }, + { + "epoch": 0.10544, + "grad_norm": 1.1034227189294226, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 10544 + }, + { + "epoch": 0.10545, + "grad_norm": 1.145569873679018, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 10545 + }, + { + "epoch": 0.10546, + "grad_norm": 1.2439976954614562, + "learning_rate": 0.003, + "loss": 4.067, + "step": 10546 + }, + { + "epoch": 0.10547, + "grad_norm": 0.9219452650458853, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 10547 + }, + { + "epoch": 0.10548, + "grad_norm": 0.9976599090984258, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 10548 + }, + { + "epoch": 0.10549, + "grad_norm": 1.2774703253681363, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 10549 + }, + { + "epoch": 0.1055, + "grad_norm": 1.1539905793697123, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 10550 + }, + { + "epoch": 0.10551, + "grad_norm": 1.135911995287088, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 10551 + }, + { + "epoch": 0.10552, + "grad_norm": 1.0527742384461596, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 10552 + }, + { + "epoch": 0.10553, + "grad_norm": 1.322002490054715, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 10553 + }, + { + "epoch": 0.10554, + "grad_norm": 1.0084031327815373, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 10554 + }, + { + "epoch": 0.10555, + "grad_norm": 1.0808986156972835, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 10555 + }, + { + "epoch": 0.10556, + "grad_norm": 1.1216797816863153, + "learning_rate": 0.003, + "loss": 4.037, + "step": 10556 + }, + { + "epoch": 0.10557, + "grad_norm": 1.1475055052985905, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 10557 + }, + { + "epoch": 0.10558, + "grad_norm": 1.1119323781963724, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 10558 + }, + { + "epoch": 0.10559, + "grad_norm": 1.2460835632404752, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 10559 + }, + { + "epoch": 0.1056, + "grad_norm": 1.2138261319567527, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 10560 + }, + { + "epoch": 0.10561, + "grad_norm": 1.0483615457288105, + "learning_rate": 0.003, + "loss": 4.024, + "step": 10561 + }, + { + "epoch": 0.10562, + "grad_norm": 1.0118320288891192, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 10562 + }, + { + "epoch": 0.10563, + "grad_norm": 1.3644503768047236, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 10563 + }, + { + "epoch": 0.10564, + "grad_norm": 0.8013831438598815, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 10564 + }, + { + "epoch": 0.10565, + "grad_norm": 0.9612754540930174, + "learning_rate": 0.003, + "loss": 4.033, + "step": 10565 + }, + { + "epoch": 0.10566, + "grad_norm": 1.2366075679735475, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 10566 + }, + { + "epoch": 0.10567, + "grad_norm": 1.155623176742999, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 10567 + }, + { + "epoch": 0.10568, + "grad_norm": 1.1220562684114699, + "learning_rate": 0.003, + "loss": 4.053, + "step": 10568 + }, + { + "epoch": 0.10569, + "grad_norm": 1.163422842918069, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 10569 + }, + { + "epoch": 0.1057, + "grad_norm": 1.1409717638103616, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 10570 + }, + { + "epoch": 0.10571, + "grad_norm": 0.9910760818265895, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 10571 + }, + { + "epoch": 0.10572, + "grad_norm": 1.1832943348904186, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 10572 + }, + { + "epoch": 0.10573, + "grad_norm": 1.0817165438920588, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 10573 + }, + { + "epoch": 0.10574, + "grad_norm": 1.2883743060276551, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 10574 + }, + { + "epoch": 0.10575, + "grad_norm": 0.8503051859413028, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 10575 + }, + { + "epoch": 0.10576, + "grad_norm": 1.0469525545832647, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 10576 + }, + { + "epoch": 0.10577, + "grad_norm": 1.509719903394313, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10577 + }, + { + "epoch": 0.10578, + "grad_norm": 0.9933503276360827, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 10578 + }, + { + "epoch": 0.10579, + "grad_norm": 1.1546805042641897, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 10579 + }, + { + "epoch": 0.1058, + "grad_norm": 1.10717494330016, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 10580 + }, + { + "epoch": 0.10581, + "grad_norm": 1.0811061670677058, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 10581 + }, + { + "epoch": 0.10582, + "grad_norm": 1.1040005473249752, + "learning_rate": 0.003, + "loss": 4.034, + "step": 10582 + }, + { + "epoch": 0.10583, + "grad_norm": 1.2803656515646102, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10583 + }, + { + "epoch": 0.10584, + "grad_norm": 1.3090939277354992, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 10584 + }, + { + "epoch": 0.10585, + "grad_norm": 1.0847880823254095, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 10585 + }, + { + "epoch": 0.10586, + "grad_norm": 1.2200835861910564, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 10586 + }, + { + "epoch": 0.10587, + "grad_norm": 0.8932627861610137, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 10587 + }, + { + "epoch": 0.10588, + "grad_norm": 1.139469360617456, + "learning_rate": 0.003, + "loss": 4.019, + "step": 10588 + }, + { + "epoch": 0.10589, + "grad_norm": 1.4344930118024855, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 10589 + }, + { + "epoch": 0.1059, + "grad_norm": 1.1102643686509654, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 10590 + }, + { + "epoch": 0.10591, + "grad_norm": 1.1481517513413337, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 10591 + }, + { + "epoch": 0.10592, + "grad_norm": 1.122474591041532, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 10592 + }, + { + "epoch": 0.10593, + "grad_norm": 1.2016781924295667, + "learning_rate": 0.003, + "loss": 4.005, + "step": 10593 + }, + { + "epoch": 0.10594, + "grad_norm": 1.07376059125612, + "learning_rate": 0.003, + "loss": 4.033, + "step": 10594 + }, + { + "epoch": 0.10595, + "grad_norm": 1.2904669765461627, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 10595 + }, + { + "epoch": 0.10596, + "grad_norm": 0.8779132609347672, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 10596 + }, + { + "epoch": 0.10597, + "grad_norm": 1.003771536112303, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 10597 + }, + { + "epoch": 0.10598, + "grad_norm": 1.0707616836208425, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 10598 + }, + { + "epoch": 0.10599, + "grad_norm": 0.9955175174902907, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 10599 + }, + { + "epoch": 0.106, + "grad_norm": 1.3851273303582732, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 10600 + }, + { + "epoch": 0.10601, + "grad_norm": 0.8326152872121492, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 10601 + }, + { + "epoch": 0.10602, + "grad_norm": 0.9405825691994378, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 10602 + }, + { + "epoch": 0.10603, + "grad_norm": 1.175319134130477, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 10603 + }, + { + "epoch": 0.10604, + "grad_norm": 1.2400190358250294, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 10604 + }, + { + "epoch": 0.10605, + "grad_norm": 1.3117115346590698, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 10605 + }, + { + "epoch": 0.10606, + "grad_norm": 0.8357581858102279, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 10606 + }, + { + "epoch": 0.10607, + "grad_norm": 1.2186083536406302, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 10607 + }, + { + "epoch": 0.10608, + "grad_norm": 1.3549499721362552, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10608 + }, + { + "epoch": 0.10609, + "grad_norm": 1.0359508960606705, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 10609 + }, + { + "epoch": 0.1061, + "grad_norm": 1.234427345167838, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 10610 + }, + { + "epoch": 0.10611, + "grad_norm": 1.1049507794457696, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 10611 + }, + { + "epoch": 0.10612, + "grad_norm": 1.1588897238356, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 10612 + }, + { + "epoch": 0.10613, + "grad_norm": 1.0882437441177095, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 10613 + }, + { + "epoch": 0.10614, + "grad_norm": 1.1661564975883256, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 10614 + }, + { + "epoch": 0.10615, + "grad_norm": 0.9756194205234838, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 10615 + }, + { + "epoch": 0.10616, + "grad_norm": 1.2136123225863367, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 10616 + }, + { + "epoch": 0.10617, + "grad_norm": 1.004520655549312, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 10617 + }, + { + "epoch": 0.10618, + "grad_norm": 1.293437758220031, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 10618 + }, + { + "epoch": 0.10619, + "grad_norm": 1.0745903694460457, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 10619 + }, + { + "epoch": 0.1062, + "grad_norm": 1.349386028002116, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 10620 + }, + { + "epoch": 0.10621, + "grad_norm": 1.0691689780916256, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 10621 + }, + { + "epoch": 0.10622, + "grad_norm": 1.224772531790401, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 10622 + }, + { + "epoch": 0.10623, + "grad_norm": 1.153790588423269, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 10623 + }, + { + "epoch": 0.10624, + "grad_norm": 1.193236273945861, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 10624 + }, + { + "epoch": 0.10625, + "grad_norm": 1.0654745501906075, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 10625 + }, + { + "epoch": 0.10626, + "grad_norm": 1.3261426114898907, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 10626 + }, + { + "epoch": 0.10627, + "grad_norm": 0.8982241457991678, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 10627 + }, + { + "epoch": 0.10628, + "grad_norm": 1.0114549713020966, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 10628 + }, + { + "epoch": 0.10629, + "grad_norm": 1.1489582478278977, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 10629 + }, + { + "epoch": 0.1063, + "grad_norm": 1.1444061673753105, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 10630 + }, + { + "epoch": 0.10631, + "grad_norm": 1.1576074418215807, + "learning_rate": 0.003, + "loss": 4.046, + "step": 10631 + }, + { + "epoch": 0.10632, + "grad_norm": 1.3627959422433176, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 10632 + }, + { + "epoch": 0.10633, + "grad_norm": 1.06565267900167, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 10633 + }, + { + "epoch": 0.10634, + "grad_norm": 1.307799627360254, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10634 + }, + { + "epoch": 0.10635, + "grad_norm": 0.865797003148197, + "learning_rate": 0.003, + "loss": 4.019, + "step": 10635 + }, + { + "epoch": 0.10636, + "grad_norm": 1.0276563707055373, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 10636 + }, + { + "epoch": 0.10637, + "grad_norm": 1.179328642984521, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 10637 + }, + { + "epoch": 0.10638, + "grad_norm": 1.1306397758583546, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 10638 + }, + { + "epoch": 0.10639, + "grad_norm": 1.1379536938728096, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 10639 + }, + { + "epoch": 0.1064, + "grad_norm": 1.0529351002129548, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 10640 + }, + { + "epoch": 0.10641, + "grad_norm": 1.4104967072137549, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 10641 + }, + { + "epoch": 0.10642, + "grad_norm": 0.9234934199965433, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 10642 + }, + { + "epoch": 0.10643, + "grad_norm": 1.060649329346942, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 10643 + }, + { + "epoch": 0.10644, + "grad_norm": 1.1653395641699265, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 10644 + }, + { + "epoch": 0.10645, + "grad_norm": 1.05447570499781, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 10645 + }, + { + "epoch": 0.10646, + "grad_norm": 1.0414075025866627, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 10646 + }, + { + "epoch": 0.10647, + "grad_norm": 1.1643043878647805, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 10647 + }, + { + "epoch": 0.10648, + "grad_norm": 1.0684722197241294, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 10648 + }, + { + "epoch": 0.10649, + "grad_norm": 1.3080634492785814, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 10649 + }, + { + "epoch": 0.1065, + "grad_norm": 1.2923711297653349, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 10650 + }, + { + "epoch": 0.10651, + "grad_norm": 1.1133144821639869, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 10651 + }, + { + "epoch": 0.10652, + "grad_norm": 1.1508639513581163, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 10652 + }, + { + "epoch": 0.10653, + "grad_norm": 1.1860732849874172, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 10653 + }, + { + "epoch": 0.10654, + "grad_norm": 1.0365296506149477, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 10654 + }, + { + "epoch": 0.10655, + "grad_norm": 0.9561188769757217, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 10655 + }, + { + "epoch": 0.10656, + "grad_norm": 1.0940765619925885, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 10656 + }, + { + "epoch": 0.10657, + "grad_norm": 1.3722387207935904, + "learning_rate": 0.003, + "loss": 4.037, + "step": 10657 + }, + { + "epoch": 0.10658, + "grad_norm": 1.1503504406176477, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 10658 + }, + { + "epoch": 0.10659, + "grad_norm": 0.966012228781889, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 10659 + }, + { + "epoch": 0.1066, + "grad_norm": 1.0144525965138471, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 10660 + }, + { + "epoch": 0.10661, + "grad_norm": 1.2051087661883464, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 10661 + }, + { + "epoch": 0.10662, + "grad_norm": 1.0764523304400608, + "learning_rate": 0.003, + "loss": 4.032, + "step": 10662 + }, + { + "epoch": 0.10663, + "grad_norm": 1.209131260208791, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10663 + }, + { + "epoch": 0.10664, + "grad_norm": 1.0568957147582319, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 10664 + }, + { + "epoch": 0.10665, + "grad_norm": 1.3417015377120707, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 10665 + }, + { + "epoch": 0.10666, + "grad_norm": 0.9362281562951543, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 10666 + }, + { + "epoch": 0.10667, + "grad_norm": 1.1589003721865672, + "learning_rate": 0.003, + "loss": 3.997, + "step": 10667 + }, + { + "epoch": 0.10668, + "grad_norm": 1.2601287656447024, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10668 + }, + { + "epoch": 0.10669, + "grad_norm": 1.061780696809955, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 10669 + }, + { + "epoch": 0.1067, + "grad_norm": 1.1903387328335544, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 10670 + }, + { + "epoch": 0.10671, + "grad_norm": 1.0301208869856568, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 10671 + }, + { + "epoch": 0.10672, + "grad_norm": 1.226842299928559, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 10672 + }, + { + "epoch": 0.10673, + "grad_norm": 1.1866348712385926, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10673 + }, + { + "epoch": 0.10674, + "grad_norm": 1.0069430418737448, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 10674 + }, + { + "epoch": 0.10675, + "grad_norm": 1.2985241581711995, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 10675 + }, + { + "epoch": 0.10676, + "grad_norm": 1.0289286257118662, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 10676 + }, + { + "epoch": 0.10677, + "grad_norm": 1.1701123709616597, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 10677 + }, + { + "epoch": 0.10678, + "grad_norm": 1.1718815214063174, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 10678 + }, + { + "epoch": 0.10679, + "grad_norm": 1.2943084912685583, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 10679 + }, + { + "epoch": 0.1068, + "grad_norm": 1.1957684694159134, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 10680 + }, + { + "epoch": 0.10681, + "grad_norm": 1.0034264649008786, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 10681 + }, + { + "epoch": 0.10682, + "grad_norm": 1.2761049621546057, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 10682 + }, + { + "epoch": 0.10683, + "grad_norm": 1.0598055696545081, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10683 + }, + { + "epoch": 0.10684, + "grad_norm": 1.218444770158993, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10684 + }, + { + "epoch": 0.10685, + "grad_norm": 0.9258871795824707, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 10685 + }, + { + "epoch": 0.10686, + "grad_norm": 1.133386229580556, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 10686 + }, + { + "epoch": 0.10687, + "grad_norm": 1.2653062418463612, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 10687 + }, + { + "epoch": 0.10688, + "grad_norm": 1.184995544981333, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 10688 + }, + { + "epoch": 0.10689, + "grad_norm": 1.0320005457605674, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 10689 + }, + { + "epoch": 0.1069, + "grad_norm": 1.2038777888203986, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 10690 + }, + { + "epoch": 0.10691, + "grad_norm": 1.129676406607381, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 10691 + }, + { + "epoch": 0.10692, + "grad_norm": 1.3575951971898041, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 10692 + }, + { + "epoch": 0.10693, + "grad_norm": 1.1196525664794985, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 10693 + }, + { + "epoch": 0.10694, + "grad_norm": 1.215242402090735, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 10694 + }, + { + "epoch": 0.10695, + "grad_norm": 1.013327742595491, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 10695 + }, + { + "epoch": 0.10696, + "grad_norm": 1.095537976839923, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 10696 + }, + { + "epoch": 0.10697, + "grad_norm": 1.2129639377799921, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 10697 + }, + { + "epoch": 0.10698, + "grad_norm": 1.1276205907442978, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 10698 + }, + { + "epoch": 0.10699, + "grad_norm": 1.4057292020168137, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 10699 + }, + { + "epoch": 0.107, + "grad_norm": 1.0889453831761258, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 10700 + }, + { + "epoch": 0.10701, + "grad_norm": 1.2636981543117896, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 10701 + }, + { + "epoch": 0.10702, + "grad_norm": 1.0444226985618574, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 10702 + }, + { + "epoch": 0.10703, + "grad_norm": 1.2951500282214232, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 10703 + }, + { + "epoch": 0.10704, + "grad_norm": 1.0393505702059902, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 10704 + }, + { + "epoch": 0.10705, + "grad_norm": 1.1797188685102178, + "learning_rate": 0.003, + "loss": 4.037, + "step": 10705 + }, + { + "epoch": 0.10706, + "grad_norm": 1.1975777074687664, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 10706 + }, + { + "epoch": 0.10707, + "grad_norm": 1.194075879308646, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 10707 + }, + { + "epoch": 0.10708, + "grad_norm": 1.0102354176185093, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 10708 + }, + { + "epoch": 0.10709, + "grad_norm": 1.3131806649777147, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 10709 + }, + { + "epoch": 0.1071, + "grad_norm": 1.2373109858890716, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 10710 + }, + { + "epoch": 0.10711, + "grad_norm": 1.2003702638346156, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 10711 + }, + { + "epoch": 0.10712, + "grad_norm": 0.9983255919354295, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 10712 + }, + { + "epoch": 0.10713, + "grad_norm": 0.9877113113359729, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 10713 + }, + { + "epoch": 0.10714, + "grad_norm": 1.3460252903807608, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 10714 + }, + { + "epoch": 0.10715, + "grad_norm": 0.9965502362486884, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 10715 + }, + { + "epoch": 0.10716, + "grad_norm": 1.2725890037766499, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 10716 + }, + { + "epoch": 0.10717, + "grad_norm": 1.1113072976703275, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10717 + }, + { + "epoch": 0.10718, + "grad_norm": 1.0100186651689251, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 10718 + }, + { + "epoch": 0.10719, + "grad_norm": 1.0640701601768137, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 10719 + }, + { + "epoch": 0.1072, + "grad_norm": 1.0593061211004513, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 10720 + }, + { + "epoch": 0.10721, + "grad_norm": 1.3099431143363085, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 10721 + }, + { + "epoch": 0.10722, + "grad_norm": 0.9809915726075293, + "learning_rate": 0.003, + "loss": 4.016, + "step": 10722 + }, + { + "epoch": 0.10723, + "grad_norm": 1.3406390414189182, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 10723 + }, + { + "epoch": 0.10724, + "grad_norm": 1.1349839036395613, + "learning_rate": 0.003, + "loss": 4.023, + "step": 10724 + }, + { + "epoch": 0.10725, + "grad_norm": 1.1603735826420933, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 10725 + }, + { + "epoch": 0.10726, + "grad_norm": 1.1306634640881634, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 10726 + }, + { + "epoch": 0.10727, + "grad_norm": 1.0401524644138374, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 10727 + }, + { + "epoch": 0.10728, + "grad_norm": 1.2861473387032047, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 10728 + }, + { + "epoch": 0.10729, + "grad_norm": 0.8700171921619617, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 10729 + }, + { + "epoch": 0.1073, + "grad_norm": 1.1494268855327163, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 10730 + }, + { + "epoch": 0.10731, + "grad_norm": 1.0235158899515824, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 10731 + }, + { + "epoch": 0.10732, + "grad_norm": 1.3171688814462412, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 10732 + }, + { + "epoch": 0.10733, + "grad_norm": 1.1618346197242664, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 10733 + }, + { + "epoch": 0.10734, + "grad_norm": 1.299839347131802, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 10734 + }, + { + "epoch": 0.10735, + "grad_norm": 0.9596941109018636, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 10735 + }, + { + "epoch": 0.10736, + "grad_norm": 1.047297478533331, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 10736 + }, + { + "epoch": 0.10737, + "grad_norm": 1.0623586974481836, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 10737 + }, + { + "epoch": 0.10738, + "grad_norm": 0.9993600202006488, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 10738 + }, + { + "epoch": 0.10739, + "grad_norm": 1.254811328838207, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 10739 + }, + { + "epoch": 0.1074, + "grad_norm": 1.02443107388267, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 10740 + }, + { + "epoch": 0.10741, + "grad_norm": 1.053670941966742, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 10741 + }, + { + "epoch": 0.10742, + "grad_norm": 1.3054887606936714, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 10742 + }, + { + "epoch": 0.10743, + "grad_norm": 1.1815464442351478, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 10743 + }, + { + "epoch": 0.10744, + "grad_norm": 1.0439503530541596, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 10744 + }, + { + "epoch": 0.10745, + "grad_norm": 1.2910029674176995, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 10745 + }, + { + "epoch": 0.10746, + "grad_norm": 1.1171478726339876, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 10746 + }, + { + "epoch": 0.10747, + "grad_norm": 1.316025255296466, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 10747 + }, + { + "epoch": 0.10748, + "grad_norm": 1.0185714857385384, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 10748 + }, + { + "epoch": 0.10749, + "grad_norm": 1.2319455923474587, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 10749 + }, + { + "epoch": 0.1075, + "grad_norm": 1.0698239902071434, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 10750 + }, + { + "epoch": 0.10751, + "grad_norm": 1.1767304687464153, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 10751 + }, + { + "epoch": 0.10752, + "grad_norm": 1.200346526551157, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 10752 + }, + { + "epoch": 0.10753, + "grad_norm": 0.8004910788946202, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 10753 + }, + { + "epoch": 0.10754, + "grad_norm": 0.9299169462294767, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 10754 + }, + { + "epoch": 0.10755, + "grad_norm": 1.1496170197907911, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 10755 + }, + { + "epoch": 0.10756, + "grad_norm": 1.3269009046033973, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 10756 + }, + { + "epoch": 0.10757, + "grad_norm": 1.01836707032746, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 10757 + }, + { + "epoch": 0.10758, + "grad_norm": 1.4360512920311785, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 10758 + }, + { + "epoch": 0.10759, + "grad_norm": 0.8198014966019104, + "learning_rate": 0.003, + "loss": 3.997, + "step": 10759 + }, + { + "epoch": 0.1076, + "grad_norm": 0.9805123439382194, + "learning_rate": 0.003, + "loss": 4.033, + "step": 10760 + }, + { + "epoch": 0.10761, + "grad_norm": 1.2404719998449543, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 10761 + }, + { + "epoch": 0.10762, + "grad_norm": 1.100158666709582, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10762 + }, + { + "epoch": 0.10763, + "grad_norm": 1.255637890490154, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 10763 + }, + { + "epoch": 0.10764, + "grad_norm": 1.0281623834755365, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 10764 + }, + { + "epoch": 0.10765, + "grad_norm": 1.4149964948090226, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10765 + }, + { + "epoch": 0.10766, + "grad_norm": 0.9836393890872247, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 10766 + }, + { + "epoch": 0.10767, + "grad_norm": 1.2741151461584197, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 10767 + }, + { + "epoch": 0.10768, + "grad_norm": 1.1238950688453537, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 10768 + }, + { + "epoch": 0.10769, + "grad_norm": 1.0948648776611019, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 10769 + }, + { + "epoch": 0.1077, + "grad_norm": 1.201055603123503, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 10770 + }, + { + "epoch": 0.10771, + "grad_norm": 1.0193171022277756, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 10771 + }, + { + "epoch": 0.10772, + "grad_norm": 1.3771536426901734, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 10772 + }, + { + "epoch": 0.10773, + "grad_norm": 1.036422327432776, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 10773 + }, + { + "epoch": 0.10774, + "grad_norm": 1.3787602954979676, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 10774 + }, + { + "epoch": 0.10775, + "grad_norm": 0.9876267992932676, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 10775 + }, + { + "epoch": 0.10776, + "grad_norm": 1.2287074239953188, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10776 + }, + { + "epoch": 0.10777, + "grad_norm": 1.1159465296748252, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 10777 + }, + { + "epoch": 0.10778, + "grad_norm": 1.2143634624449668, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10778 + }, + { + "epoch": 0.10779, + "grad_norm": 0.8371084824202994, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 10779 + }, + { + "epoch": 0.1078, + "grad_norm": 1.0448266717192551, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 10780 + }, + { + "epoch": 0.10781, + "grad_norm": 1.339724186849472, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 10781 + }, + { + "epoch": 0.10782, + "grad_norm": 0.9542629472880868, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 10782 + }, + { + "epoch": 0.10783, + "grad_norm": 1.1863204682968913, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10783 + }, + { + "epoch": 0.10784, + "grad_norm": 1.4462798896316145, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 10784 + }, + { + "epoch": 0.10785, + "grad_norm": 1.0951546615964138, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 10785 + }, + { + "epoch": 0.10786, + "grad_norm": 1.296430995678855, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 10786 + }, + { + "epoch": 0.10787, + "grad_norm": 0.8997746711593995, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 10787 + }, + { + "epoch": 0.10788, + "grad_norm": 1.170708369493236, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 10788 + }, + { + "epoch": 0.10789, + "grad_norm": 1.1961220945918547, + "learning_rate": 0.003, + "loss": 3.981, + "step": 10789 + }, + { + "epoch": 0.1079, + "grad_norm": 1.2476033414148846, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 10790 + }, + { + "epoch": 0.10791, + "grad_norm": 1.3938502880572137, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 10791 + }, + { + "epoch": 0.10792, + "grad_norm": 0.8287063955222098, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 10792 + }, + { + "epoch": 0.10793, + "grad_norm": 0.9968889789622127, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 10793 + }, + { + "epoch": 0.10794, + "grad_norm": 1.2636523367148476, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 10794 + }, + { + "epoch": 0.10795, + "grad_norm": 0.9320188951631827, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 10795 + }, + { + "epoch": 0.10796, + "grad_norm": 1.13871483259714, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 10796 + }, + { + "epoch": 0.10797, + "grad_norm": 1.1430569400531418, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 10797 + }, + { + "epoch": 0.10798, + "grad_norm": 1.2223707770106316, + "learning_rate": 0.003, + "loss": 4.027, + "step": 10798 + }, + { + "epoch": 0.10799, + "grad_norm": 1.0797391466920707, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 10799 + }, + { + "epoch": 0.108, + "grad_norm": 1.5331772300169273, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 10800 + }, + { + "epoch": 0.10801, + "grad_norm": 0.8462632416981025, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 10801 + }, + { + "epoch": 0.10802, + "grad_norm": 1.1323280804246525, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 10802 + }, + { + "epoch": 0.10803, + "grad_norm": 1.2472360350753866, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 10803 + }, + { + "epoch": 0.10804, + "grad_norm": 1.0221991986540284, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 10804 + }, + { + "epoch": 0.10805, + "grad_norm": 1.193633649724736, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 10805 + }, + { + "epoch": 0.10806, + "grad_norm": 1.128446301055873, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 10806 + }, + { + "epoch": 0.10807, + "grad_norm": 1.3429005701975214, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 10807 + }, + { + "epoch": 0.10808, + "grad_norm": 1.1916229558274094, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 10808 + }, + { + "epoch": 0.10809, + "grad_norm": 1.1666311030390335, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 10809 + }, + { + "epoch": 0.1081, + "grad_norm": 1.0948133461800402, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 10810 + }, + { + "epoch": 0.10811, + "grad_norm": 1.070956586984476, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 10811 + }, + { + "epoch": 0.10812, + "grad_norm": 1.1405962531093388, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 10812 + }, + { + "epoch": 0.10813, + "grad_norm": 1.1923710781482288, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 10813 + }, + { + "epoch": 0.10814, + "grad_norm": 0.9637789414068072, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 10814 + }, + { + "epoch": 0.10815, + "grad_norm": 1.1303684936509142, + "learning_rate": 0.003, + "loss": 4.033, + "step": 10815 + }, + { + "epoch": 0.10816, + "grad_norm": 1.1277334351478485, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 10816 + }, + { + "epoch": 0.10817, + "grad_norm": 1.394920688077164, + "learning_rate": 0.003, + "loss": 4.036, + "step": 10817 + }, + { + "epoch": 0.10818, + "grad_norm": 0.9247339438137491, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 10818 + }, + { + "epoch": 0.10819, + "grad_norm": 1.2827723765119825, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 10819 + }, + { + "epoch": 0.1082, + "grad_norm": 1.1027026712478345, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 10820 + }, + { + "epoch": 0.10821, + "grad_norm": 1.130402720324394, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 10821 + }, + { + "epoch": 0.10822, + "grad_norm": 1.1822569814525696, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 10822 + }, + { + "epoch": 0.10823, + "grad_norm": 1.3160765140161788, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 10823 + }, + { + "epoch": 0.10824, + "grad_norm": 1.2344419298288998, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 10824 + }, + { + "epoch": 0.10825, + "grad_norm": 0.9804550860424565, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 10825 + }, + { + "epoch": 0.10826, + "grad_norm": 1.2393166369862887, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 10826 + }, + { + "epoch": 0.10827, + "grad_norm": 1.1285558425268936, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 10827 + }, + { + "epoch": 0.10828, + "grad_norm": 1.2190487151901024, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 10828 + }, + { + "epoch": 0.10829, + "grad_norm": 1.0051177849528747, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 10829 + }, + { + "epoch": 0.1083, + "grad_norm": 1.2475552966341636, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 10830 + }, + { + "epoch": 0.10831, + "grad_norm": 1.086747469310628, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 10831 + }, + { + "epoch": 0.10832, + "grad_norm": 1.2521787369584434, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 10832 + }, + { + "epoch": 0.10833, + "grad_norm": 1.1044210264497303, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 10833 + }, + { + "epoch": 0.10834, + "grad_norm": 1.121459039289483, + "learning_rate": 0.003, + "loss": 3.988, + "step": 10834 + }, + { + "epoch": 0.10835, + "grad_norm": 1.1128708174680606, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 10835 + }, + { + "epoch": 0.10836, + "grad_norm": 1.5980637509276823, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 10836 + }, + { + "epoch": 0.10837, + "grad_norm": 1.2324712817974364, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10837 + }, + { + "epoch": 0.10838, + "grad_norm": 1.0657770844010468, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 10838 + }, + { + "epoch": 0.10839, + "grad_norm": 1.0927040171769402, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 10839 + }, + { + "epoch": 0.1084, + "grad_norm": 1.1359360509882592, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 10840 + }, + { + "epoch": 0.10841, + "grad_norm": 0.903405768475979, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 10841 + }, + { + "epoch": 0.10842, + "grad_norm": 1.227696001765903, + "learning_rate": 0.003, + "loss": 4.046, + "step": 10842 + }, + { + "epoch": 0.10843, + "grad_norm": 1.2265984387203541, + "learning_rate": 0.003, + "loss": 4.031, + "step": 10843 + }, + { + "epoch": 0.10844, + "grad_norm": 1.2973674659059795, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 10844 + }, + { + "epoch": 0.10845, + "grad_norm": 1.0539275266266204, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 10845 + }, + { + "epoch": 0.10846, + "grad_norm": 1.2771787744436187, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 10846 + }, + { + "epoch": 0.10847, + "grad_norm": 1.1019062809673246, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 10847 + }, + { + "epoch": 0.10848, + "grad_norm": 1.170952800012063, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 10848 + }, + { + "epoch": 0.10849, + "grad_norm": 1.199786332475287, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 10849 + }, + { + "epoch": 0.1085, + "grad_norm": 1.0400770912095054, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 10850 + }, + { + "epoch": 0.10851, + "grad_norm": 1.169933869139355, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 10851 + }, + { + "epoch": 0.10852, + "grad_norm": 1.134282690971657, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 10852 + }, + { + "epoch": 0.10853, + "grad_norm": 1.1214683357708541, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 10853 + }, + { + "epoch": 0.10854, + "grad_norm": 1.1718280449767655, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 10854 + }, + { + "epoch": 0.10855, + "grad_norm": 1.3835381568865763, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 10855 + }, + { + "epoch": 0.10856, + "grad_norm": 0.8314414595792815, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 10856 + }, + { + "epoch": 0.10857, + "grad_norm": 0.9164952496161265, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 10857 + }, + { + "epoch": 0.10858, + "grad_norm": 1.1610451470170327, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 10858 + }, + { + "epoch": 0.10859, + "grad_norm": 1.1965855517444213, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 10859 + }, + { + "epoch": 0.1086, + "grad_norm": 1.128342345854473, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 10860 + }, + { + "epoch": 0.10861, + "grad_norm": 1.2002004859159896, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 10861 + }, + { + "epoch": 0.10862, + "grad_norm": 1.2403320918171405, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 10862 + }, + { + "epoch": 0.10863, + "grad_norm": 0.9972196835269065, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 10863 + }, + { + "epoch": 0.10864, + "grad_norm": 1.387585591578004, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 10864 + }, + { + "epoch": 0.10865, + "grad_norm": 0.9421221076213664, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 10865 + }, + { + "epoch": 0.10866, + "grad_norm": 1.1948737529367885, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 10866 + }, + { + "epoch": 0.10867, + "grad_norm": 1.022212784674362, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 10867 + }, + { + "epoch": 0.10868, + "grad_norm": 1.2732584760819687, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 10868 + }, + { + "epoch": 0.10869, + "grad_norm": 1.1047154277724311, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 10869 + }, + { + "epoch": 0.1087, + "grad_norm": 1.1854424862823498, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 10870 + }, + { + "epoch": 0.10871, + "grad_norm": 1.1741520672533723, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 10871 + }, + { + "epoch": 0.10872, + "grad_norm": 1.0480602065379063, + "learning_rate": 0.003, + "loss": 4.018, + "step": 10872 + }, + { + "epoch": 0.10873, + "grad_norm": 1.2571567165899944, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 10873 + }, + { + "epoch": 0.10874, + "grad_norm": 1.337309331111073, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 10874 + }, + { + "epoch": 0.10875, + "grad_norm": 1.2565095414668002, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 10875 + }, + { + "epoch": 0.10876, + "grad_norm": 0.9744590907381745, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 10876 + }, + { + "epoch": 0.10877, + "grad_norm": 1.3835777093829644, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 10877 + }, + { + "epoch": 0.10878, + "grad_norm": 0.9441806514467737, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 10878 + }, + { + "epoch": 0.10879, + "grad_norm": 1.1451762360336781, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 10879 + }, + { + "epoch": 0.1088, + "grad_norm": 1.1541770622613978, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 10880 + }, + { + "epoch": 0.10881, + "grad_norm": 1.1458016315587218, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 10881 + }, + { + "epoch": 0.10882, + "grad_norm": 1.070149396095645, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 10882 + }, + { + "epoch": 0.10883, + "grad_norm": 1.177388796693508, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 10883 + }, + { + "epoch": 0.10884, + "grad_norm": 1.097237368443832, + "learning_rate": 0.003, + "loss": 4.03, + "step": 10884 + }, + { + "epoch": 0.10885, + "grad_norm": 1.200352068902575, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 10885 + }, + { + "epoch": 0.10886, + "grad_norm": 1.104974642016532, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 10886 + }, + { + "epoch": 0.10887, + "grad_norm": 1.2279008727498621, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 10887 + }, + { + "epoch": 0.10888, + "grad_norm": 1.202420390338381, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 10888 + }, + { + "epoch": 0.10889, + "grad_norm": 1.0304295601558235, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 10889 + }, + { + "epoch": 0.1089, + "grad_norm": 1.3568156975479342, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 10890 + }, + { + "epoch": 0.10891, + "grad_norm": 1.106559013336969, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10891 + }, + { + "epoch": 0.10892, + "grad_norm": 1.175561573120817, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 10892 + }, + { + "epoch": 0.10893, + "grad_norm": 1.02974471285148, + "learning_rate": 0.003, + "loss": 4.011, + "step": 10893 + }, + { + "epoch": 0.10894, + "grad_norm": 1.186667566372747, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 10894 + }, + { + "epoch": 0.10895, + "grad_norm": 0.9317722895107462, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 10895 + }, + { + "epoch": 0.10896, + "grad_norm": 1.1702513574970357, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 10896 + }, + { + "epoch": 0.10897, + "grad_norm": 1.2355323699421268, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 10897 + }, + { + "epoch": 0.10898, + "grad_norm": 1.0319769089185715, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 10898 + }, + { + "epoch": 0.10899, + "grad_norm": 1.1782040466486867, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 10899 + }, + { + "epoch": 0.109, + "grad_norm": 1.0467952536854728, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 10900 + }, + { + "epoch": 0.10901, + "grad_norm": 1.230063746087194, + "learning_rate": 0.003, + "loss": 4.023, + "step": 10901 + }, + { + "epoch": 0.10902, + "grad_norm": 1.256377588857215, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10902 + }, + { + "epoch": 0.10903, + "grad_norm": 1.1730572753174304, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 10903 + }, + { + "epoch": 0.10904, + "grad_norm": 1.1307723050871854, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 10904 + }, + { + "epoch": 0.10905, + "grad_norm": 1.1510216931005048, + "learning_rate": 0.003, + "loss": 4.006, + "step": 10905 + }, + { + "epoch": 0.10906, + "grad_norm": 1.0156123289857955, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 10906 + }, + { + "epoch": 0.10907, + "grad_norm": 1.298179684529916, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 10907 + }, + { + "epoch": 0.10908, + "grad_norm": 1.017892580324516, + "learning_rate": 0.003, + "loss": 4.026, + "step": 10908 + }, + { + "epoch": 0.10909, + "grad_norm": 1.4152613686138085, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 10909 + }, + { + "epoch": 0.1091, + "grad_norm": 0.9104381971108158, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 10910 + }, + { + "epoch": 0.10911, + "grad_norm": 1.1706988285504234, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 10911 + }, + { + "epoch": 0.10912, + "grad_norm": 1.1246788359272881, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 10912 + }, + { + "epoch": 0.10913, + "grad_norm": 1.2615350196980175, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 10913 + }, + { + "epoch": 0.10914, + "grad_norm": 1.0529171886939046, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 10914 + }, + { + "epoch": 0.10915, + "grad_norm": 1.181984560865992, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 10915 + }, + { + "epoch": 0.10916, + "grad_norm": 0.9398122380586809, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 10916 + }, + { + "epoch": 0.10917, + "grad_norm": 1.3986893076006122, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 10917 + }, + { + "epoch": 0.10918, + "grad_norm": 1.1386917109617494, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 10918 + }, + { + "epoch": 0.10919, + "grad_norm": 1.388526998048085, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 10919 + }, + { + "epoch": 0.1092, + "grad_norm": 0.9553367529462157, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 10920 + }, + { + "epoch": 0.10921, + "grad_norm": 1.1249097029410848, + "learning_rate": 0.003, + "loss": 4.04, + "step": 10921 + }, + { + "epoch": 0.10922, + "grad_norm": 1.1588772915067118, + "learning_rate": 0.003, + "loss": 4.036, + "step": 10922 + }, + { + "epoch": 0.10923, + "grad_norm": 1.1594370172230615, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 10923 + }, + { + "epoch": 0.10924, + "grad_norm": 1.1587584233463495, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 10924 + }, + { + "epoch": 0.10925, + "grad_norm": 1.199402284930254, + "learning_rate": 0.003, + "loss": 4.039, + "step": 10925 + }, + { + "epoch": 0.10926, + "grad_norm": 1.3780799925530476, + "learning_rate": 0.003, + "loss": 3.998, + "step": 10926 + }, + { + "epoch": 0.10927, + "grad_norm": 0.9547337146677488, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10927 + }, + { + "epoch": 0.10928, + "grad_norm": 1.2312732084033924, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 10928 + }, + { + "epoch": 0.10929, + "grad_norm": 1.124638240875568, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 10929 + }, + { + "epoch": 0.1093, + "grad_norm": 1.0996981207669076, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 10930 + }, + { + "epoch": 0.10931, + "grad_norm": 0.9997241567513053, + "learning_rate": 0.003, + "loss": 4.037, + "step": 10931 + }, + { + "epoch": 0.10932, + "grad_norm": 1.2572811468968264, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10932 + }, + { + "epoch": 0.10933, + "grad_norm": 1.235322426229629, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10933 + }, + { + "epoch": 0.10934, + "grad_norm": 1.203499600140372, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 10934 + }, + { + "epoch": 0.10935, + "grad_norm": 1.0560801016003458, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 10935 + }, + { + "epoch": 0.10936, + "grad_norm": 1.4290113078300473, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 10936 + }, + { + "epoch": 0.10937, + "grad_norm": 0.8838300191155496, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 10937 + }, + { + "epoch": 0.10938, + "grad_norm": 1.0970734661026584, + "learning_rate": 0.003, + "loss": 4.042, + "step": 10938 + }, + { + "epoch": 0.10939, + "grad_norm": 1.1495776401685247, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 10939 + }, + { + "epoch": 0.1094, + "grad_norm": 1.2065630154932683, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 10940 + }, + { + "epoch": 0.10941, + "grad_norm": 1.2374225552061804, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 10941 + }, + { + "epoch": 0.10942, + "grad_norm": 1.2524313252271473, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 10942 + }, + { + "epoch": 0.10943, + "grad_norm": 1.0258331148557027, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 10943 + }, + { + "epoch": 0.10944, + "grad_norm": 1.1367149424101854, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 10944 + }, + { + "epoch": 0.10945, + "grad_norm": 1.264314678398142, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10945 + }, + { + "epoch": 0.10946, + "grad_norm": 1.1470856240128633, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 10946 + }, + { + "epoch": 0.10947, + "grad_norm": 1.343562409477934, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 10947 + }, + { + "epoch": 0.10948, + "grad_norm": 1.1340548072968755, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 10948 + }, + { + "epoch": 0.10949, + "grad_norm": 1.1527721030646023, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 10949 + }, + { + "epoch": 0.1095, + "grad_norm": 1.0245060674231437, + "learning_rate": 0.003, + "loss": 4.006, + "step": 10950 + }, + { + "epoch": 0.10951, + "grad_norm": 1.094903481253029, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 10951 + }, + { + "epoch": 0.10952, + "grad_norm": 1.138549421081459, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 10952 + }, + { + "epoch": 0.10953, + "grad_norm": 1.1286975653189033, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 10953 + }, + { + "epoch": 0.10954, + "grad_norm": 1.1602190952130804, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 10954 + }, + { + "epoch": 0.10955, + "grad_norm": 1.2130192000896474, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 10955 + }, + { + "epoch": 0.10956, + "grad_norm": 1.2258014219563267, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 10956 + }, + { + "epoch": 0.10957, + "grad_norm": 1.0776761162714221, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 10957 + }, + { + "epoch": 0.10958, + "grad_norm": 1.3034903704468932, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 10958 + }, + { + "epoch": 0.10959, + "grad_norm": 1.0464894790504649, + "learning_rate": 0.003, + "loss": 4.034, + "step": 10959 + }, + { + "epoch": 0.1096, + "grad_norm": 1.14488623119123, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 10960 + }, + { + "epoch": 0.10961, + "grad_norm": 1.115487082003804, + "learning_rate": 0.003, + "loss": 4.016, + "step": 10961 + }, + { + "epoch": 0.10962, + "grad_norm": 1.2309886332824755, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 10962 + }, + { + "epoch": 0.10963, + "grad_norm": 1.1406460952529542, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 10963 + }, + { + "epoch": 0.10964, + "grad_norm": 1.2092082481970878, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 10964 + }, + { + "epoch": 0.10965, + "grad_norm": 1.1089678629170563, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 10965 + }, + { + "epoch": 0.10966, + "grad_norm": 1.453982552663511, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10966 + }, + { + "epoch": 0.10967, + "grad_norm": 1.0132525838939015, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 10967 + }, + { + "epoch": 0.10968, + "grad_norm": 1.2921121470526886, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 10968 + }, + { + "epoch": 0.10969, + "grad_norm": 1.0273381027526496, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 10969 + }, + { + "epoch": 0.1097, + "grad_norm": 1.402973935592333, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 10970 + }, + { + "epoch": 0.10971, + "grad_norm": 1.1233713113959605, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 10971 + }, + { + "epoch": 0.10972, + "grad_norm": 1.1972507884324581, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 10972 + }, + { + "epoch": 0.10973, + "grad_norm": 1.0689380384676346, + "learning_rate": 0.003, + "loss": 4.031, + "step": 10973 + }, + { + "epoch": 0.10974, + "grad_norm": 1.2935838341305652, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 10974 + }, + { + "epoch": 0.10975, + "grad_norm": 1.2148835363852442, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 10975 + }, + { + "epoch": 0.10976, + "grad_norm": 1.1196265735397568, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 10976 + }, + { + "epoch": 0.10977, + "grad_norm": 1.0766132416152376, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10977 + }, + { + "epoch": 0.10978, + "grad_norm": 1.262804962527681, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 10978 + }, + { + "epoch": 0.10979, + "grad_norm": 0.9073294804824369, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 10979 + }, + { + "epoch": 0.1098, + "grad_norm": 1.2605259139634963, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 10980 + }, + { + "epoch": 0.10981, + "grad_norm": 0.9680594140582308, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10981 + }, + { + "epoch": 0.10982, + "grad_norm": 1.164726097839411, + "learning_rate": 0.003, + "loss": 4.004, + "step": 10982 + }, + { + "epoch": 0.10983, + "grad_norm": 1.078100686604328, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 10983 + }, + { + "epoch": 0.10984, + "grad_norm": 1.27951824157926, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 10984 + }, + { + "epoch": 0.10985, + "grad_norm": 1.2483088905929303, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 10985 + }, + { + "epoch": 0.10986, + "grad_norm": 1.0764913660255253, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 10986 + }, + { + "epoch": 0.10987, + "grad_norm": 1.2325305120406893, + "learning_rate": 0.003, + "loss": 4.048, + "step": 10987 + }, + { + "epoch": 0.10988, + "grad_norm": 1.1074761108120266, + "learning_rate": 0.003, + "loss": 4.04, + "step": 10988 + }, + { + "epoch": 0.10989, + "grad_norm": 1.1438099942455424, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 10989 + }, + { + "epoch": 0.1099, + "grad_norm": 0.9531903580271849, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 10990 + }, + { + "epoch": 0.10991, + "grad_norm": 1.1261491344660406, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 10991 + }, + { + "epoch": 0.10992, + "grad_norm": 1.2458886185212843, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 10992 + }, + { + "epoch": 0.10993, + "grad_norm": 0.9909464766748146, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 10993 + }, + { + "epoch": 0.10994, + "grad_norm": 1.2594072221547077, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 10994 + }, + { + "epoch": 0.10995, + "grad_norm": 1.0286576604311943, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 10995 + }, + { + "epoch": 0.10996, + "grad_norm": 1.0879885186981424, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 10996 + }, + { + "epoch": 0.10997, + "grad_norm": 1.2251558646601735, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 10997 + }, + { + "epoch": 0.10998, + "grad_norm": 1.520992898306493, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 10998 + }, + { + "epoch": 0.10999, + "grad_norm": 0.9974930439447344, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 10999 + }, + { + "epoch": 0.11, + "grad_norm": 1.3984769150579184, + "learning_rate": 0.003, + "loss": 4.074, + "step": 11000 + }, + { + "epoch": 0.11001, + "grad_norm": 0.9811644529227863, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 11001 + }, + { + "epoch": 0.11002, + "grad_norm": 1.1450918449781147, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 11002 + }, + { + "epoch": 0.11003, + "grad_norm": 1.0480431049370729, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 11003 + }, + { + "epoch": 0.11004, + "grad_norm": 1.4695224754258749, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 11004 + }, + { + "epoch": 0.11005, + "grad_norm": 1.0012488506178907, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 11005 + }, + { + "epoch": 0.11006, + "grad_norm": 1.1097531270381813, + "learning_rate": 0.003, + "loss": 4.044, + "step": 11006 + }, + { + "epoch": 0.11007, + "grad_norm": 1.3457776983843623, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 11007 + }, + { + "epoch": 0.11008, + "grad_norm": 0.9945908240179561, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 11008 + }, + { + "epoch": 0.11009, + "grad_norm": 1.146031306723099, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 11009 + }, + { + "epoch": 0.1101, + "grad_norm": 1.3041512694151578, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 11010 + }, + { + "epoch": 0.11011, + "grad_norm": 1.0561894869979496, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 11011 + }, + { + "epoch": 0.11012, + "grad_norm": 1.2575408616044952, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 11012 + }, + { + "epoch": 0.11013, + "grad_norm": 0.9634053404348774, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 11013 + }, + { + "epoch": 0.11014, + "grad_norm": 1.1455573340874126, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 11014 + }, + { + "epoch": 0.11015, + "grad_norm": 1.0979373194946567, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11015 + }, + { + "epoch": 0.11016, + "grad_norm": 1.0909505512602309, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 11016 + }, + { + "epoch": 0.11017, + "grad_norm": 1.070802229909584, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 11017 + }, + { + "epoch": 0.11018, + "grad_norm": 1.0655397569106997, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 11018 + }, + { + "epoch": 0.11019, + "grad_norm": 1.2632535966366136, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 11019 + }, + { + "epoch": 0.1102, + "grad_norm": 0.958588750896633, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 11020 + }, + { + "epoch": 0.11021, + "grad_norm": 1.1602559223950129, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 11021 + }, + { + "epoch": 0.11022, + "grad_norm": 0.9786123700993407, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 11022 + }, + { + "epoch": 0.11023, + "grad_norm": 1.0897353532712468, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 11023 + }, + { + "epoch": 0.11024, + "grad_norm": 1.3783770276098104, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 11024 + }, + { + "epoch": 0.11025, + "grad_norm": 0.9920237745810896, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 11025 + }, + { + "epoch": 0.11026, + "grad_norm": 1.2990419941936762, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11026 + }, + { + "epoch": 0.11027, + "grad_norm": 1.1054453581494064, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 11027 + }, + { + "epoch": 0.11028, + "grad_norm": 1.1552917683687645, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 11028 + }, + { + "epoch": 0.11029, + "grad_norm": 1.1868727922087963, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 11029 + }, + { + "epoch": 0.1103, + "grad_norm": 0.9825451090946785, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 11030 + }, + { + "epoch": 0.11031, + "grad_norm": 1.1804760582921996, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 11031 + }, + { + "epoch": 0.11032, + "grad_norm": 1.0624183712164799, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 11032 + }, + { + "epoch": 0.11033, + "grad_norm": 1.279437929756929, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 11033 + }, + { + "epoch": 0.11034, + "grad_norm": 0.9050557877634389, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 11034 + }, + { + "epoch": 0.11035, + "grad_norm": 1.0758467424852527, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 11035 + }, + { + "epoch": 0.11036, + "grad_norm": 1.4004957620496878, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 11036 + }, + { + "epoch": 0.11037, + "grad_norm": 1.148007136705928, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 11037 + }, + { + "epoch": 0.11038, + "grad_norm": 1.1327536508610503, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 11038 + }, + { + "epoch": 0.11039, + "grad_norm": 1.1848546858736662, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 11039 + }, + { + "epoch": 0.1104, + "grad_norm": 1.323480908182648, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 11040 + }, + { + "epoch": 0.11041, + "grad_norm": 0.8971359841886503, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 11041 + }, + { + "epoch": 0.11042, + "grad_norm": 1.1769094334599115, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 11042 + }, + { + "epoch": 0.11043, + "grad_norm": 1.1666462489026705, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11043 + }, + { + "epoch": 0.11044, + "grad_norm": 1.2253457217850752, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 11044 + }, + { + "epoch": 0.11045, + "grad_norm": 1.1206251489359325, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 11045 + }, + { + "epoch": 0.11046, + "grad_norm": 1.3392782992158903, + "learning_rate": 0.003, + "loss": 4.019, + "step": 11046 + }, + { + "epoch": 0.11047, + "grad_norm": 1.0113482407477978, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 11047 + }, + { + "epoch": 0.11048, + "grad_norm": 1.2434741216349872, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 11048 + }, + { + "epoch": 0.11049, + "grad_norm": 1.0690557967832832, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 11049 + }, + { + "epoch": 0.1105, + "grad_norm": 1.2906453831596687, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 11050 + }, + { + "epoch": 0.11051, + "grad_norm": 1.049645251796059, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 11051 + }, + { + "epoch": 0.11052, + "grad_norm": 1.1022370813626103, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 11052 + }, + { + "epoch": 0.11053, + "grad_norm": 1.404602485307846, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11053 + }, + { + "epoch": 0.11054, + "grad_norm": 0.9778708210072532, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 11054 + }, + { + "epoch": 0.11055, + "grad_norm": 1.0849146668394676, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11055 + }, + { + "epoch": 0.11056, + "grad_norm": 1.3525665780096308, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 11056 + }, + { + "epoch": 0.11057, + "grad_norm": 0.8663546618979914, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 11057 + }, + { + "epoch": 0.11058, + "grad_norm": 0.9349712450004345, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 11058 + }, + { + "epoch": 0.11059, + "grad_norm": 1.2419152885907019, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 11059 + }, + { + "epoch": 0.1106, + "grad_norm": 1.2015131996587158, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 11060 + }, + { + "epoch": 0.11061, + "grad_norm": 1.1726168292055932, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 11061 + }, + { + "epoch": 0.11062, + "grad_norm": 1.235236596847503, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 11062 + }, + { + "epoch": 0.11063, + "grad_norm": 1.1430465957610831, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 11063 + }, + { + "epoch": 0.11064, + "grad_norm": 1.1694916513185551, + "learning_rate": 0.003, + "loss": 4.027, + "step": 11064 + }, + { + "epoch": 0.11065, + "grad_norm": 1.1843858564427805, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 11065 + }, + { + "epoch": 0.11066, + "grad_norm": 1.1731005960048595, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 11066 + }, + { + "epoch": 0.11067, + "grad_norm": 1.3610597815554937, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 11067 + }, + { + "epoch": 0.11068, + "grad_norm": 0.9388571621226179, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 11068 + }, + { + "epoch": 0.11069, + "grad_norm": 1.1845720962926707, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 11069 + }, + { + "epoch": 0.1107, + "grad_norm": 1.0184914945024592, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 11070 + }, + { + "epoch": 0.11071, + "grad_norm": 1.1695843280280722, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 11071 + }, + { + "epoch": 0.11072, + "grad_norm": 1.2124753448260535, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 11072 + }, + { + "epoch": 0.11073, + "grad_norm": 1.1600313038561658, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 11073 + }, + { + "epoch": 0.11074, + "grad_norm": 1.2553421574935741, + "learning_rate": 0.003, + "loss": 4.024, + "step": 11074 + }, + { + "epoch": 0.11075, + "grad_norm": 1.1166830328732873, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 11075 + }, + { + "epoch": 0.11076, + "grad_norm": 1.362951644650086, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 11076 + }, + { + "epoch": 0.11077, + "grad_norm": 1.0206627313768633, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 11077 + }, + { + "epoch": 0.11078, + "grad_norm": 1.137921297110709, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 11078 + }, + { + "epoch": 0.11079, + "grad_norm": 1.0636974748852701, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 11079 + }, + { + "epoch": 0.1108, + "grad_norm": 1.243205584369264, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 11080 + }, + { + "epoch": 0.11081, + "grad_norm": 1.1466850373196094, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 11081 + }, + { + "epoch": 0.11082, + "grad_norm": 1.0332780369799524, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 11082 + }, + { + "epoch": 0.11083, + "grad_norm": 1.1547474519658834, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 11083 + }, + { + "epoch": 0.11084, + "grad_norm": 1.2630821564699577, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 11084 + }, + { + "epoch": 0.11085, + "grad_norm": 0.9616550782921469, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 11085 + }, + { + "epoch": 0.11086, + "grad_norm": 1.0337612711666608, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 11086 + }, + { + "epoch": 0.11087, + "grad_norm": 1.246325517565779, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 11087 + }, + { + "epoch": 0.11088, + "grad_norm": 1.1210534817234898, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 11088 + }, + { + "epoch": 0.11089, + "grad_norm": 1.1210922280085747, + "learning_rate": 0.003, + "loss": 4.025, + "step": 11089 + }, + { + "epoch": 0.1109, + "grad_norm": 1.2076356656569776, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 11090 + }, + { + "epoch": 0.11091, + "grad_norm": 1.0505164908267228, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 11091 + }, + { + "epoch": 0.11092, + "grad_norm": 1.116407928167723, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11092 + }, + { + "epoch": 0.11093, + "grad_norm": 1.3679759693794233, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 11093 + }, + { + "epoch": 0.11094, + "grad_norm": 0.9149580813928061, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 11094 + }, + { + "epoch": 0.11095, + "grad_norm": 1.2127377272290825, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 11095 + }, + { + "epoch": 0.11096, + "grad_norm": 1.2654770177418224, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11096 + }, + { + "epoch": 0.11097, + "grad_norm": 1.005101859952365, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 11097 + }, + { + "epoch": 0.11098, + "grad_norm": 1.1559183971161302, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 11098 + }, + { + "epoch": 0.11099, + "grad_norm": 1.2125995577106417, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 11099 + }, + { + "epoch": 0.111, + "grad_norm": 1.1741835677043042, + "learning_rate": 0.003, + "loss": 4.014, + "step": 11100 + }, + { + "epoch": 0.11101, + "grad_norm": 1.0431649381898074, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 11101 + }, + { + "epoch": 0.11102, + "grad_norm": 1.1127524956437145, + "learning_rate": 0.003, + "loss": 4.019, + "step": 11102 + }, + { + "epoch": 0.11103, + "grad_norm": 1.4390556394886567, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 11103 + }, + { + "epoch": 0.11104, + "grad_norm": 0.8695810012178421, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 11104 + }, + { + "epoch": 0.11105, + "grad_norm": 1.2898692002393344, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 11105 + }, + { + "epoch": 0.11106, + "grad_norm": 0.9339284088375288, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 11106 + }, + { + "epoch": 0.11107, + "grad_norm": 1.002203567794351, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 11107 + }, + { + "epoch": 0.11108, + "grad_norm": 1.212964255808357, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 11108 + }, + { + "epoch": 0.11109, + "grad_norm": 1.1363148006340882, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11109 + }, + { + "epoch": 0.1111, + "grad_norm": 1.0654823793341264, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 11110 + }, + { + "epoch": 0.11111, + "grad_norm": 1.0410737532990986, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 11111 + }, + { + "epoch": 0.11112, + "grad_norm": 1.2755289571055364, + "learning_rate": 0.003, + "loss": 4.057, + "step": 11112 + }, + { + "epoch": 0.11113, + "grad_norm": 1.1301440513445171, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 11113 + }, + { + "epoch": 0.11114, + "grad_norm": 1.3210030487586704, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 11114 + }, + { + "epoch": 0.11115, + "grad_norm": 1.0405738153003652, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 11115 + }, + { + "epoch": 0.11116, + "grad_norm": 1.4674668716202037, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 11116 + }, + { + "epoch": 0.11117, + "grad_norm": 1.126768367155695, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 11117 + }, + { + "epoch": 0.11118, + "grad_norm": 1.4499818568572693, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 11118 + }, + { + "epoch": 0.11119, + "grad_norm": 0.8723773760283889, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 11119 + }, + { + "epoch": 0.1112, + "grad_norm": 0.9880766918243952, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 11120 + }, + { + "epoch": 0.11121, + "grad_norm": 1.233724451927589, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 11121 + }, + { + "epoch": 0.11122, + "grad_norm": 0.9592358462304813, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 11122 + }, + { + "epoch": 0.11123, + "grad_norm": 0.9374151178986565, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 11123 + }, + { + "epoch": 0.11124, + "grad_norm": 1.0746579089776649, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 11124 + }, + { + "epoch": 0.11125, + "grad_norm": 1.0272355456391729, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 11125 + }, + { + "epoch": 0.11126, + "grad_norm": 1.4696931867563632, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 11126 + }, + { + "epoch": 0.11127, + "grad_norm": 0.9306122996930648, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 11127 + }, + { + "epoch": 0.11128, + "grad_norm": 1.259426522382338, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 11128 + }, + { + "epoch": 0.11129, + "grad_norm": 0.9952764357013387, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 11129 + }, + { + "epoch": 0.1113, + "grad_norm": 1.2922884263779666, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 11130 + }, + { + "epoch": 0.11131, + "grad_norm": 1.0933870097789917, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 11131 + }, + { + "epoch": 0.11132, + "grad_norm": 1.222132471274821, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 11132 + }, + { + "epoch": 0.11133, + "grad_norm": 1.086913343875556, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 11133 + }, + { + "epoch": 0.11134, + "grad_norm": 1.2038589551456618, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 11134 + }, + { + "epoch": 0.11135, + "grad_norm": 1.1630263270645305, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11135 + }, + { + "epoch": 0.11136, + "grad_norm": 1.2109273926288262, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 11136 + }, + { + "epoch": 0.11137, + "grad_norm": 1.133767888763714, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 11137 + }, + { + "epoch": 0.11138, + "grad_norm": 1.3304986578383604, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 11138 + }, + { + "epoch": 0.11139, + "grad_norm": 1.0703377791647029, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 11139 + }, + { + "epoch": 0.1114, + "grad_norm": 1.2710614152570363, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 11140 + }, + { + "epoch": 0.11141, + "grad_norm": 1.23803074569592, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 11141 + }, + { + "epoch": 0.11142, + "grad_norm": 1.3481665548815753, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11142 + }, + { + "epoch": 0.11143, + "grad_norm": 1.07936134160658, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 11143 + }, + { + "epoch": 0.11144, + "grad_norm": 1.2330660101095827, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 11144 + }, + { + "epoch": 0.11145, + "grad_norm": 1.1001097752215603, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 11145 + }, + { + "epoch": 0.11146, + "grad_norm": 1.2439381573445438, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 11146 + }, + { + "epoch": 0.11147, + "grad_norm": 0.99219418207008, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 11147 + }, + { + "epoch": 0.11148, + "grad_norm": 1.2162762932312685, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 11148 + }, + { + "epoch": 0.11149, + "grad_norm": 1.1639544600650762, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 11149 + }, + { + "epoch": 0.1115, + "grad_norm": 1.1006116288127763, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 11150 + }, + { + "epoch": 0.11151, + "grad_norm": 1.2672886873563878, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 11151 + }, + { + "epoch": 0.11152, + "grad_norm": 0.910987023246572, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 11152 + }, + { + "epoch": 0.11153, + "grad_norm": 1.101946526056744, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 11153 + }, + { + "epoch": 0.11154, + "grad_norm": 1.0947773638822873, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 11154 + }, + { + "epoch": 0.11155, + "grad_norm": 1.2702655690371203, + "learning_rate": 0.003, + "loss": 4.041, + "step": 11155 + }, + { + "epoch": 0.11156, + "grad_norm": 1.007909951916138, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 11156 + }, + { + "epoch": 0.11157, + "grad_norm": 1.3319173530862718, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 11157 + }, + { + "epoch": 0.11158, + "grad_norm": 0.9207828461156834, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11158 + }, + { + "epoch": 0.11159, + "grad_norm": 1.197285438055082, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 11159 + }, + { + "epoch": 0.1116, + "grad_norm": 1.0433813495595061, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 11160 + }, + { + "epoch": 0.11161, + "grad_norm": 1.081328185673201, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 11161 + }, + { + "epoch": 0.11162, + "grad_norm": 1.2926210086878027, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 11162 + }, + { + "epoch": 0.11163, + "grad_norm": 0.9963164319297045, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 11163 + }, + { + "epoch": 0.11164, + "grad_norm": 1.3281076066936632, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 11164 + }, + { + "epoch": 0.11165, + "grad_norm": 0.9999767623058433, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 11165 + }, + { + "epoch": 0.11166, + "grad_norm": 1.3510910761006925, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 11166 + }, + { + "epoch": 0.11167, + "grad_norm": 0.9733408617008773, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 11167 + }, + { + "epoch": 0.11168, + "grad_norm": 1.247793495062047, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11168 + }, + { + "epoch": 0.11169, + "grad_norm": 0.8953094218520867, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 11169 + }, + { + "epoch": 0.1117, + "grad_norm": 1.1780557955764552, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 11170 + }, + { + "epoch": 0.11171, + "grad_norm": 1.1811278342129226, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 11171 + }, + { + "epoch": 0.11172, + "grad_norm": 1.2689331311338854, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 11172 + }, + { + "epoch": 0.11173, + "grad_norm": 1.1683719180902132, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 11173 + }, + { + "epoch": 0.11174, + "grad_norm": 1.0094485713003973, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 11174 + }, + { + "epoch": 0.11175, + "grad_norm": 1.3250881012136548, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 11175 + }, + { + "epoch": 0.11176, + "grad_norm": 1.0043723462137841, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 11176 + }, + { + "epoch": 0.11177, + "grad_norm": 1.1386272400480386, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 11177 + }, + { + "epoch": 0.11178, + "grad_norm": 1.2126593447984106, + "learning_rate": 0.003, + "loss": 4.049, + "step": 11178 + }, + { + "epoch": 0.11179, + "grad_norm": 1.2307687560202298, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11179 + }, + { + "epoch": 0.1118, + "grad_norm": 1.1602217088708568, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11180 + }, + { + "epoch": 0.11181, + "grad_norm": 1.057733354940848, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 11181 + }, + { + "epoch": 0.11182, + "grad_norm": 1.1720682053385487, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 11182 + }, + { + "epoch": 0.11183, + "grad_norm": 1.1436907781386656, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 11183 + }, + { + "epoch": 0.11184, + "grad_norm": 1.3870042277723742, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 11184 + }, + { + "epoch": 0.11185, + "grad_norm": 1.0936717631849686, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 11185 + }, + { + "epoch": 0.11186, + "grad_norm": 1.3772654953768122, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11186 + }, + { + "epoch": 0.11187, + "grad_norm": 1.1790597204228255, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 11187 + }, + { + "epoch": 0.11188, + "grad_norm": 1.1073857604366244, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 11188 + }, + { + "epoch": 0.11189, + "grad_norm": 1.130277787761296, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 11189 + }, + { + "epoch": 0.1119, + "grad_norm": 1.2313249191247733, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 11190 + }, + { + "epoch": 0.11191, + "grad_norm": 1.0068096645265725, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 11191 + }, + { + "epoch": 0.11192, + "grad_norm": 1.4926464646825244, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 11192 + }, + { + "epoch": 0.11193, + "grad_norm": 1.075022433453773, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 11193 + }, + { + "epoch": 0.11194, + "grad_norm": 1.097191178968125, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 11194 + }, + { + "epoch": 0.11195, + "grad_norm": 1.1810849460786332, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11195 + }, + { + "epoch": 0.11196, + "grad_norm": 1.1503825490398842, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 11196 + }, + { + "epoch": 0.11197, + "grad_norm": 1.0421171730999512, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11197 + }, + { + "epoch": 0.11198, + "grad_norm": 1.0783422809281649, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 11198 + }, + { + "epoch": 0.11199, + "grad_norm": 1.133756805734178, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 11199 + }, + { + "epoch": 0.112, + "grad_norm": 1.132914344900601, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 11200 + }, + { + "epoch": 0.11201, + "grad_norm": 1.3917889105852121, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 11201 + }, + { + "epoch": 0.11202, + "grad_norm": 0.9826420636753783, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11202 + }, + { + "epoch": 0.11203, + "grad_norm": 1.35378770158763, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11203 + }, + { + "epoch": 0.11204, + "grad_norm": 1.1631858332700105, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 11204 + }, + { + "epoch": 0.11205, + "grad_norm": 1.2921914277100586, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 11205 + }, + { + "epoch": 0.11206, + "grad_norm": 1.0898547835621788, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 11206 + }, + { + "epoch": 0.11207, + "grad_norm": 1.2652008774282912, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 11207 + }, + { + "epoch": 0.11208, + "grad_norm": 0.9934464327612474, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 11208 + }, + { + "epoch": 0.11209, + "grad_norm": 1.2684033075929582, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 11209 + }, + { + "epoch": 0.1121, + "grad_norm": 1.0481317867453344, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 11210 + }, + { + "epoch": 0.11211, + "grad_norm": 1.2804818902359865, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 11211 + }, + { + "epoch": 0.11212, + "grad_norm": 1.1905349539801156, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 11212 + }, + { + "epoch": 0.11213, + "grad_norm": 1.1425360307860808, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 11213 + }, + { + "epoch": 0.11214, + "grad_norm": 1.0912668598159907, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 11214 + }, + { + "epoch": 0.11215, + "grad_norm": 1.2077706236952057, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 11215 + }, + { + "epoch": 0.11216, + "grad_norm": 1.139473012377748, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 11216 + }, + { + "epoch": 0.11217, + "grad_norm": 1.2737949872278453, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 11217 + }, + { + "epoch": 0.11218, + "grad_norm": 1.1462612962051975, + "learning_rate": 0.003, + "loss": 4.033, + "step": 11218 + }, + { + "epoch": 0.11219, + "grad_norm": 1.0951419504062592, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 11219 + }, + { + "epoch": 0.1122, + "grad_norm": 1.4045206799945849, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 11220 + }, + { + "epoch": 0.11221, + "grad_norm": 0.950211704268758, + "learning_rate": 0.003, + "loss": 4.025, + "step": 11221 + }, + { + "epoch": 0.11222, + "grad_norm": 1.1123561273367195, + "learning_rate": 0.003, + "loss": 4.045, + "step": 11222 + }, + { + "epoch": 0.11223, + "grad_norm": 1.1138231860334924, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11223 + }, + { + "epoch": 0.11224, + "grad_norm": 1.33255470843806, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 11224 + }, + { + "epoch": 0.11225, + "grad_norm": 1.0201189422663082, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 11225 + }, + { + "epoch": 0.11226, + "grad_norm": 1.2296700017941067, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 11226 + }, + { + "epoch": 0.11227, + "grad_norm": 1.2361560337024013, + "learning_rate": 0.003, + "loss": 4.015, + "step": 11227 + }, + { + "epoch": 0.11228, + "grad_norm": 1.359208100581836, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 11228 + }, + { + "epoch": 0.11229, + "grad_norm": 0.9721364087537138, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11229 + }, + { + "epoch": 0.1123, + "grad_norm": 1.056450608774184, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 11230 + }, + { + "epoch": 0.11231, + "grad_norm": 1.1151470529423464, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11231 + }, + { + "epoch": 0.11232, + "grad_norm": 1.349357130637738, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 11232 + }, + { + "epoch": 0.11233, + "grad_norm": 1.2692980366831477, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 11233 + }, + { + "epoch": 0.11234, + "grad_norm": 1.0600042315155727, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 11234 + }, + { + "epoch": 0.11235, + "grad_norm": 0.9730948794785325, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 11235 + }, + { + "epoch": 0.11236, + "grad_norm": 1.2005508877119877, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 11236 + }, + { + "epoch": 0.11237, + "grad_norm": 1.2147190669139891, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 11237 + }, + { + "epoch": 0.11238, + "grad_norm": 1.1338750404710822, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 11238 + }, + { + "epoch": 0.11239, + "grad_norm": 1.138170848080884, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 11239 + }, + { + "epoch": 0.1124, + "grad_norm": 1.031187833440196, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 11240 + }, + { + "epoch": 0.11241, + "grad_norm": 1.2789914501541244, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 11241 + }, + { + "epoch": 0.11242, + "grad_norm": 1.2236110455515656, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 11242 + }, + { + "epoch": 0.11243, + "grad_norm": 1.1093808126477966, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 11243 + }, + { + "epoch": 0.11244, + "grad_norm": 1.288173452117589, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 11244 + }, + { + "epoch": 0.11245, + "grad_norm": 1.0828260500666966, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 11245 + }, + { + "epoch": 0.11246, + "grad_norm": 1.1567543117559596, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 11246 + }, + { + "epoch": 0.11247, + "grad_norm": 1.3948568681025382, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 11247 + }, + { + "epoch": 0.11248, + "grad_norm": 1.0116804641863193, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 11248 + }, + { + "epoch": 0.11249, + "grad_norm": 1.3046509496910825, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11249 + }, + { + "epoch": 0.1125, + "grad_norm": 0.9991437863954199, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 11250 + }, + { + "epoch": 0.11251, + "grad_norm": 1.6013468281261287, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11251 + }, + { + "epoch": 0.11252, + "grad_norm": 0.8687536172748702, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 11252 + }, + { + "epoch": 0.11253, + "grad_norm": 1.0983958258583018, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 11253 + }, + { + "epoch": 0.11254, + "grad_norm": 1.226302806954953, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 11254 + }, + { + "epoch": 0.11255, + "grad_norm": 1.0662642116642365, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 11255 + }, + { + "epoch": 0.11256, + "grad_norm": 1.243463378805967, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 11256 + }, + { + "epoch": 0.11257, + "grad_norm": 1.0726761813569627, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 11257 + }, + { + "epoch": 0.11258, + "grad_norm": 1.101242624393574, + "learning_rate": 0.003, + "loss": 4.025, + "step": 11258 + }, + { + "epoch": 0.11259, + "grad_norm": 1.2600622231655116, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 11259 + }, + { + "epoch": 0.1126, + "grad_norm": 0.9980365205693852, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 11260 + }, + { + "epoch": 0.11261, + "grad_norm": 1.2636683024025575, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 11261 + }, + { + "epoch": 0.11262, + "grad_norm": 0.91784299165787, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 11262 + }, + { + "epoch": 0.11263, + "grad_norm": 1.2597297519227169, + "learning_rate": 0.003, + "loss": 4.02, + "step": 11263 + }, + { + "epoch": 0.11264, + "grad_norm": 1.1932659497244533, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 11264 + }, + { + "epoch": 0.11265, + "grad_norm": 0.9816454818813927, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 11265 + }, + { + "epoch": 0.11266, + "grad_norm": 1.2284969289133585, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 11266 + }, + { + "epoch": 0.11267, + "grad_norm": 1.1261765260002052, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 11267 + }, + { + "epoch": 0.11268, + "grad_norm": 1.08799226517971, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 11268 + }, + { + "epoch": 0.11269, + "grad_norm": 1.1922244816252106, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 11269 + }, + { + "epoch": 0.1127, + "grad_norm": 0.9472450427826526, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 11270 + }, + { + "epoch": 0.11271, + "grad_norm": 1.1932608996712035, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 11271 + }, + { + "epoch": 0.11272, + "grad_norm": 1.1892191152480909, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 11272 + }, + { + "epoch": 0.11273, + "grad_norm": 1.2513453159910775, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 11273 + }, + { + "epoch": 0.11274, + "grad_norm": 1.2756382922479819, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 11274 + }, + { + "epoch": 0.11275, + "grad_norm": 1.0956380848789105, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 11275 + }, + { + "epoch": 0.11276, + "grad_norm": 1.166088340374118, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 11276 + }, + { + "epoch": 0.11277, + "grad_norm": 0.9706800035472809, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 11277 + }, + { + "epoch": 0.11278, + "grad_norm": 1.277416659504229, + "learning_rate": 0.003, + "loss": 4.053, + "step": 11278 + }, + { + "epoch": 0.11279, + "grad_norm": 1.1485716440991571, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 11279 + }, + { + "epoch": 0.1128, + "grad_norm": 1.2519360307904746, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 11280 + }, + { + "epoch": 0.11281, + "grad_norm": 1.1269013514569461, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 11281 + }, + { + "epoch": 0.11282, + "grad_norm": 1.2633224979713635, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 11282 + }, + { + "epoch": 0.11283, + "grad_norm": 0.9853227805426616, + "learning_rate": 0.003, + "loss": 4.03, + "step": 11283 + }, + { + "epoch": 0.11284, + "grad_norm": 1.3468725308834935, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11284 + }, + { + "epoch": 0.11285, + "grad_norm": 0.9626469410998565, + "learning_rate": 0.003, + "loss": 4.024, + "step": 11285 + }, + { + "epoch": 0.11286, + "grad_norm": 1.2746488595767842, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 11286 + }, + { + "epoch": 0.11287, + "grad_norm": 1.023668641008112, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 11287 + }, + { + "epoch": 0.11288, + "grad_norm": 1.2629926336246062, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 11288 + }, + { + "epoch": 0.11289, + "grad_norm": 1.1125863518225823, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11289 + }, + { + "epoch": 0.1129, + "grad_norm": 1.2085679726727696, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 11290 + }, + { + "epoch": 0.11291, + "grad_norm": 1.020769473245544, + "learning_rate": 0.003, + "loss": 4.047, + "step": 11291 + }, + { + "epoch": 0.11292, + "grad_norm": 1.5500356961334352, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 11292 + }, + { + "epoch": 0.11293, + "grad_norm": 1.1685657842109745, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 11293 + }, + { + "epoch": 0.11294, + "grad_norm": 1.3386643011461636, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 11294 + }, + { + "epoch": 0.11295, + "grad_norm": 1.1580125532963097, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 11295 + }, + { + "epoch": 0.11296, + "grad_norm": 1.0489120902824944, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 11296 + }, + { + "epoch": 0.11297, + "grad_norm": 1.103839461492198, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 11297 + }, + { + "epoch": 0.11298, + "grad_norm": 1.2294443402904203, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 11298 + }, + { + "epoch": 0.11299, + "grad_norm": 1.2428348430774223, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 11299 + }, + { + "epoch": 0.113, + "grad_norm": 1.188214219956524, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 11300 + }, + { + "epoch": 0.11301, + "grad_norm": 1.091582106196238, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 11301 + }, + { + "epoch": 0.11302, + "grad_norm": 0.9288460311574318, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 11302 + }, + { + "epoch": 0.11303, + "grad_norm": 1.4007101429225426, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 11303 + }, + { + "epoch": 0.11304, + "grad_norm": 1.0982722121566038, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 11304 + }, + { + "epoch": 0.11305, + "grad_norm": 1.2279359120748694, + "learning_rate": 0.003, + "loss": 4.012, + "step": 11305 + }, + { + "epoch": 0.11306, + "grad_norm": 1.0604973608241537, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 11306 + }, + { + "epoch": 0.11307, + "grad_norm": 1.1270170083873274, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 11307 + }, + { + "epoch": 0.11308, + "grad_norm": 1.0777135494150742, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 11308 + }, + { + "epoch": 0.11309, + "grad_norm": 1.3544518294662256, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 11309 + }, + { + "epoch": 0.1131, + "grad_norm": 0.9727494707141395, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 11310 + }, + { + "epoch": 0.11311, + "grad_norm": 1.3272901030217268, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 11311 + }, + { + "epoch": 0.11312, + "grad_norm": 1.2076378589872812, + "learning_rate": 0.003, + "loss": 4.036, + "step": 11312 + }, + { + "epoch": 0.11313, + "grad_norm": 1.1833510307690753, + "learning_rate": 0.003, + "loss": 4.016, + "step": 11313 + }, + { + "epoch": 0.11314, + "grad_norm": 1.2380294898068176, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 11314 + }, + { + "epoch": 0.11315, + "grad_norm": 0.9562005645157645, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 11315 + }, + { + "epoch": 0.11316, + "grad_norm": 1.2025637701941305, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11316 + }, + { + "epoch": 0.11317, + "grad_norm": 1.1779191244129885, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 11317 + }, + { + "epoch": 0.11318, + "grad_norm": 1.2236972961058097, + "learning_rate": 0.003, + "loss": 4.046, + "step": 11318 + }, + { + "epoch": 0.11319, + "grad_norm": 1.0549576319959617, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 11319 + }, + { + "epoch": 0.1132, + "grad_norm": 1.2105786067689994, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 11320 + }, + { + "epoch": 0.11321, + "grad_norm": 1.3397052436576664, + "learning_rate": 0.003, + "loss": 4.004, + "step": 11321 + }, + { + "epoch": 0.11322, + "grad_norm": 1.2893426572692654, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11322 + }, + { + "epoch": 0.11323, + "grad_norm": 0.9407951953734294, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 11323 + }, + { + "epoch": 0.11324, + "grad_norm": 1.177515193019718, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 11324 + }, + { + "epoch": 0.11325, + "grad_norm": 1.2549973557620873, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 11325 + }, + { + "epoch": 0.11326, + "grad_norm": 1.0497216278305863, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11326 + }, + { + "epoch": 0.11327, + "grad_norm": 1.2211928944892607, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 11327 + }, + { + "epoch": 0.11328, + "grad_norm": 1.1575302001609644, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 11328 + }, + { + "epoch": 0.11329, + "grad_norm": 1.353259362168543, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11329 + }, + { + "epoch": 0.1133, + "grad_norm": 1.1673317612996932, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 11330 + }, + { + "epoch": 0.11331, + "grad_norm": 1.110344451425524, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 11331 + }, + { + "epoch": 0.11332, + "grad_norm": 0.9832494841416515, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 11332 + }, + { + "epoch": 0.11333, + "grad_norm": 1.1341803958872791, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 11333 + }, + { + "epoch": 0.11334, + "grad_norm": 1.0713985717412855, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 11334 + }, + { + "epoch": 0.11335, + "grad_norm": 1.4947599748518885, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 11335 + }, + { + "epoch": 0.11336, + "grad_norm": 1.0176108077782644, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 11336 + }, + { + "epoch": 0.11337, + "grad_norm": 1.0630112841599184, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 11337 + }, + { + "epoch": 0.11338, + "grad_norm": 1.0818296952259, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 11338 + }, + { + "epoch": 0.11339, + "grad_norm": 1.2490039926297583, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 11339 + }, + { + "epoch": 0.1134, + "grad_norm": 1.094214947697734, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 11340 + }, + { + "epoch": 0.11341, + "grad_norm": 1.1634860925483297, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 11341 + }, + { + "epoch": 0.11342, + "grad_norm": 1.1087200576928156, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 11342 + }, + { + "epoch": 0.11343, + "grad_norm": 1.0264504151598612, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 11343 + }, + { + "epoch": 0.11344, + "grad_norm": 1.1105495732679151, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 11344 + }, + { + "epoch": 0.11345, + "grad_norm": 0.94706032963158, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 11345 + }, + { + "epoch": 0.11346, + "grad_norm": 1.2356018822595631, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 11346 + }, + { + "epoch": 0.11347, + "grad_norm": 1.1999112899120898, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 11347 + }, + { + "epoch": 0.11348, + "grad_norm": 1.123422414749545, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 11348 + }, + { + "epoch": 0.11349, + "grad_norm": 1.1936480899464423, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 11349 + }, + { + "epoch": 0.1135, + "grad_norm": 1.3146543934382486, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 11350 + }, + { + "epoch": 0.11351, + "grad_norm": 1.2860271723469208, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 11351 + }, + { + "epoch": 0.11352, + "grad_norm": 1.0410866520299848, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 11352 + }, + { + "epoch": 0.11353, + "grad_norm": 1.09507802259332, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 11353 + }, + { + "epoch": 0.11354, + "grad_norm": 1.2049447545318868, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 11354 + }, + { + "epoch": 0.11355, + "grad_norm": 1.1115701247127443, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 11355 + }, + { + "epoch": 0.11356, + "grad_norm": 1.235219941310057, + "learning_rate": 0.003, + "loss": 4.028, + "step": 11356 + }, + { + "epoch": 0.11357, + "grad_norm": 1.165360102724017, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 11357 + }, + { + "epoch": 0.11358, + "grad_norm": 1.3339213697650871, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 11358 + }, + { + "epoch": 0.11359, + "grad_norm": 0.972211541313835, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 11359 + }, + { + "epoch": 0.1136, + "grad_norm": 1.0879757952074331, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 11360 + }, + { + "epoch": 0.11361, + "grad_norm": 1.065115109105123, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 11361 + }, + { + "epoch": 0.11362, + "grad_norm": 1.24872681243813, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11362 + }, + { + "epoch": 0.11363, + "grad_norm": 1.072079099300088, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 11363 + }, + { + "epoch": 0.11364, + "grad_norm": 1.055551473984547, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 11364 + }, + { + "epoch": 0.11365, + "grad_norm": 1.2328441149042377, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 11365 + }, + { + "epoch": 0.11366, + "grad_norm": 1.0839732791788177, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 11366 + }, + { + "epoch": 0.11367, + "grad_norm": 1.1584220106389211, + "learning_rate": 0.003, + "loss": 4.0, + "step": 11367 + }, + { + "epoch": 0.11368, + "grad_norm": 1.0721506976287616, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 11368 + }, + { + "epoch": 0.11369, + "grad_norm": 1.3646659075049767, + "learning_rate": 0.003, + "loss": 4.043, + "step": 11369 + }, + { + "epoch": 0.1137, + "grad_norm": 1.1322442704014541, + "learning_rate": 0.003, + "loss": 4.05, + "step": 11370 + }, + { + "epoch": 0.11371, + "grad_norm": 1.0558050074955512, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 11371 + }, + { + "epoch": 0.11372, + "grad_norm": 1.1301617951796612, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 11372 + }, + { + "epoch": 0.11373, + "grad_norm": 0.9812153815857593, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11373 + }, + { + "epoch": 0.11374, + "grad_norm": 1.359681063168288, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 11374 + }, + { + "epoch": 0.11375, + "grad_norm": 1.1989683827617195, + "learning_rate": 0.003, + "loss": 4.041, + "step": 11375 + }, + { + "epoch": 0.11376, + "grad_norm": 0.9981750577047487, + "learning_rate": 0.003, + "loss": 4.061, + "step": 11376 + }, + { + "epoch": 0.11377, + "grad_norm": 1.2247933958516897, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 11377 + }, + { + "epoch": 0.11378, + "grad_norm": 1.219591085402563, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 11378 + }, + { + "epoch": 0.11379, + "grad_norm": 1.1739864083872635, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 11379 + }, + { + "epoch": 0.1138, + "grad_norm": 1.1150848974614544, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 11380 + }, + { + "epoch": 0.11381, + "grad_norm": 1.32267660459148, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 11381 + }, + { + "epoch": 0.11382, + "grad_norm": 1.19364357962154, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 11382 + }, + { + "epoch": 0.11383, + "grad_norm": 1.1210298550024658, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 11383 + }, + { + "epoch": 0.11384, + "grad_norm": 1.2945719303722119, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 11384 + }, + { + "epoch": 0.11385, + "grad_norm": 0.8328895137306691, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 11385 + }, + { + "epoch": 0.11386, + "grad_norm": 0.9686545135829541, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 11386 + }, + { + "epoch": 0.11387, + "grad_norm": 1.3533951425487776, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 11387 + }, + { + "epoch": 0.11388, + "grad_norm": 0.9410384710036733, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 11388 + }, + { + "epoch": 0.11389, + "grad_norm": 1.2925355506565999, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 11389 + }, + { + "epoch": 0.1139, + "grad_norm": 1.187327748759508, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 11390 + }, + { + "epoch": 0.11391, + "grad_norm": 1.2938651872263744, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 11391 + }, + { + "epoch": 0.11392, + "grad_norm": 1.1411249301402717, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 11392 + }, + { + "epoch": 0.11393, + "grad_norm": 1.1337183382980245, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 11393 + }, + { + "epoch": 0.11394, + "grad_norm": 1.0889052885198094, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 11394 + }, + { + "epoch": 0.11395, + "grad_norm": 1.1071420401026428, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 11395 + }, + { + "epoch": 0.11396, + "grad_norm": 0.9858679783214278, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 11396 + }, + { + "epoch": 0.11397, + "grad_norm": 1.2081766368555051, + "learning_rate": 0.003, + "loss": 4.017, + "step": 11397 + }, + { + "epoch": 0.11398, + "grad_norm": 1.1904796842320102, + "learning_rate": 0.003, + "loss": 4.003, + "step": 11398 + }, + { + "epoch": 0.11399, + "grad_norm": 1.1652838732036046, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 11399 + }, + { + "epoch": 0.114, + "grad_norm": 1.1994199478656578, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 11400 + }, + { + "epoch": 0.11401, + "grad_norm": 1.0582784467349977, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 11401 + }, + { + "epoch": 0.11402, + "grad_norm": 1.1358520028873116, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 11402 + }, + { + "epoch": 0.11403, + "grad_norm": 1.0954439597350976, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 11403 + }, + { + "epoch": 0.11404, + "grad_norm": 1.2934994659161199, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 11404 + }, + { + "epoch": 0.11405, + "grad_norm": 1.1821009379113836, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 11405 + }, + { + "epoch": 0.11406, + "grad_norm": 0.9323614807006916, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 11406 + }, + { + "epoch": 0.11407, + "grad_norm": 1.044288690455784, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 11407 + }, + { + "epoch": 0.11408, + "grad_norm": 1.288371712660551, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 11408 + }, + { + "epoch": 0.11409, + "grad_norm": 1.1013793581481566, + "learning_rate": 0.003, + "loss": 4.038, + "step": 11409 + }, + { + "epoch": 0.1141, + "grad_norm": 1.2399270262890425, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 11410 + }, + { + "epoch": 0.11411, + "grad_norm": 1.198910800893651, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 11411 + }, + { + "epoch": 0.11412, + "grad_norm": 1.2722748826966932, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11412 + }, + { + "epoch": 0.11413, + "grad_norm": 1.2237602676428068, + "learning_rate": 0.003, + "loss": 4.022, + "step": 11413 + }, + { + "epoch": 0.11414, + "grad_norm": 1.054751554499624, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 11414 + }, + { + "epoch": 0.11415, + "grad_norm": 1.3859608251730753, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 11415 + }, + { + "epoch": 0.11416, + "grad_norm": 0.9023833580320342, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 11416 + }, + { + "epoch": 0.11417, + "grad_norm": 1.163717032717626, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 11417 + }, + { + "epoch": 0.11418, + "grad_norm": 1.1635066782698764, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 11418 + }, + { + "epoch": 0.11419, + "grad_norm": 1.2785442673812115, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 11419 + }, + { + "epoch": 0.1142, + "grad_norm": 1.1457574905694237, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 11420 + }, + { + "epoch": 0.11421, + "grad_norm": 1.2404496422928504, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 11421 + }, + { + "epoch": 0.11422, + "grad_norm": 0.9198614788403456, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 11422 + }, + { + "epoch": 0.11423, + "grad_norm": 0.9771207760414389, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 11423 + }, + { + "epoch": 0.11424, + "grad_norm": 1.0105210455470817, + "learning_rate": 0.003, + "loss": 4.05, + "step": 11424 + }, + { + "epoch": 0.11425, + "grad_norm": 1.209588882831824, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 11425 + }, + { + "epoch": 0.11426, + "grad_norm": 1.0050650721043324, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 11426 + }, + { + "epoch": 0.11427, + "grad_norm": 1.2129381435966735, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 11427 + }, + { + "epoch": 0.11428, + "grad_norm": 1.1725659285639696, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 11428 + }, + { + "epoch": 0.11429, + "grad_norm": 1.3772603501943919, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 11429 + }, + { + "epoch": 0.1143, + "grad_norm": 1.1189640923482154, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 11430 + }, + { + "epoch": 0.11431, + "grad_norm": 1.0788431637790905, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 11431 + }, + { + "epoch": 0.11432, + "grad_norm": 1.3054445233916165, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 11432 + }, + { + "epoch": 0.11433, + "grad_norm": 1.0880911650168383, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 11433 + }, + { + "epoch": 0.11434, + "grad_norm": 1.1337028141735888, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 11434 + }, + { + "epoch": 0.11435, + "grad_norm": 1.1674027454792053, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 11435 + }, + { + "epoch": 0.11436, + "grad_norm": 1.0782480965612236, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11436 + }, + { + "epoch": 0.11437, + "grad_norm": 1.0494128681386032, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 11437 + }, + { + "epoch": 0.11438, + "grad_norm": 1.2824338283149883, + "learning_rate": 0.003, + "loss": 4.034, + "step": 11438 + }, + { + "epoch": 0.11439, + "grad_norm": 1.1874792640867795, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 11439 + }, + { + "epoch": 0.1144, + "grad_norm": 1.1122036447087196, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 11440 + }, + { + "epoch": 0.11441, + "grad_norm": 1.190227594977014, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 11441 + }, + { + "epoch": 0.11442, + "grad_norm": 1.270616577468654, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 11442 + }, + { + "epoch": 0.11443, + "grad_norm": 1.3070052964269097, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11443 + }, + { + "epoch": 0.11444, + "grad_norm": 1.03052198866508, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 11444 + }, + { + "epoch": 0.11445, + "grad_norm": 1.1390169230493974, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 11445 + }, + { + "epoch": 0.11446, + "grad_norm": 1.3887008590353196, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11446 + }, + { + "epoch": 0.11447, + "grad_norm": 1.3046931809126903, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 11447 + }, + { + "epoch": 0.11448, + "grad_norm": 1.1339224847291296, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 11448 + }, + { + "epoch": 0.11449, + "grad_norm": 1.2471236226671056, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 11449 + }, + { + "epoch": 0.1145, + "grad_norm": 1.1731443345300483, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 11450 + }, + { + "epoch": 0.11451, + "grad_norm": 1.051150483846822, + "learning_rate": 0.003, + "loss": 4.032, + "step": 11451 + }, + { + "epoch": 0.11452, + "grad_norm": 1.2092227172081997, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 11452 + }, + { + "epoch": 0.11453, + "grad_norm": 1.167343479702236, + "learning_rate": 0.003, + "loss": 4.02, + "step": 11453 + }, + { + "epoch": 0.11454, + "grad_norm": 1.3660105566993612, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 11454 + }, + { + "epoch": 0.11455, + "grad_norm": 1.1629557310949954, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 11455 + }, + { + "epoch": 0.11456, + "grad_norm": 1.1739663234451223, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 11456 + }, + { + "epoch": 0.11457, + "grad_norm": 1.1156521901007033, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 11457 + }, + { + "epoch": 0.11458, + "grad_norm": 1.2592133833338512, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 11458 + }, + { + "epoch": 0.11459, + "grad_norm": 0.9615728049014559, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11459 + }, + { + "epoch": 0.1146, + "grad_norm": 1.2024834313430042, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 11460 + }, + { + "epoch": 0.11461, + "grad_norm": 1.2257514801570328, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 11461 + }, + { + "epoch": 0.11462, + "grad_norm": 1.0895605253494356, + "learning_rate": 0.003, + "loss": 4.029, + "step": 11462 + }, + { + "epoch": 0.11463, + "grad_norm": 1.1193074344631706, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 11463 + }, + { + "epoch": 0.11464, + "grad_norm": 1.294479238346295, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 11464 + }, + { + "epoch": 0.11465, + "grad_norm": 1.2104511450000375, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 11465 + }, + { + "epoch": 0.11466, + "grad_norm": 1.3503066663860102, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 11466 + }, + { + "epoch": 0.11467, + "grad_norm": 1.0443750276791484, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 11467 + }, + { + "epoch": 0.11468, + "grad_norm": 1.0505892359693176, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 11468 + }, + { + "epoch": 0.11469, + "grad_norm": 1.2825643225762613, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11469 + }, + { + "epoch": 0.1147, + "grad_norm": 1.0516531724729168, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 11470 + }, + { + "epoch": 0.11471, + "grad_norm": 1.3346428910912866, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 11471 + }, + { + "epoch": 0.11472, + "grad_norm": 1.095097921047335, + "learning_rate": 0.003, + "loss": 4.02, + "step": 11472 + }, + { + "epoch": 0.11473, + "grad_norm": 1.0698418013171735, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 11473 + }, + { + "epoch": 0.11474, + "grad_norm": 1.248084189275581, + "learning_rate": 0.003, + "loss": 4.032, + "step": 11474 + }, + { + "epoch": 0.11475, + "grad_norm": 1.021325495043658, + "learning_rate": 0.003, + "loss": 4.004, + "step": 11475 + }, + { + "epoch": 0.11476, + "grad_norm": 1.1909200855488131, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 11476 + }, + { + "epoch": 0.11477, + "grad_norm": 0.8421493513707902, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 11477 + }, + { + "epoch": 0.11478, + "grad_norm": 1.0695750946042661, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 11478 + }, + { + "epoch": 0.11479, + "grad_norm": 1.2622542869695088, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 11479 + }, + { + "epoch": 0.1148, + "grad_norm": 1.092292616556263, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 11480 + }, + { + "epoch": 0.11481, + "grad_norm": 1.145024396755148, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 11481 + }, + { + "epoch": 0.11482, + "grad_norm": 1.2473154146510699, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 11482 + }, + { + "epoch": 0.11483, + "grad_norm": 1.2566705438815857, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 11483 + }, + { + "epoch": 0.11484, + "grad_norm": 1.0349608438987377, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 11484 + }, + { + "epoch": 0.11485, + "grad_norm": 1.4355317012427342, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 11485 + }, + { + "epoch": 0.11486, + "grad_norm": 0.9326495474339838, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 11486 + }, + { + "epoch": 0.11487, + "grad_norm": 1.192660410494905, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 11487 + }, + { + "epoch": 0.11488, + "grad_norm": 1.2529854252829626, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 11488 + }, + { + "epoch": 0.11489, + "grad_norm": 1.089982573108441, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 11489 + }, + { + "epoch": 0.1149, + "grad_norm": 1.0719423276675382, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11490 + }, + { + "epoch": 0.11491, + "grad_norm": 1.2350327715936045, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 11491 + }, + { + "epoch": 0.11492, + "grad_norm": 1.031561706435445, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 11492 + }, + { + "epoch": 0.11493, + "grad_norm": 1.2174218874516956, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 11493 + }, + { + "epoch": 0.11494, + "grad_norm": 1.1968611290413345, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11494 + }, + { + "epoch": 0.11495, + "grad_norm": 1.0535886969774535, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 11495 + }, + { + "epoch": 0.11496, + "grad_norm": 1.3975302549169397, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 11496 + }, + { + "epoch": 0.11497, + "grad_norm": 1.2029400300955522, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11497 + }, + { + "epoch": 0.11498, + "grad_norm": 1.234457060439203, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 11498 + }, + { + "epoch": 0.11499, + "grad_norm": 1.1060078725145015, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 11499 + }, + { + "epoch": 0.115, + "grad_norm": 1.2756853936300074, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 11500 + }, + { + "epoch": 0.11501, + "grad_norm": 1.2133184029419521, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 11501 + }, + { + "epoch": 0.11502, + "grad_norm": 1.0818298149149148, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 11502 + }, + { + "epoch": 0.11503, + "grad_norm": 1.2728423553765829, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 11503 + }, + { + "epoch": 0.11504, + "grad_norm": 1.059498569959539, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 11504 + }, + { + "epoch": 0.11505, + "grad_norm": 1.35683834617117, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 11505 + }, + { + "epoch": 0.11506, + "grad_norm": 0.9413672569378154, + "learning_rate": 0.003, + "loss": 4.052, + "step": 11506 + }, + { + "epoch": 0.11507, + "grad_norm": 0.9314304015475336, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 11507 + }, + { + "epoch": 0.11508, + "grad_norm": 1.2504519343724296, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11508 + }, + { + "epoch": 0.11509, + "grad_norm": 1.151992191983872, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11509 + }, + { + "epoch": 0.1151, + "grad_norm": 1.3329441485185418, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 11510 + }, + { + "epoch": 0.11511, + "grad_norm": 1.0288130767327575, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 11511 + }, + { + "epoch": 0.11512, + "grad_norm": 1.2555285931550986, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 11512 + }, + { + "epoch": 0.11513, + "grad_norm": 1.0795969059027257, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 11513 + }, + { + "epoch": 0.11514, + "grad_norm": 1.2133450305538067, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 11514 + }, + { + "epoch": 0.11515, + "grad_norm": 1.0929272209567709, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 11515 + }, + { + "epoch": 0.11516, + "grad_norm": 1.193595963865107, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 11516 + }, + { + "epoch": 0.11517, + "grad_norm": 1.1423550823219821, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 11517 + }, + { + "epoch": 0.11518, + "grad_norm": 1.2145234983955724, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 11518 + }, + { + "epoch": 0.11519, + "grad_norm": 1.1485471374131204, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 11519 + }, + { + "epoch": 0.1152, + "grad_norm": 1.1180316109358304, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 11520 + }, + { + "epoch": 0.11521, + "grad_norm": 1.2276448661153612, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 11521 + }, + { + "epoch": 0.11522, + "grad_norm": 1.175171062022422, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11522 + }, + { + "epoch": 0.11523, + "grad_norm": 1.3095154968379423, + "learning_rate": 0.003, + "loss": 4.039, + "step": 11523 + }, + { + "epoch": 0.11524, + "grad_norm": 1.1840175023993502, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 11524 + }, + { + "epoch": 0.11525, + "grad_norm": 1.2260241036155453, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 11525 + }, + { + "epoch": 0.11526, + "grad_norm": 1.2519788784999393, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 11526 + }, + { + "epoch": 0.11527, + "grad_norm": 1.2082041366767022, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 11527 + }, + { + "epoch": 0.11528, + "grad_norm": 1.070100717213944, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 11528 + }, + { + "epoch": 0.11529, + "grad_norm": 1.0301876336933986, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 11529 + }, + { + "epoch": 0.1153, + "grad_norm": 1.1693678818554514, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 11530 + }, + { + "epoch": 0.11531, + "grad_norm": 1.0247944374771782, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 11531 + }, + { + "epoch": 0.11532, + "grad_norm": 1.274822947052742, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 11532 + }, + { + "epoch": 0.11533, + "grad_norm": 0.9507147171376193, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 11533 + }, + { + "epoch": 0.11534, + "grad_norm": 1.3131155527752056, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 11534 + }, + { + "epoch": 0.11535, + "grad_norm": 1.1716121524993672, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 11535 + }, + { + "epoch": 0.11536, + "grad_norm": 1.3964603557745452, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 11536 + }, + { + "epoch": 0.11537, + "grad_norm": 1.0643667779288306, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11537 + }, + { + "epoch": 0.11538, + "grad_norm": 1.0516657590339504, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 11538 + }, + { + "epoch": 0.11539, + "grad_norm": 1.348470067591403, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 11539 + }, + { + "epoch": 0.1154, + "grad_norm": 0.9565756930424267, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 11540 + }, + { + "epoch": 0.11541, + "grad_norm": 1.2425855920676405, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 11541 + }, + { + "epoch": 0.11542, + "grad_norm": 0.9444027602069982, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 11542 + }, + { + "epoch": 0.11543, + "grad_norm": 1.121214808300316, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 11543 + }, + { + "epoch": 0.11544, + "grad_norm": 1.1647380685948656, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 11544 + }, + { + "epoch": 0.11545, + "grad_norm": 1.0922646545602452, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 11545 + }, + { + "epoch": 0.11546, + "grad_norm": 1.1959237319583436, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 11546 + }, + { + "epoch": 0.11547, + "grad_norm": 1.2168385051277104, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 11547 + }, + { + "epoch": 0.11548, + "grad_norm": 0.9906434040772247, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 11548 + }, + { + "epoch": 0.11549, + "grad_norm": 1.399388889519621, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 11549 + }, + { + "epoch": 0.1155, + "grad_norm": 1.1691281692028388, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 11550 + }, + { + "epoch": 0.11551, + "grad_norm": 1.2610503080435684, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 11551 + }, + { + "epoch": 0.11552, + "grad_norm": 1.0462781610867442, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 11552 + }, + { + "epoch": 0.11553, + "grad_norm": 1.3027344485563248, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 11553 + }, + { + "epoch": 0.11554, + "grad_norm": 0.9956780270740784, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 11554 + }, + { + "epoch": 0.11555, + "grad_norm": 1.4322465943251208, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11555 + }, + { + "epoch": 0.11556, + "grad_norm": 1.0957293213169021, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 11556 + }, + { + "epoch": 0.11557, + "grad_norm": 1.3308831178294134, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 11557 + }, + { + "epoch": 0.11558, + "grad_norm": 1.1715374522210913, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 11558 + }, + { + "epoch": 0.11559, + "grad_norm": 1.0505976322706607, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 11559 + }, + { + "epoch": 0.1156, + "grad_norm": 1.1380918415657428, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 11560 + }, + { + "epoch": 0.11561, + "grad_norm": 1.1609609825263985, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 11561 + }, + { + "epoch": 0.11562, + "grad_norm": 1.3050344178576219, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 11562 + }, + { + "epoch": 0.11563, + "grad_norm": 0.9380550906728504, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 11563 + }, + { + "epoch": 0.11564, + "grad_norm": 1.1946440621831265, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 11564 + }, + { + "epoch": 0.11565, + "grad_norm": 1.3080572257976502, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 11565 + }, + { + "epoch": 0.11566, + "grad_norm": 1.250707143374676, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 11566 + }, + { + "epoch": 0.11567, + "grad_norm": 1.2311864131520067, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 11567 + }, + { + "epoch": 0.11568, + "grad_norm": 1.311111463929809, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 11568 + }, + { + "epoch": 0.11569, + "grad_norm": 0.8868310857413003, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 11569 + }, + { + "epoch": 0.1157, + "grad_norm": 1.1027450881292422, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 11570 + }, + { + "epoch": 0.11571, + "grad_norm": 1.1161081563397963, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 11571 + }, + { + "epoch": 0.11572, + "grad_norm": 0.9089775562151213, + "learning_rate": 0.003, + "loss": 3.976, + "step": 11572 + }, + { + "epoch": 0.11573, + "grad_norm": 1.003694953680766, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 11573 + }, + { + "epoch": 0.11574, + "grad_norm": 1.3694504994237875, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 11574 + }, + { + "epoch": 0.11575, + "grad_norm": 1.1741369881532855, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11575 + }, + { + "epoch": 0.11576, + "grad_norm": 1.436922790332464, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 11576 + }, + { + "epoch": 0.11577, + "grad_norm": 0.953413832089103, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 11577 + }, + { + "epoch": 0.11578, + "grad_norm": 1.0815315504459275, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 11578 + }, + { + "epoch": 0.11579, + "grad_norm": 1.331353293169261, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 11579 + }, + { + "epoch": 0.1158, + "grad_norm": 1.2014803447115574, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 11580 + }, + { + "epoch": 0.11581, + "grad_norm": 1.307264485319865, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 11581 + }, + { + "epoch": 0.11582, + "grad_norm": 1.0474446025518378, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 11582 + }, + { + "epoch": 0.11583, + "grad_norm": 1.1339925874103132, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 11583 + }, + { + "epoch": 0.11584, + "grad_norm": 1.0857258271985775, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 11584 + }, + { + "epoch": 0.11585, + "grad_norm": 1.2405218315074111, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 11585 + }, + { + "epoch": 0.11586, + "grad_norm": 1.14159513185874, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 11586 + }, + { + "epoch": 0.11587, + "grad_norm": 1.171361791538229, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 11587 + }, + { + "epoch": 0.11588, + "grad_norm": 0.8984976334031136, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11588 + }, + { + "epoch": 0.11589, + "grad_norm": 1.0483798909522963, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 11589 + }, + { + "epoch": 0.1159, + "grad_norm": 1.3752191283722144, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 11590 + }, + { + "epoch": 0.11591, + "grad_norm": 0.9271251406957484, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11591 + }, + { + "epoch": 0.11592, + "grad_norm": 1.1839711004608278, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 11592 + }, + { + "epoch": 0.11593, + "grad_norm": 1.327326801192664, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 11593 + }, + { + "epoch": 0.11594, + "grad_norm": 1.1228264344894299, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 11594 + }, + { + "epoch": 0.11595, + "grad_norm": 1.3993924632845647, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 11595 + }, + { + "epoch": 0.11596, + "grad_norm": 0.9955718638323818, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 11596 + }, + { + "epoch": 0.11597, + "grad_norm": 1.2034336458201005, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 11597 + }, + { + "epoch": 0.11598, + "grad_norm": 1.0008043287213872, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 11598 + }, + { + "epoch": 0.11599, + "grad_norm": 1.386252254726889, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 11599 + }, + { + "epoch": 0.116, + "grad_norm": 0.8982486028150353, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 11600 + }, + { + "epoch": 0.11601, + "grad_norm": 1.184347234907959, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11601 + }, + { + "epoch": 0.11602, + "grad_norm": 1.1247974246340537, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 11602 + }, + { + "epoch": 0.11603, + "grad_norm": 1.1459727868895968, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 11603 + }, + { + "epoch": 0.11604, + "grad_norm": 1.200299318110692, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 11604 + }, + { + "epoch": 0.11605, + "grad_norm": 1.1221538219177465, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 11605 + }, + { + "epoch": 0.11606, + "grad_norm": 1.3896592029006039, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 11606 + }, + { + "epoch": 0.11607, + "grad_norm": 1.167856625001811, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 11607 + }, + { + "epoch": 0.11608, + "grad_norm": 1.2487032951275419, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11608 + }, + { + "epoch": 0.11609, + "grad_norm": 0.9357902119911607, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 11609 + }, + { + "epoch": 0.1161, + "grad_norm": 1.3783721164193854, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 11610 + }, + { + "epoch": 0.11611, + "grad_norm": 1.0580641425258053, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 11611 + }, + { + "epoch": 0.11612, + "grad_norm": 1.0737145006413418, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 11612 + }, + { + "epoch": 0.11613, + "grad_norm": 1.2046751425123334, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11613 + }, + { + "epoch": 0.11614, + "grad_norm": 1.1911854189282736, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 11614 + }, + { + "epoch": 0.11615, + "grad_norm": 1.2731578822836134, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 11615 + }, + { + "epoch": 0.11616, + "grad_norm": 1.486744540343558, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 11616 + }, + { + "epoch": 0.11617, + "grad_norm": 0.8968819287679518, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 11617 + }, + { + "epoch": 0.11618, + "grad_norm": 1.0047111452204807, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 11618 + }, + { + "epoch": 0.11619, + "grad_norm": 1.176206234770719, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 11619 + }, + { + "epoch": 0.1162, + "grad_norm": 1.2253326151523196, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 11620 + }, + { + "epoch": 0.11621, + "grad_norm": 1.1956657043524104, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 11621 + }, + { + "epoch": 0.11622, + "grad_norm": 0.9435279690248589, + "learning_rate": 0.003, + "loss": 4.035, + "step": 11622 + }, + { + "epoch": 0.11623, + "grad_norm": 1.056882142632435, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 11623 + }, + { + "epoch": 0.11624, + "grad_norm": 1.1106539056889895, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 11624 + }, + { + "epoch": 0.11625, + "grad_norm": 1.0205099002386708, + "learning_rate": 0.003, + "loss": 4.011, + "step": 11625 + }, + { + "epoch": 0.11626, + "grad_norm": 1.258461909905812, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11626 + }, + { + "epoch": 0.11627, + "grad_norm": 1.2874361312056184, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11627 + }, + { + "epoch": 0.11628, + "grad_norm": 1.1743309619531834, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 11628 + }, + { + "epoch": 0.11629, + "grad_norm": 1.2243600857600132, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 11629 + }, + { + "epoch": 0.1163, + "grad_norm": 1.0499602492021, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 11630 + }, + { + "epoch": 0.11631, + "grad_norm": 1.2573505771904938, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 11631 + }, + { + "epoch": 0.11632, + "grad_norm": 1.2413025869166792, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 11632 + }, + { + "epoch": 0.11633, + "grad_norm": 1.4540436444418998, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 11633 + }, + { + "epoch": 0.11634, + "grad_norm": 0.9548533573974899, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 11634 + }, + { + "epoch": 0.11635, + "grad_norm": 1.3071307922264086, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 11635 + }, + { + "epoch": 0.11636, + "grad_norm": 1.03525412132836, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 11636 + }, + { + "epoch": 0.11637, + "grad_norm": 1.2135127098907612, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 11637 + }, + { + "epoch": 0.11638, + "grad_norm": 1.3223760615299698, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11638 + }, + { + "epoch": 0.11639, + "grad_norm": 0.9699089271963789, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 11639 + }, + { + "epoch": 0.1164, + "grad_norm": 1.2811922369560198, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 11640 + }, + { + "epoch": 0.11641, + "grad_norm": 1.036699916635101, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 11641 + }, + { + "epoch": 0.11642, + "grad_norm": 1.116072859573732, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 11642 + }, + { + "epoch": 0.11643, + "grad_norm": 1.1737632356245493, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 11643 + }, + { + "epoch": 0.11644, + "grad_norm": 1.105651145301224, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 11644 + }, + { + "epoch": 0.11645, + "grad_norm": 1.1206281967173415, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 11645 + }, + { + "epoch": 0.11646, + "grad_norm": 1.2792254145674709, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 11646 + }, + { + "epoch": 0.11647, + "grad_norm": 1.1472479706618874, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 11647 + }, + { + "epoch": 0.11648, + "grad_norm": 1.3828212140161604, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 11648 + }, + { + "epoch": 0.11649, + "grad_norm": 1.1798512117694304, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 11649 + }, + { + "epoch": 0.1165, + "grad_norm": 1.1784795673577373, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 11650 + }, + { + "epoch": 0.11651, + "grad_norm": 1.10280548005316, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 11651 + }, + { + "epoch": 0.11652, + "grad_norm": 1.2084602073133586, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 11652 + }, + { + "epoch": 0.11653, + "grad_norm": 1.1644747868763254, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 11653 + }, + { + "epoch": 0.11654, + "grad_norm": 1.1934077001382124, + "learning_rate": 0.003, + "loss": 4.004, + "step": 11654 + }, + { + "epoch": 0.11655, + "grad_norm": 0.9579468700893862, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 11655 + }, + { + "epoch": 0.11656, + "grad_norm": 1.2244849985241453, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 11656 + }, + { + "epoch": 0.11657, + "grad_norm": 1.2201292215580293, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 11657 + }, + { + "epoch": 0.11658, + "grad_norm": 1.008379460556807, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 11658 + }, + { + "epoch": 0.11659, + "grad_norm": 1.177586242231204, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11659 + }, + { + "epoch": 0.1166, + "grad_norm": 1.0891812559060503, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 11660 + }, + { + "epoch": 0.11661, + "grad_norm": 1.3079283675775348, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 11661 + }, + { + "epoch": 0.11662, + "grad_norm": 1.0842584854688044, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 11662 + }, + { + "epoch": 0.11663, + "grad_norm": 1.0420405487472146, + "learning_rate": 0.003, + "loss": 4.004, + "step": 11663 + }, + { + "epoch": 0.11664, + "grad_norm": 1.3155796441920706, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 11664 + }, + { + "epoch": 0.11665, + "grad_norm": 1.152392525026187, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 11665 + }, + { + "epoch": 0.11666, + "grad_norm": 1.217393356467726, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 11666 + }, + { + "epoch": 0.11667, + "grad_norm": 1.3174299861589436, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 11667 + }, + { + "epoch": 0.11668, + "grad_norm": 1.2379665248144123, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11668 + }, + { + "epoch": 0.11669, + "grad_norm": 1.2462022619969693, + "learning_rate": 0.003, + "loss": 4.031, + "step": 11669 + }, + { + "epoch": 0.1167, + "grad_norm": 1.0813016807485878, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11670 + }, + { + "epoch": 0.11671, + "grad_norm": 1.581906417209871, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 11671 + }, + { + "epoch": 0.11672, + "grad_norm": 0.8933295146366401, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 11672 + }, + { + "epoch": 0.11673, + "grad_norm": 1.3454666952243168, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 11673 + }, + { + "epoch": 0.11674, + "grad_norm": 1.254430804934053, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11674 + }, + { + "epoch": 0.11675, + "grad_norm": 1.1407305796856528, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 11675 + }, + { + "epoch": 0.11676, + "grad_norm": 1.1240412766616164, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 11676 + }, + { + "epoch": 0.11677, + "grad_norm": 1.1039857268337923, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 11677 + }, + { + "epoch": 0.11678, + "grad_norm": 1.1382453808865138, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 11678 + }, + { + "epoch": 0.11679, + "grad_norm": 1.1297436535746768, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 11679 + }, + { + "epoch": 0.1168, + "grad_norm": 1.3532016024947384, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 11680 + }, + { + "epoch": 0.11681, + "grad_norm": 0.924056872724392, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 11681 + }, + { + "epoch": 0.11682, + "grad_norm": 1.3574144651574285, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 11682 + }, + { + "epoch": 0.11683, + "grad_norm": 0.9768276300170212, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 11683 + }, + { + "epoch": 0.11684, + "grad_norm": 1.2699611558933386, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 11684 + }, + { + "epoch": 0.11685, + "grad_norm": 1.2501909792103931, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 11685 + }, + { + "epoch": 0.11686, + "grad_norm": 1.1022171786236412, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 11686 + }, + { + "epoch": 0.11687, + "grad_norm": 1.3814616040574366, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 11687 + }, + { + "epoch": 0.11688, + "grad_norm": 0.9287871350220055, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 11688 + }, + { + "epoch": 0.11689, + "grad_norm": 1.2281964388187652, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 11689 + }, + { + "epoch": 0.1169, + "grad_norm": 1.1127225127033677, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 11690 + }, + { + "epoch": 0.11691, + "grad_norm": 1.2909281509347093, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 11691 + }, + { + "epoch": 0.11692, + "grad_norm": 1.3619954406747572, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 11692 + }, + { + "epoch": 0.11693, + "grad_norm": 1.1057049514943214, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 11693 + }, + { + "epoch": 0.11694, + "grad_norm": 1.1428819960705745, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 11694 + }, + { + "epoch": 0.11695, + "grad_norm": 1.1667755384935443, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 11695 + }, + { + "epoch": 0.11696, + "grad_norm": 1.4469227506294298, + "learning_rate": 0.003, + "loss": 4.037, + "step": 11696 + }, + { + "epoch": 0.11697, + "grad_norm": 1.005450115587486, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 11697 + }, + { + "epoch": 0.11698, + "grad_norm": 1.2238213012885069, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 11698 + }, + { + "epoch": 0.11699, + "grad_norm": 1.054385789261448, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 11699 + }, + { + "epoch": 0.117, + "grad_norm": 1.122631863282793, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 11700 + }, + { + "epoch": 0.11701, + "grad_norm": 1.0125507091879444, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 11701 + }, + { + "epoch": 0.11702, + "grad_norm": 1.4368068506374105, + "learning_rate": 0.003, + "loss": 4.029, + "step": 11702 + }, + { + "epoch": 0.11703, + "grad_norm": 1.1065201692933635, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 11703 + }, + { + "epoch": 0.11704, + "grad_norm": 1.1390282672127734, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 11704 + }, + { + "epoch": 0.11705, + "grad_norm": 1.2237420373219998, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 11705 + }, + { + "epoch": 0.11706, + "grad_norm": 1.2701077700750183, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 11706 + }, + { + "epoch": 0.11707, + "grad_norm": 1.0572688186092254, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 11707 + }, + { + "epoch": 0.11708, + "grad_norm": 1.1488670670780203, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 11708 + }, + { + "epoch": 0.11709, + "grad_norm": 1.157519213062066, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 11709 + }, + { + "epoch": 0.1171, + "grad_norm": 1.3123712860081451, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11710 + }, + { + "epoch": 0.11711, + "grad_norm": 0.9842670475281874, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 11711 + }, + { + "epoch": 0.11712, + "grad_norm": 1.1200872070851704, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 11712 + }, + { + "epoch": 0.11713, + "grad_norm": 1.1207768888268173, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 11713 + }, + { + "epoch": 0.11714, + "grad_norm": 1.2468830968691615, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 11714 + }, + { + "epoch": 0.11715, + "grad_norm": 1.1733299447516492, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 11715 + }, + { + "epoch": 0.11716, + "grad_norm": 1.0093206271659385, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 11716 + }, + { + "epoch": 0.11717, + "grad_norm": 1.1239353189243093, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 11717 + }, + { + "epoch": 0.11718, + "grad_norm": 1.233944823509884, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 11718 + }, + { + "epoch": 0.11719, + "grad_norm": 1.2985477914439674, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 11719 + }, + { + "epoch": 0.1172, + "grad_norm": 1.1827958279910649, + "learning_rate": 0.003, + "loss": 4.012, + "step": 11720 + }, + { + "epoch": 0.11721, + "grad_norm": 1.1446307963150169, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 11721 + }, + { + "epoch": 0.11722, + "grad_norm": 1.2455474825934287, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 11722 + }, + { + "epoch": 0.11723, + "grad_norm": 1.11475125208383, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 11723 + }, + { + "epoch": 0.11724, + "grad_norm": 1.1489869640108812, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 11724 + }, + { + "epoch": 0.11725, + "grad_norm": 1.148474380544057, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 11725 + }, + { + "epoch": 0.11726, + "grad_norm": 1.1836759594720816, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 11726 + }, + { + "epoch": 0.11727, + "grad_norm": 1.1489901497136943, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 11727 + }, + { + "epoch": 0.11728, + "grad_norm": 1.2213324589045766, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 11728 + }, + { + "epoch": 0.11729, + "grad_norm": 1.2375612775771767, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 11729 + }, + { + "epoch": 0.1173, + "grad_norm": 1.0576264778912117, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 11730 + }, + { + "epoch": 0.11731, + "grad_norm": 1.1994319511524283, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 11731 + }, + { + "epoch": 0.11732, + "grad_norm": 1.1470271295502747, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 11732 + }, + { + "epoch": 0.11733, + "grad_norm": 1.0321703975881884, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11733 + }, + { + "epoch": 0.11734, + "grad_norm": 1.1708692520929904, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 11734 + }, + { + "epoch": 0.11735, + "grad_norm": 1.1217915693347118, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 11735 + }, + { + "epoch": 0.11736, + "grad_norm": 1.1484409584760604, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 11736 + }, + { + "epoch": 0.11737, + "grad_norm": 1.1665894926579745, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11737 + }, + { + "epoch": 0.11738, + "grad_norm": 1.2813020571044718, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 11738 + }, + { + "epoch": 0.11739, + "grad_norm": 1.3181678380644914, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 11739 + }, + { + "epoch": 0.1174, + "grad_norm": 1.120161267354906, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 11740 + }, + { + "epoch": 0.11741, + "grad_norm": 1.4288394398505233, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 11741 + }, + { + "epoch": 0.11742, + "grad_norm": 1.1845167399331413, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 11742 + }, + { + "epoch": 0.11743, + "grad_norm": 1.2099165641421654, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 11743 + }, + { + "epoch": 0.11744, + "grad_norm": 1.1904859574947253, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 11744 + }, + { + "epoch": 0.11745, + "grad_norm": 1.2576951023850673, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 11745 + }, + { + "epoch": 0.11746, + "grad_norm": 1.0964969976409622, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 11746 + }, + { + "epoch": 0.11747, + "grad_norm": 1.4120869523116033, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 11747 + }, + { + "epoch": 0.11748, + "grad_norm": 1.1148049335571633, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 11748 + }, + { + "epoch": 0.11749, + "grad_norm": 1.1734647930140643, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 11749 + }, + { + "epoch": 0.1175, + "grad_norm": 1.2482405487706174, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 11750 + }, + { + "epoch": 0.11751, + "grad_norm": 1.1483514385526685, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 11751 + }, + { + "epoch": 0.11752, + "grad_norm": 1.2500272890819837, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 11752 + }, + { + "epoch": 0.11753, + "grad_norm": 1.058210157236084, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 11753 + }, + { + "epoch": 0.11754, + "grad_norm": 1.1487736631750007, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11754 + }, + { + "epoch": 0.11755, + "grad_norm": 1.4715443707390885, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 11755 + }, + { + "epoch": 0.11756, + "grad_norm": 1.2303628287318422, + "learning_rate": 0.003, + "loss": 4.04, + "step": 11756 + }, + { + "epoch": 0.11757, + "grad_norm": 1.2922823982550529, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 11757 + }, + { + "epoch": 0.11758, + "grad_norm": 1.0379265137390694, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 11758 + }, + { + "epoch": 0.11759, + "grad_norm": 1.3406374253364324, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 11759 + }, + { + "epoch": 0.1176, + "grad_norm": 1.1304813742700188, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 11760 + }, + { + "epoch": 0.11761, + "grad_norm": 1.2685923899639884, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11761 + }, + { + "epoch": 0.11762, + "grad_norm": 1.1593920836432694, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 11762 + }, + { + "epoch": 0.11763, + "grad_norm": 1.1762381398269397, + "learning_rate": 0.003, + "loss": 4.024, + "step": 11763 + }, + { + "epoch": 0.11764, + "grad_norm": 1.082858055046518, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 11764 + }, + { + "epoch": 0.11765, + "grad_norm": 1.191615642444593, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 11765 + }, + { + "epoch": 0.11766, + "grad_norm": 1.236828958505827, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 11766 + }, + { + "epoch": 0.11767, + "grad_norm": 1.293279569430646, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 11767 + }, + { + "epoch": 0.11768, + "grad_norm": 1.3077046844773013, + "learning_rate": 0.003, + "loss": 4.028, + "step": 11768 + }, + { + "epoch": 0.11769, + "grad_norm": 1.0549809352357689, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 11769 + }, + { + "epoch": 0.1177, + "grad_norm": 1.0689765908038116, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11770 + }, + { + "epoch": 0.11771, + "grad_norm": 1.3866554981109134, + "learning_rate": 0.003, + "loss": 4.006, + "step": 11771 + }, + { + "epoch": 0.11772, + "grad_norm": 1.182097917949292, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 11772 + }, + { + "epoch": 0.11773, + "grad_norm": 1.2642222535276253, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 11773 + }, + { + "epoch": 0.11774, + "grad_norm": 0.9460972097814158, + "learning_rate": 0.003, + "loss": 4.023, + "step": 11774 + }, + { + "epoch": 0.11775, + "grad_norm": 1.1967480636343066, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 11775 + }, + { + "epoch": 0.11776, + "grad_norm": 0.9332242169274172, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 11776 + }, + { + "epoch": 0.11777, + "grad_norm": 1.394104158897881, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11777 + }, + { + "epoch": 0.11778, + "grad_norm": 0.8763693888556454, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 11778 + }, + { + "epoch": 0.11779, + "grad_norm": 1.1878650279892633, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 11779 + }, + { + "epoch": 0.1178, + "grad_norm": 1.2171141033704371, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11780 + }, + { + "epoch": 0.11781, + "grad_norm": 1.160898330979206, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 11781 + }, + { + "epoch": 0.11782, + "grad_norm": 1.3998726789246247, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 11782 + }, + { + "epoch": 0.11783, + "grad_norm": 1.274867591692867, + "learning_rate": 0.003, + "loss": 4.054, + "step": 11783 + }, + { + "epoch": 0.11784, + "grad_norm": 1.122679163756025, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11784 + }, + { + "epoch": 0.11785, + "grad_norm": 1.350965890967207, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 11785 + }, + { + "epoch": 0.11786, + "grad_norm": 1.1259362312209875, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 11786 + }, + { + "epoch": 0.11787, + "grad_norm": 1.1001601351081598, + "learning_rate": 0.003, + "loss": 4.044, + "step": 11787 + }, + { + "epoch": 0.11788, + "grad_norm": 1.117549418935662, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 11788 + }, + { + "epoch": 0.11789, + "grad_norm": 0.9414233692391143, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 11789 + }, + { + "epoch": 0.1179, + "grad_norm": 1.0644099136425493, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 11790 + }, + { + "epoch": 0.11791, + "grad_norm": 1.2857169916877873, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 11791 + }, + { + "epoch": 0.11792, + "grad_norm": 0.9110294228440342, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11792 + }, + { + "epoch": 0.11793, + "grad_norm": 1.1296405959354765, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 11793 + }, + { + "epoch": 0.11794, + "grad_norm": 1.098095724914582, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 11794 + }, + { + "epoch": 0.11795, + "grad_norm": 1.326195921928846, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 11795 + }, + { + "epoch": 0.11796, + "grad_norm": 0.9378070394703966, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 11796 + }, + { + "epoch": 0.11797, + "grad_norm": 1.0521633920909828, + "learning_rate": 0.003, + "loss": 4.008, + "step": 11797 + }, + { + "epoch": 0.11798, + "grad_norm": 1.1384690047707438, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 11798 + }, + { + "epoch": 0.11799, + "grad_norm": 1.2149860412806044, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 11799 + }, + { + "epoch": 0.118, + "grad_norm": 1.189248780058123, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11800 + }, + { + "epoch": 0.11801, + "grad_norm": 1.2211735751337582, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 11801 + }, + { + "epoch": 0.11802, + "grad_norm": 1.3801254418581266, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 11802 + }, + { + "epoch": 0.11803, + "grad_norm": 1.3818090508225513, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 11803 + }, + { + "epoch": 0.11804, + "grad_norm": 1.1283876629557728, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 11804 + }, + { + "epoch": 0.11805, + "grad_norm": 1.3132843310339615, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 11805 + }, + { + "epoch": 0.11806, + "grad_norm": 1.0209548234461399, + "learning_rate": 0.003, + "loss": 4.025, + "step": 11806 + }, + { + "epoch": 0.11807, + "grad_norm": 1.4212793934895955, + "learning_rate": 0.003, + "loss": 4.036, + "step": 11807 + }, + { + "epoch": 0.11808, + "grad_norm": 0.9218401109406803, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 11808 + }, + { + "epoch": 0.11809, + "grad_norm": 1.3124286668656089, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 11809 + }, + { + "epoch": 0.1181, + "grad_norm": 1.0303704151552795, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 11810 + }, + { + "epoch": 0.11811, + "grad_norm": 1.2557024156725904, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 11811 + }, + { + "epoch": 0.11812, + "grad_norm": 1.092659260298407, + "learning_rate": 0.003, + "loss": 4.034, + "step": 11812 + }, + { + "epoch": 0.11813, + "grad_norm": 1.0226098263465304, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 11813 + }, + { + "epoch": 0.11814, + "grad_norm": 1.251101082721384, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11814 + }, + { + "epoch": 0.11815, + "grad_norm": 1.1762980524741449, + "learning_rate": 0.003, + "loss": 4.043, + "step": 11815 + }, + { + "epoch": 0.11816, + "grad_norm": 1.1188651716671465, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 11816 + }, + { + "epoch": 0.11817, + "grad_norm": 0.914012325190764, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 11817 + }, + { + "epoch": 0.11818, + "grad_norm": 1.1572820201025285, + "learning_rate": 0.003, + "loss": 4.03, + "step": 11818 + }, + { + "epoch": 0.11819, + "grad_norm": 1.1097633407410619, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 11819 + }, + { + "epoch": 0.1182, + "grad_norm": 1.1315572070408753, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11820 + }, + { + "epoch": 0.11821, + "grad_norm": 1.2010201734662294, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 11821 + }, + { + "epoch": 0.11822, + "grad_norm": 1.206264909595148, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 11822 + }, + { + "epoch": 0.11823, + "grad_norm": 1.3220064172019355, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 11823 + }, + { + "epoch": 0.11824, + "grad_norm": 1.2144802838958413, + "learning_rate": 0.003, + "loss": 4.023, + "step": 11824 + }, + { + "epoch": 0.11825, + "grad_norm": 1.1598148282919596, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 11825 + }, + { + "epoch": 0.11826, + "grad_norm": 1.1618791360531502, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11826 + }, + { + "epoch": 0.11827, + "grad_norm": 1.1833304595150924, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 11827 + }, + { + "epoch": 0.11828, + "grad_norm": 1.1074286497081403, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 11828 + }, + { + "epoch": 0.11829, + "grad_norm": 1.4539648581372682, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 11829 + }, + { + "epoch": 0.1183, + "grad_norm": 1.071045817248875, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 11830 + }, + { + "epoch": 0.11831, + "grad_norm": 1.317805477765449, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 11831 + }, + { + "epoch": 0.11832, + "grad_norm": 1.129955427413547, + "learning_rate": 0.003, + "loss": 4.011, + "step": 11832 + }, + { + "epoch": 0.11833, + "grad_norm": 1.1888617474099106, + "learning_rate": 0.003, + "loss": 4.056, + "step": 11833 + }, + { + "epoch": 0.11834, + "grad_norm": 1.0290127949757615, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 11834 + }, + { + "epoch": 0.11835, + "grad_norm": 1.346252405762597, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 11835 + }, + { + "epoch": 0.11836, + "grad_norm": 1.0669477339564317, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 11836 + }, + { + "epoch": 0.11837, + "grad_norm": 1.2528885489785642, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11837 + }, + { + "epoch": 0.11838, + "grad_norm": 1.2202450943926373, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 11838 + }, + { + "epoch": 0.11839, + "grad_norm": 1.1291740083749862, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 11839 + }, + { + "epoch": 0.1184, + "grad_norm": 0.973213189261728, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 11840 + }, + { + "epoch": 0.11841, + "grad_norm": 1.3080093889925444, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 11841 + }, + { + "epoch": 0.11842, + "grad_norm": 1.0994079309142295, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 11842 + }, + { + "epoch": 0.11843, + "grad_norm": 1.2562568817457294, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 11843 + }, + { + "epoch": 0.11844, + "grad_norm": 0.9846686490918987, + "learning_rate": 0.003, + "loss": 4.006, + "step": 11844 + }, + { + "epoch": 0.11845, + "grad_norm": 1.1930430438444906, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 11845 + }, + { + "epoch": 0.11846, + "grad_norm": 1.2231257490273924, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 11846 + }, + { + "epoch": 0.11847, + "grad_norm": 1.1539088743932866, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 11847 + }, + { + "epoch": 0.11848, + "grad_norm": 1.0931038156263568, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 11848 + }, + { + "epoch": 0.11849, + "grad_norm": 0.9927389994471185, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 11849 + }, + { + "epoch": 0.1185, + "grad_norm": 1.0974827020855389, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 11850 + }, + { + "epoch": 0.11851, + "grad_norm": 1.1400193029050951, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 11851 + }, + { + "epoch": 0.11852, + "grad_norm": 1.1260088424765735, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 11852 + }, + { + "epoch": 0.11853, + "grad_norm": 1.203276342374635, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11853 + }, + { + "epoch": 0.11854, + "grad_norm": 1.2713596037206674, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 11854 + }, + { + "epoch": 0.11855, + "grad_norm": 1.0995469030408622, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 11855 + }, + { + "epoch": 0.11856, + "grad_norm": 1.526076675901606, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 11856 + }, + { + "epoch": 0.11857, + "grad_norm": 0.9423722463191191, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 11857 + }, + { + "epoch": 0.11858, + "grad_norm": 1.2391248570515367, + "learning_rate": 0.003, + "loss": 4.058, + "step": 11858 + }, + { + "epoch": 0.11859, + "grad_norm": 1.1811635655252974, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11859 + }, + { + "epoch": 0.1186, + "grad_norm": 1.2433780055764707, + "learning_rate": 0.003, + "loss": 4.024, + "step": 11860 + }, + { + "epoch": 0.11861, + "grad_norm": 1.1675448550905765, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 11861 + }, + { + "epoch": 0.11862, + "grad_norm": 1.522196016713748, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11862 + }, + { + "epoch": 0.11863, + "grad_norm": 0.8786368309806317, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 11863 + }, + { + "epoch": 0.11864, + "grad_norm": 1.189025993197089, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 11864 + }, + { + "epoch": 0.11865, + "grad_norm": 1.1997885236483348, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 11865 + }, + { + "epoch": 0.11866, + "grad_norm": 1.1371332429372165, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 11866 + }, + { + "epoch": 0.11867, + "grad_norm": 0.9698227780141195, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 11867 + }, + { + "epoch": 0.11868, + "grad_norm": 1.103930407918838, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 11868 + }, + { + "epoch": 0.11869, + "grad_norm": 1.302679313668843, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 11869 + }, + { + "epoch": 0.1187, + "grad_norm": 1.2640687849220409, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 11870 + }, + { + "epoch": 0.11871, + "grad_norm": 1.0418025168606888, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 11871 + }, + { + "epoch": 0.11872, + "grad_norm": 1.179394195887056, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 11872 + }, + { + "epoch": 0.11873, + "grad_norm": 1.000011524256758, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 11873 + }, + { + "epoch": 0.11874, + "grad_norm": 1.1944779241728058, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 11874 + }, + { + "epoch": 0.11875, + "grad_norm": 1.0502451220356457, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11875 + }, + { + "epoch": 0.11876, + "grad_norm": 1.296683390399838, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 11876 + }, + { + "epoch": 0.11877, + "grad_norm": 0.9584019380860173, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 11877 + }, + { + "epoch": 0.11878, + "grad_norm": 1.4464987936871119, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11878 + }, + { + "epoch": 0.11879, + "grad_norm": 1.0641775759664829, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 11879 + }, + { + "epoch": 0.1188, + "grad_norm": 1.1780706118325417, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 11880 + }, + { + "epoch": 0.11881, + "grad_norm": 1.1247991069312226, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 11881 + }, + { + "epoch": 0.11882, + "grad_norm": 1.1967564947594456, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 11882 + }, + { + "epoch": 0.11883, + "grad_norm": 1.1953315126688369, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 11883 + }, + { + "epoch": 0.11884, + "grad_norm": 1.1616352501275355, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 11884 + }, + { + "epoch": 0.11885, + "grad_norm": 1.2509035326277451, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 11885 + }, + { + "epoch": 0.11886, + "grad_norm": 1.0576045189685839, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 11886 + }, + { + "epoch": 0.11887, + "grad_norm": 1.298046893222415, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 11887 + }, + { + "epoch": 0.11888, + "grad_norm": 0.9499057577771591, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 11888 + }, + { + "epoch": 0.11889, + "grad_norm": 1.4214495434215852, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 11889 + }, + { + "epoch": 0.1189, + "grad_norm": 1.303172172126755, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 11890 + }, + { + "epoch": 0.11891, + "grad_norm": 1.066754223369426, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 11891 + }, + { + "epoch": 0.11892, + "grad_norm": 1.1813142465709445, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11892 + }, + { + "epoch": 0.11893, + "grad_norm": 1.1318573053901286, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 11893 + }, + { + "epoch": 0.11894, + "grad_norm": 1.330891602268722, + "learning_rate": 0.003, + "loss": 4.01, + "step": 11894 + }, + { + "epoch": 0.11895, + "grad_norm": 0.9313185148188895, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 11895 + }, + { + "epoch": 0.11896, + "grad_norm": 1.2289925546102853, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 11896 + }, + { + "epoch": 0.11897, + "grad_norm": 1.3072846874199877, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 11897 + }, + { + "epoch": 0.11898, + "grad_norm": 1.1161842409655773, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 11898 + }, + { + "epoch": 0.11899, + "grad_norm": 1.2666190795801895, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 11899 + }, + { + "epoch": 0.119, + "grad_norm": 1.1826671258994363, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 11900 + }, + { + "epoch": 0.11901, + "grad_norm": 1.137766702745759, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 11901 + }, + { + "epoch": 0.11902, + "grad_norm": 1.3498299757086438, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 11902 + }, + { + "epoch": 0.11903, + "grad_norm": 1.0286391095482723, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 11903 + }, + { + "epoch": 0.11904, + "grad_norm": 1.291367620845929, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 11904 + }, + { + "epoch": 0.11905, + "grad_norm": 1.0161169842231146, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 11905 + }, + { + "epoch": 0.11906, + "grad_norm": 1.363924299039333, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 11906 + }, + { + "epoch": 0.11907, + "grad_norm": 0.9352441437117577, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 11907 + }, + { + "epoch": 0.11908, + "grad_norm": 1.3625840410584036, + "learning_rate": 0.003, + "loss": 4.049, + "step": 11908 + }, + { + "epoch": 0.11909, + "grad_norm": 1.277088311088804, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 11909 + }, + { + "epoch": 0.1191, + "grad_norm": 1.212682027964726, + "learning_rate": 0.003, + "loss": 4.041, + "step": 11910 + }, + { + "epoch": 0.11911, + "grad_norm": 1.0288245659041568, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 11911 + }, + { + "epoch": 0.11912, + "grad_norm": 1.2480548583938877, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 11912 + }, + { + "epoch": 0.11913, + "grad_norm": 0.9880713186954828, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 11913 + }, + { + "epoch": 0.11914, + "grad_norm": 1.3610335200499963, + "learning_rate": 0.003, + "loss": 4.058, + "step": 11914 + }, + { + "epoch": 0.11915, + "grad_norm": 1.180493316410989, + "learning_rate": 0.003, + "loss": 4.039, + "step": 11915 + }, + { + "epoch": 0.11916, + "grad_norm": 1.2400050091012045, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 11916 + }, + { + "epoch": 0.11917, + "grad_norm": 1.136714338075228, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 11917 + }, + { + "epoch": 0.11918, + "grad_norm": 1.2965300537848186, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 11918 + }, + { + "epoch": 0.11919, + "grad_norm": 1.1067424063285383, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 11919 + }, + { + "epoch": 0.1192, + "grad_norm": 1.143087763102736, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 11920 + }, + { + "epoch": 0.11921, + "grad_norm": 1.1278594645317526, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 11921 + }, + { + "epoch": 0.11922, + "grad_norm": 1.1065993088756547, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 11922 + }, + { + "epoch": 0.11923, + "grad_norm": 1.0404477273555774, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 11923 + }, + { + "epoch": 0.11924, + "grad_norm": 1.098544390943939, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 11924 + }, + { + "epoch": 0.11925, + "grad_norm": 1.261762237162695, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 11925 + }, + { + "epoch": 0.11926, + "grad_norm": 1.3263357716743485, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 11926 + }, + { + "epoch": 0.11927, + "grad_norm": 1.3069181491866044, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 11927 + }, + { + "epoch": 0.11928, + "grad_norm": 1.2597036992613604, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 11928 + }, + { + "epoch": 0.11929, + "grad_norm": 1.1978030782695692, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11929 + }, + { + "epoch": 0.1193, + "grad_norm": 1.122796948660304, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 11930 + }, + { + "epoch": 0.11931, + "grad_norm": 1.28639539805541, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 11931 + }, + { + "epoch": 0.11932, + "grad_norm": 0.8855138744393674, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 11932 + }, + { + "epoch": 0.11933, + "grad_norm": 1.237861600157439, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 11933 + }, + { + "epoch": 0.11934, + "grad_norm": 1.0400155549854835, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 11934 + }, + { + "epoch": 0.11935, + "grad_norm": 1.253240275222296, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 11935 + }, + { + "epoch": 0.11936, + "grad_norm": 1.1208436384532008, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 11936 + }, + { + "epoch": 0.11937, + "grad_norm": 1.188206798413594, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 11937 + }, + { + "epoch": 0.11938, + "grad_norm": 1.1880639987904473, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 11938 + }, + { + "epoch": 0.11939, + "grad_norm": 1.1272675744243077, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 11939 + }, + { + "epoch": 0.1194, + "grad_norm": 1.4203426618154515, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 11940 + }, + { + "epoch": 0.11941, + "grad_norm": 0.8873694901506768, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 11941 + }, + { + "epoch": 0.11942, + "grad_norm": 1.1932380771885094, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 11942 + }, + { + "epoch": 0.11943, + "grad_norm": 1.172479996445272, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 11943 + }, + { + "epoch": 0.11944, + "grad_norm": 1.197483536826607, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 11944 + }, + { + "epoch": 0.11945, + "grad_norm": 1.367144021054736, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 11945 + }, + { + "epoch": 0.11946, + "grad_norm": 1.1658588956904559, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 11946 + }, + { + "epoch": 0.11947, + "grad_norm": 1.0672867133410382, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 11947 + }, + { + "epoch": 0.11948, + "grad_norm": 1.217906799639158, + "learning_rate": 0.003, + "loss": 4.029, + "step": 11948 + }, + { + "epoch": 0.11949, + "grad_norm": 1.10277312621074, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 11949 + }, + { + "epoch": 0.1195, + "grad_norm": 1.3388512544054778, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 11950 + }, + { + "epoch": 0.11951, + "grad_norm": 1.1385910637331862, + "learning_rate": 0.003, + "loss": 4.023, + "step": 11951 + }, + { + "epoch": 0.11952, + "grad_norm": 1.129078560681352, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 11952 + }, + { + "epoch": 0.11953, + "grad_norm": 1.3014140128951632, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 11953 + }, + { + "epoch": 0.11954, + "grad_norm": 1.0937861343588238, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 11954 + }, + { + "epoch": 0.11955, + "grad_norm": 1.3146945090315645, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 11955 + }, + { + "epoch": 0.11956, + "grad_norm": 1.0561747096686331, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 11956 + }, + { + "epoch": 0.11957, + "grad_norm": 1.3427476069726214, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 11957 + }, + { + "epoch": 0.11958, + "grad_norm": 0.9106335166241142, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11958 + }, + { + "epoch": 0.11959, + "grad_norm": 1.2516591304108617, + "learning_rate": 0.003, + "loss": 4.004, + "step": 11959 + }, + { + "epoch": 0.1196, + "grad_norm": 1.2013239303032632, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11960 + }, + { + "epoch": 0.11961, + "grad_norm": 1.177656487506444, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 11961 + }, + { + "epoch": 0.11962, + "grad_norm": 1.2925731396377893, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 11962 + }, + { + "epoch": 0.11963, + "grad_norm": 1.2500654665973316, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 11963 + }, + { + "epoch": 0.11964, + "grad_norm": 1.0425860200509207, + "learning_rate": 0.003, + "loss": 4.017, + "step": 11964 + }, + { + "epoch": 0.11965, + "grad_norm": 1.2377261274781197, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 11965 + }, + { + "epoch": 0.11966, + "grad_norm": 1.2032049076648743, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 11966 + }, + { + "epoch": 0.11967, + "grad_norm": 1.1633394155622445, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 11967 + }, + { + "epoch": 0.11968, + "grad_norm": 1.319184772941729, + "learning_rate": 0.003, + "loss": 4.007, + "step": 11968 + }, + { + "epoch": 0.11969, + "grad_norm": 1.2615950720517894, + "learning_rate": 0.003, + "loss": 4.043, + "step": 11969 + }, + { + "epoch": 0.1197, + "grad_norm": 1.2523901831861972, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 11970 + }, + { + "epoch": 0.11971, + "grad_norm": 1.0498669279565727, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 11971 + }, + { + "epoch": 0.11972, + "grad_norm": 1.2175500601101223, + "learning_rate": 0.003, + "loss": 4.001, + "step": 11972 + }, + { + "epoch": 0.11973, + "grad_norm": 1.130504073169973, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 11973 + }, + { + "epoch": 0.11974, + "grad_norm": 1.179537557381126, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 11974 + }, + { + "epoch": 0.11975, + "grad_norm": 1.2289925798077932, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 11975 + }, + { + "epoch": 0.11976, + "grad_norm": 1.030004757594503, + "learning_rate": 0.003, + "loss": 4.009, + "step": 11976 + }, + { + "epoch": 0.11977, + "grad_norm": 1.3000185859855784, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 11977 + }, + { + "epoch": 0.11978, + "grad_norm": 1.0874027837487101, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 11978 + }, + { + "epoch": 0.11979, + "grad_norm": 1.2673027617231618, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 11979 + }, + { + "epoch": 0.1198, + "grad_norm": 1.1353539513587794, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 11980 + }, + { + "epoch": 0.11981, + "grad_norm": 1.2446619182105716, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 11981 + }, + { + "epoch": 0.11982, + "grad_norm": 1.28462029530281, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 11982 + }, + { + "epoch": 0.11983, + "grad_norm": 1.0633275265167388, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 11983 + }, + { + "epoch": 0.11984, + "grad_norm": 1.2422112244323555, + "learning_rate": 0.003, + "loss": 4.014, + "step": 11984 + }, + { + "epoch": 0.11985, + "grad_norm": 1.0976759058442642, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 11985 + }, + { + "epoch": 0.11986, + "grad_norm": 1.2409565112326744, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 11986 + }, + { + "epoch": 0.11987, + "grad_norm": 1.091733743800522, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 11987 + }, + { + "epoch": 0.11988, + "grad_norm": 1.0535364843904362, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 11988 + }, + { + "epoch": 0.11989, + "grad_norm": 1.2117698380224653, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 11989 + }, + { + "epoch": 0.1199, + "grad_norm": 1.2107541460176305, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 11990 + }, + { + "epoch": 0.11991, + "grad_norm": 1.128519381242994, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 11991 + }, + { + "epoch": 0.11992, + "grad_norm": 1.2912774558797513, + "learning_rate": 0.003, + "loss": 4.053, + "step": 11992 + }, + { + "epoch": 0.11993, + "grad_norm": 0.9859836123745601, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 11993 + }, + { + "epoch": 0.11994, + "grad_norm": 1.4062953714578372, + "learning_rate": 0.003, + "loss": 4.038, + "step": 11994 + }, + { + "epoch": 0.11995, + "grad_norm": 0.9854278363001263, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 11995 + }, + { + "epoch": 0.11996, + "grad_norm": 1.1283327652944204, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 11996 + }, + { + "epoch": 0.11997, + "grad_norm": 1.2578860491282189, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 11997 + }, + { + "epoch": 0.11998, + "grad_norm": 1.2201013514042887, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 11998 + }, + { + "epoch": 0.11999, + "grad_norm": 1.0464428472868936, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 11999 + }, + { + "epoch": 0.12, + "grad_norm": 1.1442643913849708, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 12000 + }, + { + "epoch": 0.12001, + "grad_norm": 1.3301044375260127, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 12001 + }, + { + "epoch": 0.12002, + "grad_norm": 1.105906653032294, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 12002 + }, + { + "epoch": 0.12003, + "grad_norm": 1.2210250779728278, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 12003 + }, + { + "epoch": 0.12004, + "grad_norm": 1.1362570687956641, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 12004 + }, + { + "epoch": 0.12005, + "grad_norm": 1.2598481775607295, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 12005 + }, + { + "epoch": 0.12006, + "grad_norm": 1.238198018149289, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 12006 + }, + { + "epoch": 0.12007, + "grad_norm": 1.329447272103125, + "learning_rate": 0.003, + "loss": 4.027, + "step": 12007 + }, + { + "epoch": 0.12008, + "grad_norm": 1.1041434627768059, + "learning_rate": 0.003, + "loss": 4.05, + "step": 12008 + }, + { + "epoch": 0.12009, + "grad_norm": 1.0703280836160836, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 12009 + }, + { + "epoch": 0.1201, + "grad_norm": 1.1943502028864001, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 12010 + }, + { + "epoch": 0.12011, + "grad_norm": 1.0239676315199393, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 12011 + }, + { + "epoch": 0.12012, + "grad_norm": 1.4503121638562684, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 12012 + }, + { + "epoch": 0.12013, + "grad_norm": 1.015108621263654, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 12013 + }, + { + "epoch": 0.12014, + "grad_norm": 1.389962594266883, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 12014 + }, + { + "epoch": 0.12015, + "grad_norm": 1.1252082891848212, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 12015 + }, + { + "epoch": 0.12016, + "grad_norm": 1.0487983474875753, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12016 + }, + { + "epoch": 0.12017, + "grad_norm": 1.4431048621531977, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 12017 + }, + { + "epoch": 0.12018, + "grad_norm": 0.9166310978141474, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 12018 + }, + { + "epoch": 0.12019, + "grad_norm": 1.2506433606927407, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12019 + }, + { + "epoch": 0.1202, + "grad_norm": 1.480480476667452, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 12020 + }, + { + "epoch": 0.12021, + "grad_norm": 1.0689179540973426, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 12021 + }, + { + "epoch": 0.12022, + "grad_norm": 1.238638964578657, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 12022 + }, + { + "epoch": 0.12023, + "grad_norm": 1.217878541780031, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 12023 + }, + { + "epoch": 0.12024, + "grad_norm": 1.0851112909646494, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12024 + }, + { + "epoch": 0.12025, + "grad_norm": 1.1604224142434416, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 12025 + }, + { + "epoch": 0.12026, + "grad_norm": 1.031844987403058, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 12026 + }, + { + "epoch": 0.12027, + "grad_norm": 1.1595745336365029, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 12027 + }, + { + "epoch": 0.12028, + "grad_norm": 1.1947817751700118, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 12028 + }, + { + "epoch": 0.12029, + "grad_norm": 1.0497319637213312, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 12029 + }, + { + "epoch": 0.1203, + "grad_norm": 1.0887730490228416, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 12030 + }, + { + "epoch": 0.12031, + "grad_norm": 1.2639844964964313, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 12031 + }, + { + "epoch": 0.12032, + "grad_norm": 1.1476592744733245, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 12032 + }, + { + "epoch": 0.12033, + "grad_norm": 1.4253662214975455, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 12033 + }, + { + "epoch": 0.12034, + "grad_norm": 1.0367483017417718, + "learning_rate": 0.003, + "loss": 4.025, + "step": 12034 + }, + { + "epoch": 0.12035, + "grad_norm": 1.3672538599328856, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 12035 + }, + { + "epoch": 0.12036, + "grad_norm": 1.088814881263226, + "learning_rate": 0.003, + "loss": 4.04, + "step": 12036 + }, + { + "epoch": 0.12037, + "grad_norm": 1.2526346282054286, + "learning_rate": 0.003, + "loss": 4.035, + "step": 12037 + }, + { + "epoch": 0.12038, + "grad_norm": 1.160570553266028, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 12038 + }, + { + "epoch": 0.12039, + "grad_norm": 1.1199414106378185, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 12039 + }, + { + "epoch": 0.1204, + "grad_norm": 1.1004831372504815, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 12040 + }, + { + "epoch": 0.12041, + "grad_norm": 1.168791590589254, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 12041 + }, + { + "epoch": 0.12042, + "grad_norm": 1.1345050629943982, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 12042 + }, + { + "epoch": 0.12043, + "grad_norm": 1.1312277297552067, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 12043 + }, + { + "epoch": 0.12044, + "grad_norm": 1.1218607973489465, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12044 + }, + { + "epoch": 0.12045, + "grad_norm": 1.3439830827965968, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 12045 + }, + { + "epoch": 0.12046, + "grad_norm": 1.194576377831826, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 12046 + }, + { + "epoch": 0.12047, + "grad_norm": 1.2242887448551507, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 12047 + }, + { + "epoch": 0.12048, + "grad_norm": 1.1869139877584274, + "learning_rate": 0.003, + "loss": 4.025, + "step": 12048 + }, + { + "epoch": 0.12049, + "grad_norm": 1.23366882798688, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 12049 + }, + { + "epoch": 0.1205, + "grad_norm": 1.1321804170340066, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 12050 + }, + { + "epoch": 0.12051, + "grad_norm": 1.2506912484400732, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 12051 + }, + { + "epoch": 0.12052, + "grad_norm": 1.171321048284884, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 12052 + }, + { + "epoch": 0.12053, + "grad_norm": 1.5483651625241348, + "learning_rate": 0.003, + "loss": 4.038, + "step": 12053 + }, + { + "epoch": 0.12054, + "grad_norm": 0.9450023337340406, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 12054 + }, + { + "epoch": 0.12055, + "grad_norm": 1.2518788502247158, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 12055 + }, + { + "epoch": 0.12056, + "grad_norm": 1.2025596603154725, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 12056 + }, + { + "epoch": 0.12057, + "grad_norm": 1.2521406673057627, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 12057 + }, + { + "epoch": 0.12058, + "grad_norm": 1.0526770017752556, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 12058 + }, + { + "epoch": 0.12059, + "grad_norm": 1.0626535398232526, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 12059 + }, + { + "epoch": 0.1206, + "grad_norm": 1.3336122110831834, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 12060 + }, + { + "epoch": 0.12061, + "grad_norm": 1.118599017531215, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 12061 + }, + { + "epoch": 0.12062, + "grad_norm": 1.363473374169237, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 12062 + }, + { + "epoch": 0.12063, + "grad_norm": 0.8786979557351099, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 12063 + }, + { + "epoch": 0.12064, + "grad_norm": 1.1947466023568365, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 12064 + }, + { + "epoch": 0.12065, + "grad_norm": 1.215993473222538, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 12065 + }, + { + "epoch": 0.12066, + "grad_norm": 1.2447752310217368, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 12066 + }, + { + "epoch": 0.12067, + "grad_norm": 1.0837006580418418, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 12067 + }, + { + "epoch": 0.12068, + "grad_norm": 1.3971605398432154, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 12068 + }, + { + "epoch": 0.12069, + "grad_norm": 1.007380264646283, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12069 + }, + { + "epoch": 0.1207, + "grad_norm": 1.3122686823444492, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 12070 + }, + { + "epoch": 0.12071, + "grad_norm": 1.0684153514164216, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 12071 + }, + { + "epoch": 0.12072, + "grad_norm": 1.2604810107038449, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 12072 + }, + { + "epoch": 0.12073, + "grad_norm": 1.155441969210191, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 12073 + }, + { + "epoch": 0.12074, + "grad_norm": 1.5303984928048056, + "learning_rate": 0.003, + "loss": 4.036, + "step": 12074 + }, + { + "epoch": 0.12075, + "grad_norm": 0.8280353697240316, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 12075 + }, + { + "epoch": 0.12076, + "grad_norm": 0.9625486618108077, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 12076 + }, + { + "epoch": 0.12077, + "grad_norm": 1.2545752671715014, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 12077 + }, + { + "epoch": 0.12078, + "grad_norm": 1.1154569818666091, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 12078 + }, + { + "epoch": 0.12079, + "grad_norm": 1.141600191269537, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 12079 + }, + { + "epoch": 0.1208, + "grad_norm": 1.238179881437241, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 12080 + }, + { + "epoch": 0.12081, + "grad_norm": 1.165698596146827, + "learning_rate": 0.003, + "loss": 4.029, + "step": 12081 + }, + { + "epoch": 0.12082, + "grad_norm": 0.8889092422891364, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 12082 + }, + { + "epoch": 0.12083, + "grad_norm": 1.2543161652007244, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 12083 + }, + { + "epoch": 0.12084, + "grad_norm": 1.5327269316533685, + "learning_rate": 0.003, + "loss": 4.038, + "step": 12084 + }, + { + "epoch": 0.12085, + "grad_norm": 1.0365382149666054, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12085 + }, + { + "epoch": 0.12086, + "grad_norm": 1.316513743782211, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 12086 + }, + { + "epoch": 0.12087, + "grad_norm": 1.1582107753316935, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 12087 + }, + { + "epoch": 0.12088, + "grad_norm": 1.2026972854083593, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12088 + }, + { + "epoch": 0.12089, + "grad_norm": 1.1865818257324852, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 12089 + }, + { + "epoch": 0.1209, + "grad_norm": 1.213904883132537, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 12090 + }, + { + "epoch": 0.12091, + "grad_norm": 1.0648805961872354, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 12091 + }, + { + "epoch": 0.12092, + "grad_norm": 1.1813755883049362, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 12092 + }, + { + "epoch": 0.12093, + "grad_norm": 1.2292752165558312, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 12093 + }, + { + "epoch": 0.12094, + "grad_norm": 1.259437705088851, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 12094 + }, + { + "epoch": 0.12095, + "grad_norm": 1.1137767158004688, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 12095 + }, + { + "epoch": 0.12096, + "grad_norm": 1.3600842565366165, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 12096 + }, + { + "epoch": 0.12097, + "grad_norm": 0.8881258371739744, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 12097 + }, + { + "epoch": 0.12098, + "grad_norm": 1.2265846606485018, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 12098 + }, + { + "epoch": 0.12099, + "grad_norm": 1.2320301694381899, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 12099 + }, + { + "epoch": 0.121, + "grad_norm": 1.1555002782743249, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 12100 + }, + { + "epoch": 0.12101, + "grad_norm": 1.26009712635675, + "learning_rate": 0.003, + "loss": 4.031, + "step": 12101 + }, + { + "epoch": 0.12102, + "grad_norm": 1.2765674036907484, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 12102 + }, + { + "epoch": 0.12103, + "grad_norm": 1.2664976256713367, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 12103 + }, + { + "epoch": 0.12104, + "grad_norm": 0.9432051107004329, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 12104 + }, + { + "epoch": 0.12105, + "grad_norm": 1.326445164229193, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 12105 + }, + { + "epoch": 0.12106, + "grad_norm": 1.0473709580825368, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 12106 + }, + { + "epoch": 0.12107, + "grad_norm": 1.5389171702810274, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 12107 + }, + { + "epoch": 0.12108, + "grad_norm": 0.961160561659805, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 12108 + }, + { + "epoch": 0.12109, + "grad_norm": 1.1860213316322281, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 12109 + }, + { + "epoch": 0.1211, + "grad_norm": 1.276050655690645, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 12110 + }, + { + "epoch": 0.12111, + "grad_norm": 1.0106495220145681, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 12111 + }, + { + "epoch": 0.12112, + "grad_norm": 1.1404629571385734, + "learning_rate": 0.003, + "loss": 4.049, + "step": 12112 + }, + { + "epoch": 0.12113, + "grad_norm": 1.3379694333364878, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 12113 + }, + { + "epoch": 0.12114, + "grad_norm": 1.1057144085632993, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 12114 + }, + { + "epoch": 0.12115, + "grad_norm": 1.2389514049197021, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 12115 + }, + { + "epoch": 0.12116, + "grad_norm": 1.081032310728611, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 12116 + }, + { + "epoch": 0.12117, + "grad_norm": 1.7808747808197203, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 12117 + }, + { + "epoch": 0.12118, + "grad_norm": 0.9804521331485236, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 12118 + }, + { + "epoch": 0.12119, + "grad_norm": 1.2622545631694635, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 12119 + }, + { + "epoch": 0.1212, + "grad_norm": 1.025608924940026, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 12120 + }, + { + "epoch": 0.12121, + "grad_norm": 1.2687986530932442, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 12121 + }, + { + "epoch": 0.12122, + "grad_norm": 1.2569166469856572, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 12122 + }, + { + "epoch": 0.12123, + "grad_norm": 1.0888418797208075, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 12123 + }, + { + "epoch": 0.12124, + "grad_norm": 1.4471321534196813, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 12124 + }, + { + "epoch": 0.12125, + "grad_norm": 1.0570726579438743, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 12125 + }, + { + "epoch": 0.12126, + "grad_norm": 1.1452310973638715, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 12126 + }, + { + "epoch": 0.12127, + "grad_norm": 1.0951003459920567, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 12127 + }, + { + "epoch": 0.12128, + "grad_norm": 1.0924282991917444, + "learning_rate": 0.003, + "loss": 4.043, + "step": 12128 + }, + { + "epoch": 0.12129, + "grad_norm": 1.077336233158983, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 12129 + }, + { + "epoch": 0.1213, + "grad_norm": 1.0698116679642007, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 12130 + }, + { + "epoch": 0.12131, + "grad_norm": 1.0569838113736956, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 12131 + }, + { + "epoch": 0.12132, + "grad_norm": 1.2236310443056324, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 12132 + }, + { + "epoch": 0.12133, + "grad_norm": 1.210938464105007, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 12133 + }, + { + "epoch": 0.12134, + "grad_norm": 1.1431794281160383, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 12134 + }, + { + "epoch": 0.12135, + "grad_norm": 1.2120388488293794, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 12135 + }, + { + "epoch": 0.12136, + "grad_norm": 1.163220939636625, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 12136 + }, + { + "epoch": 0.12137, + "grad_norm": 1.408415883667614, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 12137 + }, + { + "epoch": 0.12138, + "grad_norm": 1.019972444797985, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 12138 + }, + { + "epoch": 0.12139, + "grad_norm": 1.4794420127960417, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 12139 + }, + { + "epoch": 0.1214, + "grad_norm": 0.9942984332809602, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 12140 + }, + { + "epoch": 0.12141, + "grad_norm": 1.198313384318162, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 12141 + }, + { + "epoch": 0.12142, + "grad_norm": 1.1235175502034191, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 12142 + }, + { + "epoch": 0.12143, + "grad_norm": 1.2021305789112728, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 12143 + }, + { + "epoch": 0.12144, + "grad_norm": 1.2036254675483424, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 12144 + }, + { + "epoch": 0.12145, + "grad_norm": 1.0563897356132683, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 12145 + }, + { + "epoch": 0.12146, + "grad_norm": 1.3892405792341498, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 12146 + }, + { + "epoch": 0.12147, + "grad_norm": 1.1584776401297876, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 12147 + }, + { + "epoch": 0.12148, + "grad_norm": 1.4771991665797888, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 12148 + }, + { + "epoch": 0.12149, + "grad_norm": 0.9036675444370385, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 12149 + }, + { + "epoch": 0.1215, + "grad_norm": 1.0455355580610362, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 12150 + }, + { + "epoch": 0.12151, + "grad_norm": 1.2718005495324547, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 12151 + }, + { + "epoch": 0.12152, + "grad_norm": 1.1342945095583343, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12152 + }, + { + "epoch": 0.12153, + "grad_norm": 1.2466145803684447, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 12153 + }, + { + "epoch": 0.12154, + "grad_norm": 1.0819537046373542, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 12154 + }, + { + "epoch": 0.12155, + "grad_norm": 1.1602226084211311, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 12155 + }, + { + "epoch": 0.12156, + "grad_norm": 1.4009042185837033, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 12156 + }, + { + "epoch": 0.12157, + "grad_norm": 1.1384726322964522, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 12157 + }, + { + "epoch": 0.12158, + "grad_norm": 1.1226291807229503, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 12158 + }, + { + "epoch": 0.12159, + "grad_norm": 1.188819542203544, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 12159 + }, + { + "epoch": 0.1216, + "grad_norm": 1.186378564541427, + "learning_rate": 0.003, + "loss": 4.016, + "step": 12160 + }, + { + "epoch": 0.12161, + "grad_norm": 1.1430987105677886, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 12161 + }, + { + "epoch": 0.12162, + "grad_norm": 1.0984081471542013, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 12162 + }, + { + "epoch": 0.12163, + "grad_norm": 1.2375098267751024, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 12163 + }, + { + "epoch": 0.12164, + "grad_norm": 1.3998298901471347, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 12164 + }, + { + "epoch": 0.12165, + "grad_norm": 0.9344084284754219, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 12165 + }, + { + "epoch": 0.12166, + "grad_norm": 1.512859671545336, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 12166 + }, + { + "epoch": 0.12167, + "grad_norm": 1.1164706886354052, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 12167 + }, + { + "epoch": 0.12168, + "grad_norm": 1.0269895024920574, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 12168 + }, + { + "epoch": 0.12169, + "grad_norm": 1.2437063976530573, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 12169 + }, + { + "epoch": 0.1217, + "grad_norm": 1.1616197418244247, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 12170 + }, + { + "epoch": 0.12171, + "grad_norm": 1.126266794256949, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 12171 + }, + { + "epoch": 0.12172, + "grad_norm": 1.365627633085238, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 12172 + }, + { + "epoch": 0.12173, + "grad_norm": 1.09988351709105, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 12173 + }, + { + "epoch": 0.12174, + "grad_norm": 1.199748433735864, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 12174 + }, + { + "epoch": 0.12175, + "grad_norm": 1.088769417635092, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 12175 + }, + { + "epoch": 0.12176, + "grad_norm": 1.2661905434748413, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 12176 + }, + { + "epoch": 0.12177, + "grad_norm": 1.0836807067546872, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 12177 + }, + { + "epoch": 0.12178, + "grad_norm": 1.2934576723672395, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 12178 + }, + { + "epoch": 0.12179, + "grad_norm": 1.1831772974220784, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 12179 + }, + { + "epoch": 0.1218, + "grad_norm": 1.1747496433726747, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 12180 + }, + { + "epoch": 0.12181, + "grad_norm": 1.3179604690170197, + "learning_rate": 0.003, + "loss": 4.01, + "step": 12181 + }, + { + "epoch": 0.12182, + "grad_norm": 1.135089293334139, + "learning_rate": 0.003, + "loss": 4.031, + "step": 12182 + }, + { + "epoch": 0.12183, + "grad_norm": 1.3221372241316014, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 12183 + }, + { + "epoch": 0.12184, + "grad_norm": 1.0124457856344928, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12184 + }, + { + "epoch": 0.12185, + "grad_norm": 1.3589982758664614, + "learning_rate": 0.003, + "loss": 4.019, + "step": 12185 + }, + { + "epoch": 0.12186, + "grad_norm": 1.0233570129876237, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 12186 + }, + { + "epoch": 0.12187, + "grad_norm": 1.3079895671489778, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 12187 + }, + { + "epoch": 0.12188, + "grad_norm": 1.1489865721931825, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 12188 + }, + { + "epoch": 0.12189, + "grad_norm": 1.2412768268410446, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 12189 + }, + { + "epoch": 0.1219, + "grad_norm": 1.114513378698603, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 12190 + }, + { + "epoch": 0.12191, + "grad_norm": 1.3362592660486117, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 12191 + }, + { + "epoch": 0.12192, + "grad_norm": 1.0692297451970725, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 12192 + }, + { + "epoch": 0.12193, + "grad_norm": 1.2465778007234973, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 12193 + }, + { + "epoch": 0.12194, + "grad_norm": 1.1092003193835691, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 12194 + }, + { + "epoch": 0.12195, + "grad_norm": 1.1696708395236086, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 12195 + }, + { + "epoch": 0.12196, + "grad_norm": 1.1112321590252383, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 12196 + }, + { + "epoch": 0.12197, + "grad_norm": 1.2841076485330796, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 12197 + }, + { + "epoch": 0.12198, + "grad_norm": 1.107702913323641, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 12198 + }, + { + "epoch": 0.12199, + "grad_norm": 1.1836188082240122, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 12199 + }, + { + "epoch": 0.122, + "grad_norm": 1.0702302603416496, + "learning_rate": 0.003, + "loss": 4.025, + "step": 12200 + }, + { + "epoch": 0.12201, + "grad_norm": 1.357237574537661, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 12201 + }, + { + "epoch": 0.12202, + "grad_norm": 1.107067299040779, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 12202 + }, + { + "epoch": 0.12203, + "grad_norm": 1.2133228664945446, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 12203 + }, + { + "epoch": 0.12204, + "grad_norm": 1.122185688848671, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 12204 + }, + { + "epoch": 0.12205, + "grad_norm": 1.2378973407718126, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 12205 + }, + { + "epoch": 0.12206, + "grad_norm": 1.0146025080029806, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 12206 + }, + { + "epoch": 0.12207, + "grad_norm": 1.2371951440426439, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 12207 + }, + { + "epoch": 0.12208, + "grad_norm": 1.1941790413272386, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 12208 + }, + { + "epoch": 0.12209, + "grad_norm": 1.237443502265726, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 12209 + }, + { + "epoch": 0.1221, + "grad_norm": 1.1191757624817986, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 12210 + }, + { + "epoch": 0.12211, + "grad_norm": 1.6015066005243488, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 12211 + }, + { + "epoch": 0.12212, + "grad_norm": 1.0912122851858126, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 12212 + }, + { + "epoch": 0.12213, + "grad_norm": 1.22523715982098, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 12213 + }, + { + "epoch": 0.12214, + "grad_norm": 1.025716267326302, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 12214 + }, + { + "epoch": 0.12215, + "grad_norm": 1.1976040863355408, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 12215 + }, + { + "epoch": 0.12216, + "grad_norm": 0.965569302971212, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 12216 + }, + { + "epoch": 0.12217, + "grad_norm": 1.1391293564356433, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 12217 + }, + { + "epoch": 0.12218, + "grad_norm": 1.0031823383720437, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 12218 + }, + { + "epoch": 0.12219, + "grad_norm": 1.174272945089118, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 12219 + }, + { + "epoch": 0.1222, + "grad_norm": 1.1436677312129953, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 12220 + }, + { + "epoch": 0.12221, + "grad_norm": 1.2898099150708564, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 12221 + }, + { + "epoch": 0.12222, + "grad_norm": 1.09890346663295, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 12222 + }, + { + "epoch": 0.12223, + "grad_norm": 1.3194243732788846, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 12223 + }, + { + "epoch": 0.12224, + "grad_norm": 1.1829019105273342, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 12224 + }, + { + "epoch": 0.12225, + "grad_norm": 0.9974911798449131, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 12225 + }, + { + "epoch": 0.12226, + "grad_norm": 1.2045148338615028, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 12226 + }, + { + "epoch": 0.12227, + "grad_norm": 1.2757758918468345, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12227 + }, + { + "epoch": 0.12228, + "grad_norm": 1.3030462989555127, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 12228 + }, + { + "epoch": 0.12229, + "grad_norm": 1.1447965290278583, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12229 + }, + { + "epoch": 0.1223, + "grad_norm": 1.1830532774411873, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 12230 + }, + { + "epoch": 0.12231, + "grad_norm": 0.9927189165650173, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 12231 + }, + { + "epoch": 0.12232, + "grad_norm": 1.1343637660969617, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 12232 + }, + { + "epoch": 0.12233, + "grad_norm": 1.2339551740369135, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 12233 + }, + { + "epoch": 0.12234, + "grad_norm": 1.2153175541284331, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 12234 + }, + { + "epoch": 0.12235, + "grad_norm": 1.2112370667362282, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 12235 + }, + { + "epoch": 0.12236, + "grad_norm": 1.2680813133769673, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 12236 + }, + { + "epoch": 0.12237, + "grad_norm": 1.2440893879161208, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 12237 + }, + { + "epoch": 0.12238, + "grad_norm": 1.0938588058052665, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 12238 + }, + { + "epoch": 0.12239, + "grad_norm": 1.3079644744847825, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 12239 + }, + { + "epoch": 0.1224, + "grad_norm": 1.2032893752750062, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 12240 + }, + { + "epoch": 0.12241, + "grad_norm": 1.5263423425857818, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 12241 + }, + { + "epoch": 0.12242, + "grad_norm": 1.1123122789832365, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 12242 + }, + { + "epoch": 0.12243, + "grad_norm": 1.0556139683461747, + "learning_rate": 0.003, + "loss": 4.028, + "step": 12243 + }, + { + "epoch": 0.12244, + "grad_norm": 1.3289277594180184, + "learning_rate": 0.003, + "loss": 4.049, + "step": 12244 + }, + { + "epoch": 0.12245, + "grad_norm": 1.0073269421814348, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 12245 + }, + { + "epoch": 0.12246, + "grad_norm": 1.4046094869371852, + "learning_rate": 0.003, + "loss": 4.03, + "step": 12246 + }, + { + "epoch": 0.12247, + "grad_norm": 1.051725583525556, + "learning_rate": 0.003, + "loss": 4.014, + "step": 12247 + }, + { + "epoch": 0.12248, + "grad_norm": 1.2368654496027767, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12248 + }, + { + "epoch": 0.12249, + "grad_norm": 1.1041322243002467, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 12249 + }, + { + "epoch": 0.1225, + "grad_norm": 1.2524463553074452, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 12250 + }, + { + "epoch": 0.12251, + "grad_norm": 1.1622632606319239, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 12251 + }, + { + "epoch": 0.12252, + "grad_norm": 1.3179849114848947, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 12252 + }, + { + "epoch": 0.12253, + "grad_norm": 1.0090202086355522, + "learning_rate": 0.003, + "loss": 4.027, + "step": 12253 + }, + { + "epoch": 0.12254, + "grad_norm": 1.3230018748827679, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 12254 + }, + { + "epoch": 0.12255, + "grad_norm": 1.040049687066231, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 12255 + }, + { + "epoch": 0.12256, + "grad_norm": 1.390916117154564, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 12256 + }, + { + "epoch": 0.12257, + "grad_norm": 1.1859129806624236, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 12257 + }, + { + "epoch": 0.12258, + "grad_norm": 1.0671361782302644, + "learning_rate": 0.003, + "loss": 4.014, + "step": 12258 + }, + { + "epoch": 0.12259, + "grad_norm": 1.2161759605513203, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12259 + }, + { + "epoch": 0.1226, + "grad_norm": 1.2804176653736112, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 12260 + }, + { + "epoch": 0.12261, + "grad_norm": 0.983275306429123, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 12261 + }, + { + "epoch": 0.12262, + "grad_norm": 1.2709090680095843, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 12262 + }, + { + "epoch": 0.12263, + "grad_norm": 1.1752535790764227, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 12263 + }, + { + "epoch": 0.12264, + "grad_norm": 1.276868126228694, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12264 + }, + { + "epoch": 0.12265, + "grad_norm": 1.1411677072819246, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 12265 + }, + { + "epoch": 0.12266, + "grad_norm": 1.0902306136530755, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 12266 + }, + { + "epoch": 0.12267, + "grad_norm": 1.3205667988804515, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 12267 + }, + { + "epoch": 0.12268, + "grad_norm": 1.0465548404396692, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12268 + }, + { + "epoch": 0.12269, + "grad_norm": 1.525340885064841, + "learning_rate": 0.003, + "loss": 4.024, + "step": 12269 + }, + { + "epoch": 0.1227, + "grad_norm": 0.898532904288709, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 12270 + }, + { + "epoch": 0.12271, + "grad_norm": 1.1625001192804503, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 12271 + }, + { + "epoch": 0.12272, + "grad_norm": 1.2506428969604266, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 12272 + }, + { + "epoch": 0.12273, + "grad_norm": 1.1400665488880517, + "learning_rate": 0.003, + "loss": 4.032, + "step": 12273 + }, + { + "epoch": 0.12274, + "grad_norm": 1.4807402303846313, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 12274 + }, + { + "epoch": 0.12275, + "grad_norm": 0.9718032014863593, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 12275 + }, + { + "epoch": 0.12276, + "grad_norm": 1.1004371990631832, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 12276 + }, + { + "epoch": 0.12277, + "grad_norm": 1.2316751692589583, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 12277 + }, + { + "epoch": 0.12278, + "grad_norm": 1.2692996495372502, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 12278 + }, + { + "epoch": 0.12279, + "grad_norm": 1.1298346790675529, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 12279 + }, + { + "epoch": 0.1228, + "grad_norm": 1.0241566953582193, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 12280 + }, + { + "epoch": 0.12281, + "grad_norm": 1.2259128020142214, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 12281 + }, + { + "epoch": 0.12282, + "grad_norm": 1.1727117627822834, + "learning_rate": 0.003, + "loss": 4.056, + "step": 12282 + }, + { + "epoch": 0.12283, + "grad_norm": 1.4802551869138976, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 12283 + }, + { + "epoch": 0.12284, + "grad_norm": 0.8562657072581173, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 12284 + }, + { + "epoch": 0.12285, + "grad_norm": 1.143034764039563, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 12285 + }, + { + "epoch": 0.12286, + "grad_norm": 1.2625712453206883, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 12286 + }, + { + "epoch": 0.12287, + "grad_norm": 1.0830382002896604, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 12287 + }, + { + "epoch": 0.12288, + "grad_norm": 1.1182157999246587, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 12288 + }, + { + "epoch": 0.12289, + "grad_norm": 0.9493864192646565, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 12289 + }, + { + "epoch": 0.1229, + "grad_norm": 1.1563869494269485, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12290 + }, + { + "epoch": 0.12291, + "grad_norm": 1.2541160596246155, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 12291 + }, + { + "epoch": 0.12292, + "grad_norm": 1.2248213033814566, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 12292 + }, + { + "epoch": 0.12293, + "grad_norm": 1.2000885923117754, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 12293 + }, + { + "epoch": 0.12294, + "grad_norm": 1.0981859496339343, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 12294 + }, + { + "epoch": 0.12295, + "grad_norm": 1.1575443564037111, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 12295 + }, + { + "epoch": 0.12296, + "grad_norm": 1.301359422856533, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 12296 + }, + { + "epoch": 0.12297, + "grad_norm": 1.2514385607651537, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 12297 + }, + { + "epoch": 0.12298, + "grad_norm": 1.273736329357012, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 12298 + }, + { + "epoch": 0.12299, + "grad_norm": 1.3886346632204654, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 12299 + }, + { + "epoch": 0.123, + "grad_norm": 1.0872428215026426, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 12300 + }, + { + "epoch": 0.12301, + "grad_norm": 1.370513264865828, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 12301 + }, + { + "epoch": 0.12302, + "grad_norm": 0.8505983228963824, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 12302 + }, + { + "epoch": 0.12303, + "grad_norm": 0.9457348435365074, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12303 + }, + { + "epoch": 0.12304, + "grad_norm": 0.9951681172464344, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 12304 + }, + { + "epoch": 0.12305, + "grad_norm": 1.0681593913605725, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 12305 + }, + { + "epoch": 0.12306, + "grad_norm": 1.2916709766050543, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 12306 + }, + { + "epoch": 0.12307, + "grad_norm": 1.2292127687463348, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 12307 + }, + { + "epoch": 0.12308, + "grad_norm": 1.139218468833114, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 12308 + }, + { + "epoch": 0.12309, + "grad_norm": 1.2485231059799966, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 12309 + }, + { + "epoch": 0.1231, + "grad_norm": 1.285310561114578, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12310 + }, + { + "epoch": 0.12311, + "grad_norm": 1.2794706200625143, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 12311 + }, + { + "epoch": 0.12312, + "grad_norm": 1.0817853226112992, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 12312 + }, + { + "epoch": 0.12313, + "grad_norm": 1.4834866026044042, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 12313 + }, + { + "epoch": 0.12314, + "grad_norm": 0.9875551326088596, + "learning_rate": 0.003, + "loss": 4.016, + "step": 12314 + }, + { + "epoch": 0.12315, + "grad_norm": 1.4829417421206113, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 12315 + }, + { + "epoch": 0.12316, + "grad_norm": 1.0408210412276793, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 12316 + }, + { + "epoch": 0.12317, + "grad_norm": 1.4264169348104334, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 12317 + }, + { + "epoch": 0.12318, + "grad_norm": 1.0989074999876036, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 12318 + }, + { + "epoch": 0.12319, + "grad_norm": 1.2243058952350059, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 12319 + }, + { + "epoch": 0.1232, + "grad_norm": 1.0845777556091951, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 12320 + }, + { + "epoch": 0.12321, + "grad_norm": 1.2109554941735854, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 12321 + }, + { + "epoch": 0.12322, + "grad_norm": 1.0902149341547924, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 12322 + }, + { + "epoch": 0.12323, + "grad_norm": 1.207304856681732, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 12323 + }, + { + "epoch": 0.12324, + "grad_norm": 1.1095092917775558, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12324 + }, + { + "epoch": 0.12325, + "grad_norm": 1.2357130353809473, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 12325 + }, + { + "epoch": 0.12326, + "grad_norm": 1.0963499157957226, + "learning_rate": 0.003, + "loss": 4.019, + "step": 12326 + }, + { + "epoch": 0.12327, + "grad_norm": 1.340824324855678, + "learning_rate": 0.003, + "loss": 3.999, + "step": 12327 + }, + { + "epoch": 0.12328, + "grad_norm": 1.2449364637709306, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 12328 + }, + { + "epoch": 0.12329, + "grad_norm": 1.1876948176946969, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 12329 + }, + { + "epoch": 0.1233, + "grad_norm": 1.1120658251348128, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 12330 + }, + { + "epoch": 0.12331, + "grad_norm": 1.2766533808324558, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 12331 + }, + { + "epoch": 0.12332, + "grad_norm": 1.0419909057814578, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 12332 + }, + { + "epoch": 0.12333, + "grad_norm": 1.2900410395289634, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 12333 + }, + { + "epoch": 0.12334, + "grad_norm": 1.194625300357013, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12334 + }, + { + "epoch": 0.12335, + "grad_norm": 1.3193993205603087, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12335 + }, + { + "epoch": 0.12336, + "grad_norm": 1.341608532912959, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 12336 + }, + { + "epoch": 0.12337, + "grad_norm": 1.1486692164623977, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12337 + }, + { + "epoch": 0.12338, + "grad_norm": 1.2242326996905912, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 12338 + }, + { + "epoch": 0.12339, + "grad_norm": 1.0095451474374366, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 12339 + }, + { + "epoch": 0.1234, + "grad_norm": 1.3942157585638366, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 12340 + }, + { + "epoch": 0.12341, + "grad_norm": 0.9436108135164665, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12341 + }, + { + "epoch": 0.12342, + "grad_norm": 1.1302393729306983, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 12342 + }, + { + "epoch": 0.12343, + "grad_norm": 1.3295607670812282, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 12343 + }, + { + "epoch": 0.12344, + "grad_norm": 1.194555142662113, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 12344 + }, + { + "epoch": 0.12345, + "grad_norm": 1.3101897952776536, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 12345 + }, + { + "epoch": 0.12346, + "grad_norm": 0.9886687025758016, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 12346 + }, + { + "epoch": 0.12347, + "grad_norm": 1.048339053066772, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 12347 + }, + { + "epoch": 0.12348, + "grad_norm": 1.384508645610422, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 12348 + }, + { + "epoch": 0.12349, + "grad_norm": 1.0486243460691573, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 12349 + }, + { + "epoch": 0.1235, + "grad_norm": 1.2688673285137406, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 12350 + }, + { + "epoch": 0.12351, + "grad_norm": 0.9688609196568242, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 12351 + }, + { + "epoch": 0.12352, + "grad_norm": 1.3053088708802922, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 12352 + }, + { + "epoch": 0.12353, + "grad_norm": 1.012117558257402, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 12353 + }, + { + "epoch": 0.12354, + "grad_norm": 1.4775072585954483, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 12354 + }, + { + "epoch": 0.12355, + "grad_norm": 0.9673416007518341, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 12355 + }, + { + "epoch": 0.12356, + "grad_norm": 1.1498611705148574, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 12356 + }, + { + "epoch": 0.12357, + "grad_norm": 0.9580734048464713, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 12357 + }, + { + "epoch": 0.12358, + "grad_norm": 1.3371631264382817, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 12358 + }, + { + "epoch": 0.12359, + "grad_norm": 1.139746107919948, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 12359 + }, + { + "epoch": 0.1236, + "grad_norm": 1.2004566124253264, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 12360 + }, + { + "epoch": 0.12361, + "grad_norm": 1.1116903585599232, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 12361 + }, + { + "epoch": 0.12362, + "grad_norm": 1.0685972050756347, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 12362 + }, + { + "epoch": 0.12363, + "grad_norm": 1.0483608668790687, + "learning_rate": 0.003, + "loss": 4.015, + "step": 12363 + }, + { + "epoch": 0.12364, + "grad_norm": 1.4918723329058845, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 12364 + }, + { + "epoch": 0.12365, + "grad_norm": 1.0178238571752845, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 12365 + }, + { + "epoch": 0.12366, + "grad_norm": 1.3030151120524025, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 12366 + }, + { + "epoch": 0.12367, + "grad_norm": 1.141857602951121, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 12367 + }, + { + "epoch": 0.12368, + "grad_norm": 1.2565604375779478, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 12368 + }, + { + "epoch": 0.12369, + "grad_norm": 1.214027777384337, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 12369 + }, + { + "epoch": 0.1237, + "grad_norm": 1.0915755444135937, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 12370 + }, + { + "epoch": 0.12371, + "grad_norm": 1.1867414654709663, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 12371 + }, + { + "epoch": 0.12372, + "grad_norm": 1.0546431117703352, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 12372 + }, + { + "epoch": 0.12373, + "grad_norm": 1.4647953108218288, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 12373 + }, + { + "epoch": 0.12374, + "grad_norm": 1.04762106343006, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 12374 + }, + { + "epoch": 0.12375, + "grad_norm": 1.1963079339111795, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 12375 + }, + { + "epoch": 0.12376, + "grad_norm": 1.1499302069922561, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 12376 + }, + { + "epoch": 0.12377, + "grad_norm": 1.0968516310925092, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 12377 + }, + { + "epoch": 0.12378, + "grad_norm": 1.385819083799469, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 12378 + }, + { + "epoch": 0.12379, + "grad_norm": 1.1485024350100335, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 12379 + }, + { + "epoch": 0.1238, + "grad_norm": 1.0950675066447881, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 12380 + }, + { + "epoch": 0.12381, + "grad_norm": 1.1638382018028037, + "learning_rate": 0.003, + "loss": 4.031, + "step": 12381 + }, + { + "epoch": 0.12382, + "grad_norm": 1.129457787007263, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12382 + }, + { + "epoch": 0.12383, + "grad_norm": 1.1036814048890862, + "learning_rate": 0.003, + "loss": 4.004, + "step": 12383 + }, + { + "epoch": 0.12384, + "grad_norm": 1.2376100778668477, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 12384 + }, + { + "epoch": 0.12385, + "grad_norm": 1.1084913141813695, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 12385 + }, + { + "epoch": 0.12386, + "grad_norm": 1.178199736592679, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 12386 + }, + { + "epoch": 0.12387, + "grad_norm": 1.1779043826583042, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 12387 + }, + { + "epoch": 0.12388, + "grad_norm": 1.2916996558932137, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 12388 + }, + { + "epoch": 0.12389, + "grad_norm": 1.0404310442375322, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 12389 + }, + { + "epoch": 0.1239, + "grad_norm": 1.2897160473158022, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 12390 + }, + { + "epoch": 0.12391, + "grad_norm": 1.1756647311322306, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 12391 + }, + { + "epoch": 0.12392, + "grad_norm": 1.1316623880446204, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 12392 + }, + { + "epoch": 0.12393, + "grad_norm": 1.238766697242648, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 12393 + }, + { + "epoch": 0.12394, + "grad_norm": 1.088911073223708, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 12394 + }, + { + "epoch": 0.12395, + "grad_norm": 1.1453594979436852, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 12395 + }, + { + "epoch": 0.12396, + "grad_norm": 1.2257856525238398, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 12396 + }, + { + "epoch": 0.12397, + "grad_norm": 1.1932329492927123, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 12397 + }, + { + "epoch": 0.12398, + "grad_norm": 1.263429821271588, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 12398 + }, + { + "epoch": 0.12399, + "grad_norm": 1.1410816377835378, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 12399 + }, + { + "epoch": 0.124, + "grad_norm": 1.2684277262143193, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 12400 + }, + { + "epoch": 0.12401, + "grad_norm": 1.3163035517059634, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 12401 + }, + { + "epoch": 0.12402, + "grad_norm": 1.0710840595995579, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 12402 + }, + { + "epoch": 0.12403, + "grad_norm": 1.3799174340236744, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 12403 + }, + { + "epoch": 0.12404, + "grad_norm": 0.9632650030245177, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12404 + }, + { + "epoch": 0.12405, + "grad_norm": 1.5065254113151083, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 12405 + }, + { + "epoch": 0.12406, + "grad_norm": 1.2014138932290637, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 12406 + }, + { + "epoch": 0.12407, + "grad_norm": 1.3139679781057607, + "learning_rate": 0.003, + "loss": 4.018, + "step": 12407 + }, + { + "epoch": 0.12408, + "grad_norm": 0.9751484488901144, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 12408 + }, + { + "epoch": 0.12409, + "grad_norm": 1.3002152359013281, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 12409 + }, + { + "epoch": 0.1241, + "grad_norm": 1.0966298279249171, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 12410 + }, + { + "epoch": 0.12411, + "grad_norm": 1.2426091821974081, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 12411 + }, + { + "epoch": 0.12412, + "grad_norm": 1.176010755554183, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 12412 + }, + { + "epoch": 0.12413, + "grad_norm": 1.286184335849363, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 12413 + }, + { + "epoch": 0.12414, + "grad_norm": 1.1895946884822244, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 12414 + }, + { + "epoch": 0.12415, + "grad_norm": 1.1707059442835113, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 12415 + }, + { + "epoch": 0.12416, + "grad_norm": 0.9363912878676401, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 12416 + }, + { + "epoch": 0.12417, + "grad_norm": 1.046953318675986, + "learning_rate": 0.003, + "loss": 4.04, + "step": 12417 + }, + { + "epoch": 0.12418, + "grad_norm": 1.317720438395889, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 12418 + }, + { + "epoch": 0.12419, + "grad_norm": 1.0657548477063905, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 12419 + }, + { + "epoch": 0.1242, + "grad_norm": 1.1089272329972197, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 12420 + }, + { + "epoch": 0.12421, + "grad_norm": 1.3485884933715704, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 12421 + }, + { + "epoch": 0.12422, + "grad_norm": 1.222601076789822, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12422 + }, + { + "epoch": 0.12423, + "grad_norm": 1.3597254489602004, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 12423 + }, + { + "epoch": 0.12424, + "grad_norm": 1.1677550597946742, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 12424 + }, + { + "epoch": 0.12425, + "grad_norm": 1.1477029620568557, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 12425 + }, + { + "epoch": 0.12426, + "grad_norm": 1.2385626889982522, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 12426 + }, + { + "epoch": 0.12427, + "grad_norm": 0.9655680600257277, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 12427 + }, + { + "epoch": 0.12428, + "grad_norm": 1.4020376789614961, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 12428 + }, + { + "epoch": 0.12429, + "grad_norm": 1.1747627795440823, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 12429 + }, + { + "epoch": 0.1243, + "grad_norm": 1.3579254765229938, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 12430 + }, + { + "epoch": 0.12431, + "grad_norm": 0.9294601734714742, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 12431 + }, + { + "epoch": 0.12432, + "grad_norm": 1.2390178128721387, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 12432 + }, + { + "epoch": 0.12433, + "grad_norm": 0.9563357941981651, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 12433 + }, + { + "epoch": 0.12434, + "grad_norm": 1.3376711308452687, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 12434 + }, + { + "epoch": 0.12435, + "grad_norm": 1.0536428112043779, + "learning_rate": 0.003, + "loss": 4.024, + "step": 12435 + }, + { + "epoch": 0.12436, + "grad_norm": 1.30700540627271, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 12436 + }, + { + "epoch": 0.12437, + "grad_norm": 1.1830783159612694, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 12437 + }, + { + "epoch": 0.12438, + "grad_norm": 1.0559360395390067, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 12438 + }, + { + "epoch": 0.12439, + "grad_norm": 1.3789067519353553, + "learning_rate": 0.003, + "loss": 4.039, + "step": 12439 + }, + { + "epoch": 0.1244, + "grad_norm": 1.0241235840308207, + "learning_rate": 0.003, + "loss": 4.014, + "step": 12440 + }, + { + "epoch": 0.12441, + "grad_norm": 1.3409960385311457, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 12441 + }, + { + "epoch": 0.12442, + "grad_norm": 0.9732899619983509, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12442 + }, + { + "epoch": 0.12443, + "grad_norm": 1.2486871041590437, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 12443 + }, + { + "epoch": 0.12444, + "grad_norm": 1.2134843237041595, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12444 + }, + { + "epoch": 0.12445, + "grad_norm": 1.269281860860668, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 12445 + }, + { + "epoch": 0.12446, + "grad_norm": 1.3188318275217783, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 12446 + }, + { + "epoch": 0.12447, + "grad_norm": 1.1481372478074385, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 12447 + }, + { + "epoch": 0.12448, + "grad_norm": 1.2369333239119498, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 12448 + }, + { + "epoch": 0.12449, + "grad_norm": 1.2694235265392477, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 12449 + }, + { + "epoch": 0.1245, + "grad_norm": 1.1317858601417061, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 12450 + }, + { + "epoch": 0.12451, + "grad_norm": 1.074137700501863, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 12451 + }, + { + "epoch": 0.12452, + "grad_norm": 1.2885266855909137, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 12452 + }, + { + "epoch": 0.12453, + "grad_norm": 1.0396878994563437, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 12453 + }, + { + "epoch": 0.12454, + "grad_norm": 1.4662506221052487, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 12454 + }, + { + "epoch": 0.12455, + "grad_norm": 0.9532945933163989, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12455 + }, + { + "epoch": 0.12456, + "grad_norm": 1.2584449951050858, + "learning_rate": 0.003, + "loss": 4.003, + "step": 12456 + }, + { + "epoch": 0.12457, + "grad_norm": 1.1411104987476233, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 12457 + }, + { + "epoch": 0.12458, + "grad_norm": 1.1800010265166672, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 12458 + }, + { + "epoch": 0.12459, + "grad_norm": 1.2595512640292814, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 12459 + }, + { + "epoch": 0.1246, + "grad_norm": 1.1244644633947707, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 12460 + }, + { + "epoch": 0.12461, + "grad_norm": 1.4047730844954986, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 12461 + }, + { + "epoch": 0.12462, + "grad_norm": 1.065955962651246, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 12462 + }, + { + "epoch": 0.12463, + "grad_norm": 1.23712644506417, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 12463 + }, + { + "epoch": 0.12464, + "grad_norm": 1.048205712021745, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 12464 + }, + { + "epoch": 0.12465, + "grad_norm": 1.2119882374602933, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 12465 + }, + { + "epoch": 0.12466, + "grad_norm": 1.0136502432256684, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 12466 + }, + { + "epoch": 0.12467, + "grad_norm": 1.3500074196062515, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 12467 + }, + { + "epoch": 0.12468, + "grad_norm": 1.1123455140298126, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 12468 + }, + { + "epoch": 0.12469, + "grad_norm": 1.1752785623116864, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 12469 + }, + { + "epoch": 0.1247, + "grad_norm": 1.0239397030257342, + "learning_rate": 0.003, + "loss": 3.994, + "step": 12470 + }, + { + "epoch": 0.12471, + "grad_norm": 1.3169454051060006, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 12471 + }, + { + "epoch": 0.12472, + "grad_norm": 1.076334168779626, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 12472 + }, + { + "epoch": 0.12473, + "grad_norm": 1.1950416553824108, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 12473 + }, + { + "epoch": 0.12474, + "grad_norm": 1.3179792713272818, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 12474 + }, + { + "epoch": 0.12475, + "grad_norm": 1.245756788047074, + "learning_rate": 0.003, + "loss": 4.023, + "step": 12475 + }, + { + "epoch": 0.12476, + "grad_norm": 1.1900834831639637, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 12476 + }, + { + "epoch": 0.12477, + "grad_norm": 1.2777258402774527, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 12477 + }, + { + "epoch": 0.12478, + "grad_norm": 1.3540644784807185, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 12478 + }, + { + "epoch": 0.12479, + "grad_norm": 1.250859757659591, + "learning_rate": 0.003, + "loss": 4.034, + "step": 12479 + }, + { + "epoch": 0.1248, + "grad_norm": 1.0011661308076665, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 12480 + }, + { + "epoch": 0.12481, + "grad_norm": 1.0944248906550968, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 12481 + }, + { + "epoch": 0.12482, + "grad_norm": 1.0766623803287114, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 12482 + }, + { + "epoch": 0.12483, + "grad_norm": 1.247377497569275, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 12483 + }, + { + "epoch": 0.12484, + "grad_norm": 1.2590475389724518, + "learning_rate": 0.003, + "loss": 3.985, + "step": 12484 + }, + { + "epoch": 0.12485, + "grad_norm": 1.2563221922601575, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 12485 + }, + { + "epoch": 0.12486, + "grad_norm": 1.366666755852874, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 12486 + }, + { + "epoch": 0.12487, + "grad_norm": 1.0391130894161864, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 12487 + }, + { + "epoch": 0.12488, + "grad_norm": 1.252739650762261, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 12488 + }, + { + "epoch": 0.12489, + "grad_norm": 1.1169828804543638, + "learning_rate": 0.003, + "loss": 4.005, + "step": 12489 + }, + { + "epoch": 0.1249, + "grad_norm": 1.1550846912653996, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 12490 + }, + { + "epoch": 0.12491, + "grad_norm": 1.2273757973022745, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 12491 + }, + { + "epoch": 0.12492, + "grad_norm": 1.138029638011022, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 12492 + }, + { + "epoch": 0.12493, + "grad_norm": 1.1633599584548278, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 12493 + }, + { + "epoch": 0.12494, + "grad_norm": 1.1071951401003601, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 12494 + }, + { + "epoch": 0.12495, + "grad_norm": 1.1728962633082556, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 12495 + }, + { + "epoch": 0.12496, + "grad_norm": 1.1412993656234558, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12496 + }, + { + "epoch": 0.12497, + "grad_norm": 1.1836701462957342, + "learning_rate": 0.003, + "loss": 4.023, + "step": 12497 + }, + { + "epoch": 0.12498, + "grad_norm": 1.3444553399522738, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 12498 + }, + { + "epoch": 0.12499, + "grad_norm": 1.055299434095236, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 12499 + }, + { + "epoch": 0.125, + "grad_norm": 1.1681452892889812, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 12500 + }, + { + "epoch": 0.12501, + "grad_norm": 1.1566911568476128, + "learning_rate": 0.003, + "loss": 4.039, + "step": 12501 + }, + { + "epoch": 0.12502, + "grad_norm": 1.2648808509254572, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 12502 + }, + { + "epoch": 0.12503, + "grad_norm": 1.1849013409331248, + "learning_rate": 0.003, + "loss": 4.029, + "step": 12503 + }, + { + "epoch": 0.12504, + "grad_norm": 1.2885672439504576, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 12504 + }, + { + "epoch": 0.12505, + "grad_norm": 1.1165062827533234, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 12505 + }, + { + "epoch": 0.12506, + "grad_norm": 1.312246783818512, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 12506 + }, + { + "epoch": 0.12507, + "grad_norm": 1.3021217260244191, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 12507 + }, + { + "epoch": 0.12508, + "grad_norm": 1.080380448706609, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 12508 + }, + { + "epoch": 0.12509, + "grad_norm": 1.4439839528646126, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 12509 + }, + { + "epoch": 0.1251, + "grad_norm": 1.0678739097641072, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 12510 + }, + { + "epoch": 0.12511, + "grad_norm": 1.239905906705609, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 12511 + }, + { + "epoch": 0.12512, + "grad_norm": 1.0887393820962412, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12512 + }, + { + "epoch": 0.12513, + "grad_norm": 1.1044326141571423, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 12513 + }, + { + "epoch": 0.12514, + "grad_norm": 1.296916204033871, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12514 + }, + { + "epoch": 0.12515, + "grad_norm": 1.1111976158805856, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 12515 + }, + { + "epoch": 0.12516, + "grad_norm": 1.4345634992074552, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12516 + }, + { + "epoch": 0.12517, + "grad_norm": 1.0038106730731644, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 12517 + }, + { + "epoch": 0.12518, + "grad_norm": 1.233815776710949, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 12518 + }, + { + "epoch": 0.12519, + "grad_norm": 1.0644914590833376, + "learning_rate": 0.003, + "loss": 4.028, + "step": 12519 + }, + { + "epoch": 0.1252, + "grad_norm": 1.1510047747949075, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 12520 + }, + { + "epoch": 0.12521, + "grad_norm": 1.2458741278187768, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 12521 + }, + { + "epoch": 0.12522, + "grad_norm": 1.2658284858594022, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 12522 + }, + { + "epoch": 0.12523, + "grad_norm": 1.0684459848883463, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 12523 + }, + { + "epoch": 0.12524, + "grad_norm": 1.1841532183619132, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 12524 + }, + { + "epoch": 0.12525, + "grad_norm": 1.2159923676158797, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 12525 + }, + { + "epoch": 0.12526, + "grad_norm": 1.3025781911250225, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 12526 + }, + { + "epoch": 0.12527, + "grad_norm": 1.0656354531638161, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 12527 + }, + { + "epoch": 0.12528, + "grad_norm": 1.5808452423937671, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 12528 + }, + { + "epoch": 0.12529, + "grad_norm": 1.1645359316338186, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 12529 + }, + { + "epoch": 0.1253, + "grad_norm": 1.0732087825270942, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 12530 + }, + { + "epoch": 0.12531, + "grad_norm": 1.1743639789562694, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 12531 + }, + { + "epoch": 0.12532, + "grad_norm": 1.1258543261386147, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 12532 + }, + { + "epoch": 0.12533, + "grad_norm": 1.2074492118607754, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 12533 + }, + { + "epoch": 0.12534, + "grad_norm": 1.1335294582199398, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 12534 + }, + { + "epoch": 0.12535, + "grad_norm": 1.393436023522577, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 12535 + }, + { + "epoch": 0.12536, + "grad_norm": 0.8943166326444245, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 12536 + }, + { + "epoch": 0.12537, + "grad_norm": 1.0287110754049926, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 12537 + }, + { + "epoch": 0.12538, + "grad_norm": 1.255488963677731, + "learning_rate": 0.003, + "loss": 4.031, + "step": 12538 + }, + { + "epoch": 0.12539, + "grad_norm": 1.0084876103414888, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 12539 + }, + { + "epoch": 0.1254, + "grad_norm": 1.4860849068921025, + "learning_rate": 0.003, + "loss": 4.03, + "step": 12540 + }, + { + "epoch": 0.12541, + "grad_norm": 0.9581597732122125, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 12541 + }, + { + "epoch": 0.12542, + "grad_norm": 1.4035696364661694, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 12542 + }, + { + "epoch": 0.12543, + "grad_norm": 1.0541027313016913, + "learning_rate": 0.003, + "loss": 3.99, + "step": 12543 + }, + { + "epoch": 0.12544, + "grad_norm": 1.1821544967346649, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 12544 + }, + { + "epoch": 0.12545, + "grad_norm": 1.2469621369638744, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 12545 + }, + { + "epoch": 0.12546, + "grad_norm": 1.4233370848984432, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 12546 + }, + { + "epoch": 0.12547, + "grad_norm": 1.0233959256284908, + "learning_rate": 0.003, + "loss": 4.053, + "step": 12547 + }, + { + "epoch": 0.12548, + "grad_norm": 1.242054122735897, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 12548 + }, + { + "epoch": 0.12549, + "grad_norm": 1.1070600937614252, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 12549 + }, + { + "epoch": 0.1255, + "grad_norm": 1.1087147779990503, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 12550 + }, + { + "epoch": 0.12551, + "grad_norm": 1.1665980055570004, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 12551 + }, + { + "epoch": 0.12552, + "grad_norm": 1.1086804659085676, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 12552 + }, + { + "epoch": 0.12553, + "grad_norm": 1.3121353361094357, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 12553 + }, + { + "epoch": 0.12554, + "grad_norm": 1.0719336426573771, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 12554 + }, + { + "epoch": 0.12555, + "grad_norm": 1.1223239153580142, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 12555 + }, + { + "epoch": 0.12556, + "grad_norm": 1.3267498467952836, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 12556 + }, + { + "epoch": 0.12557, + "grad_norm": 1.3268383865328555, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 12557 + }, + { + "epoch": 0.12558, + "grad_norm": 1.4931899187677369, + "learning_rate": 0.003, + "loss": 4.037, + "step": 12558 + }, + { + "epoch": 0.12559, + "grad_norm": 1.053612760407219, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 12559 + }, + { + "epoch": 0.1256, + "grad_norm": 1.3211965283864615, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 12560 + }, + { + "epoch": 0.12561, + "grad_norm": 1.2521128264959986, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 12561 + }, + { + "epoch": 0.12562, + "grad_norm": 1.3252559218711522, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12562 + }, + { + "epoch": 0.12563, + "grad_norm": 1.1034688777336168, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 12563 + }, + { + "epoch": 0.12564, + "grad_norm": 1.245795847005279, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 12564 + }, + { + "epoch": 0.12565, + "grad_norm": 1.0799959151676444, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 12565 + }, + { + "epoch": 0.12566, + "grad_norm": 1.1399011458856696, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 12566 + }, + { + "epoch": 0.12567, + "grad_norm": 1.1637302807696905, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 12567 + }, + { + "epoch": 0.12568, + "grad_norm": 1.3881251582640086, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12568 + }, + { + "epoch": 0.12569, + "grad_norm": 0.9148944974251467, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 12569 + }, + { + "epoch": 0.1257, + "grad_norm": 1.3720674020991435, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 12570 + }, + { + "epoch": 0.12571, + "grad_norm": 1.1589754593303436, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 12571 + }, + { + "epoch": 0.12572, + "grad_norm": 1.288935836145991, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 12572 + }, + { + "epoch": 0.12573, + "grad_norm": 1.3761421534077674, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 12573 + }, + { + "epoch": 0.12574, + "grad_norm": 1.0527095557807116, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 12574 + }, + { + "epoch": 0.12575, + "grad_norm": 1.2462154382753858, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 12575 + }, + { + "epoch": 0.12576, + "grad_norm": 1.1767019385647244, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 12576 + }, + { + "epoch": 0.12577, + "grad_norm": 1.2310642908036986, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12577 + }, + { + "epoch": 0.12578, + "grad_norm": 1.0681347288973733, + "learning_rate": 0.003, + "loss": 4.018, + "step": 12578 + }, + { + "epoch": 0.12579, + "grad_norm": 1.200121848817965, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 12579 + }, + { + "epoch": 0.1258, + "grad_norm": 1.0547714047591237, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 12580 + }, + { + "epoch": 0.12581, + "grad_norm": 1.399432745523962, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 12581 + }, + { + "epoch": 0.12582, + "grad_norm": 0.9891791775910421, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 12582 + }, + { + "epoch": 0.12583, + "grad_norm": 1.2865457649563121, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 12583 + }, + { + "epoch": 0.12584, + "grad_norm": 1.0318108268908344, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 12584 + }, + { + "epoch": 0.12585, + "grad_norm": 1.2833427957706554, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 12585 + }, + { + "epoch": 0.12586, + "grad_norm": 1.093722855738362, + "learning_rate": 0.003, + "loss": 4.031, + "step": 12586 + }, + { + "epoch": 0.12587, + "grad_norm": 1.1534754913393301, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 12587 + }, + { + "epoch": 0.12588, + "grad_norm": 1.3439763061886305, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 12588 + }, + { + "epoch": 0.12589, + "grad_norm": 1.2091801981891621, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 12589 + }, + { + "epoch": 0.1259, + "grad_norm": 1.1011477287499383, + "learning_rate": 0.003, + "loss": 4.021, + "step": 12590 + }, + { + "epoch": 0.12591, + "grad_norm": 1.1312006138545256, + "learning_rate": 0.003, + "loss": 4.016, + "step": 12591 + }, + { + "epoch": 0.12592, + "grad_norm": 1.1523264528504664, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 12592 + }, + { + "epoch": 0.12593, + "grad_norm": 1.255533301220495, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 12593 + }, + { + "epoch": 0.12594, + "grad_norm": 1.2058516957260115, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 12594 + }, + { + "epoch": 0.12595, + "grad_norm": 0.9821636958326797, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 12595 + }, + { + "epoch": 0.12596, + "grad_norm": 1.2542787438402092, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 12596 + }, + { + "epoch": 0.12597, + "grad_norm": 1.2121687973888942, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12597 + }, + { + "epoch": 0.12598, + "grad_norm": 1.3235298233111423, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 12598 + }, + { + "epoch": 0.12599, + "grad_norm": 1.000802003255851, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 12599 + }, + { + "epoch": 0.126, + "grad_norm": 1.2650303129189735, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 12600 + }, + { + "epoch": 0.12601, + "grad_norm": 1.2518624621845913, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 12601 + }, + { + "epoch": 0.12602, + "grad_norm": 1.4759094497892915, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 12602 + }, + { + "epoch": 0.12603, + "grad_norm": 1.2729425663847465, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 12603 + }, + { + "epoch": 0.12604, + "grad_norm": 1.191215597044836, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 12604 + }, + { + "epoch": 0.12605, + "grad_norm": 1.2346027461305622, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12605 + }, + { + "epoch": 0.12606, + "grad_norm": 1.0890178656815974, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 12606 + }, + { + "epoch": 0.12607, + "grad_norm": 1.2208264878432329, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 12607 + }, + { + "epoch": 0.12608, + "grad_norm": 1.0725139583720165, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 12608 + }, + { + "epoch": 0.12609, + "grad_norm": 1.0438403055998695, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 12609 + }, + { + "epoch": 0.1261, + "grad_norm": 1.423012084634035, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 12610 + }, + { + "epoch": 0.12611, + "grad_norm": 0.9883128164819032, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 12611 + }, + { + "epoch": 0.12612, + "grad_norm": 1.4515971588087238, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 12612 + }, + { + "epoch": 0.12613, + "grad_norm": 0.9323036378252141, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 12613 + }, + { + "epoch": 0.12614, + "grad_norm": 1.3119744482601774, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 12614 + }, + { + "epoch": 0.12615, + "grad_norm": 1.3243048548778869, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 12615 + }, + { + "epoch": 0.12616, + "grad_norm": 1.0687568535775176, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 12616 + }, + { + "epoch": 0.12617, + "grad_norm": 1.2329522629931724, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 12617 + }, + { + "epoch": 0.12618, + "grad_norm": 1.2447679099075022, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 12618 + }, + { + "epoch": 0.12619, + "grad_norm": 1.1574700071462647, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 12619 + }, + { + "epoch": 0.1262, + "grad_norm": 1.127223199736362, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 12620 + }, + { + "epoch": 0.12621, + "grad_norm": 1.2178554897602358, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 12621 + }, + { + "epoch": 0.12622, + "grad_norm": 1.5252586533504058, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 12622 + }, + { + "epoch": 0.12623, + "grad_norm": 0.9953370520103634, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 12623 + }, + { + "epoch": 0.12624, + "grad_norm": 1.2288885934582412, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 12624 + }, + { + "epoch": 0.12625, + "grad_norm": 1.2039239879403167, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 12625 + }, + { + "epoch": 0.12626, + "grad_norm": 1.0854130128519532, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 12626 + }, + { + "epoch": 0.12627, + "grad_norm": 1.1745590891614175, + "learning_rate": 0.003, + "loss": 4.021, + "step": 12627 + }, + { + "epoch": 0.12628, + "grad_norm": 1.1566678215804154, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 12628 + }, + { + "epoch": 0.12629, + "grad_norm": 1.233599095010633, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 12629 + }, + { + "epoch": 0.1263, + "grad_norm": 1.270380200195146, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 12630 + }, + { + "epoch": 0.12631, + "grad_norm": 1.1062440334986967, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 12631 + }, + { + "epoch": 0.12632, + "grad_norm": 1.3055152614095051, + "learning_rate": 0.003, + "loss": 4.026, + "step": 12632 + }, + { + "epoch": 0.12633, + "grad_norm": 1.0713326524662674, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 12633 + }, + { + "epoch": 0.12634, + "grad_norm": 1.328625658674953, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 12634 + }, + { + "epoch": 0.12635, + "grad_norm": 0.9992586114619201, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 12635 + }, + { + "epoch": 0.12636, + "grad_norm": 1.4006756734928287, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 12636 + }, + { + "epoch": 0.12637, + "grad_norm": 1.0884890054982994, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12637 + }, + { + "epoch": 0.12638, + "grad_norm": 1.2337201777333826, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 12638 + }, + { + "epoch": 0.12639, + "grad_norm": 1.1474779657673895, + "learning_rate": 0.003, + "loss": 4.039, + "step": 12639 + }, + { + "epoch": 0.1264, + "grad_norm": 1.2550079297749897, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 12640 + }, + { + "epoch": 0.12641, + "grad_norm": 1.2570253479523554, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 12641 + }, + { + "epoch": 0.12642, + "grad_norm": 1.2545340832026206, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 12642 + }, + { + "epoch": 0.12643, + "grad_norm": 1.1479312545047873, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 12643 + }, + { + "epoch": 0.12644, + "grad_norm": 1.3820539223152442, + "learning_rate": 0.003, + "loss": 4.019, + "step": 12644 + }, + { + "epoch": 0.12645, + "grad_norm": 1.0105513446878904, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 12645 + }, + { + "epoch": 0.12646, + "grad_norm": 1.2048119372599568, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 12646 + }, + { + "epoch": 0.12647, + "grad_norm": 1.010959049838853, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 12647 + }, + { + "epoch": 0.12648, + "grad_norm": 1.2821540434209036, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 12648 + }, + { + "epoch": 0.12649, + "grad_norm": 1.0810220087640319, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 12649 + }, + { + "epoch": 0.1265, + "grad_norm": 1.204458455749565, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 12650 + }, + { + "epoch": 0.12651, + "grad_norm": 0.9792762762076124, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 12651 + }, + { + "epoch": 0.12652, + "grad_norm": 1.5682234906370067, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 12652 + }, + { + "epoch": 0.12653, + "grad_norm": 0.8164817667847991, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 12653 + }, + { + "epoch": 0.12654, + "grad_norm": 1.2008138076892603, + "learning_rate": 0.003, + "loss": 4.018, + "step": 12654 + }, + { + "epoch": 0.12655, + "grad_norm": 1.2201295724632517, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 12655 + }, + { + "epoch": 0.12656, + "grad_norm": 1.3418859859435595, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 12656 + }, + { + "epoch": 0.12657, + "grad_norm": 1.2531819255827479, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 12657 + }, + { + "epoch": 0.12658, + "grad_norm": 1.2676038809684318, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 12658 + }, + { + "epoch": 0.12659, + "grad_norm": 1.0091691228634534, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 12659 + }, + { + "epoch": 0.1266, + "grad_norm": 1.2238608274423066, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 12660 + }, + { + "epoch": 0.12661, + "grad_norm": 0.9105821714736707, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 12661 + }, + { + "epoch": 0.12662, + "grad_norm": 1.2446462516408656, + "learning_rate": 0.003, + "loss": 4.02, + "step": 12662 + }, + { + "epoch": 0.12663, + "grad_norm": 1.3444189852230655, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12663 + }, + { + "epoch": 0.12664, + "grad_norm": 1.215579372076641, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 12664 + }, + { + "epoch": 0.12665, + "grad_norm": 1.1894756167487517, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 12665 + }, + { + "epoch": 0.12666, + "grad_norm": 1.0796402434304548, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 12666 + }, + { + "epoch": 0.12667, + "grad_norm": 1.2729192473905775, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 12667 + }, + { + "epoch": 0.12668, + "grad_norm": 1.148234082366681, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 12668 + }, + { + "epoch": 0.12669, + "grad_norm": 1.1418967246731544, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 12669 + }, + { + "epoch": 0.1267, + "grad_norm": 1.2700756456120366, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 12670 + }, + { + "epoch": 0.12671, + "grad_norm": 1.0764358520395991, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 12671 + }, + { + "epoch": 0.12672, + "grad_norm": 1.409785311657584, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 12672 + }, + { + "epoch": 0.12673, + "grad_norm": 1.1401916417528815, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 12673 + }, + { + "epoch": 0.12674, + "grad_norm": 1.1669071703585918, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 12674 + }, + { + "epoch": 0.12675, + "grad_norm": 1.1980384256902983, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 12675 + }, + { + "epoch": 0.12676, + "grad_norm": 1.2130331742254143, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 12676 + }, + { + "epoch": 0.12677, + "grad_norm": 1.1165493248220288, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 12677 + }, + { + "epoch": 0.12678, + "grad_norm": 1.3433766034805201, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12678 + }, + { + "epoch": 0.12679, + "grad_norm": 1.1252445488214073, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 12679 + }, + { + "epoch": 0.1268, + "grad_norm": 1.0754874878584735, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 12680 + }, + { + "epoch": 0.12681, + "grad_norm": 1.1485296840321575, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 12681 + }, + { + "epoch": 0.12682, + "grad_norm": 1.1760583571845962, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 12682 + }, + { + "epoch": 0.12683, + "grad_norm": 1.0880121539775605, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12683 + }, + { + "epoch": 0.12684, + "grad_norm": 1.3040924045912328, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 12684 + }, + { + "epoch": 0.12685, + "grad_norm": 1.1876837315600164, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 12685 + }, + { + "epoch": 0.12686, + "grad_norm": 1.3387931565528755, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 12686 + }, + { + "epoch": 0.12687, + "grad_norm": 1.0300693277032922, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 12687 + }, + { + "epoch": 0.12688, + "grad_norm": 1.297604962549656, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 12688 + }, + { + "epoch": 0.12689, + "grad_norm": 1.100287623472564, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 12689 + }, + { + "epoch": 0.1269, + "grad_norm": 1.2595777690018868, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 12690 + }, + { + "epoch": 0.12691, + "grad_norm": 1.1988396498532445, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 12691 + }, + { + "epoch": 0.12692, + "grad_norm": 1.130646827975255, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 12692 + }, + { + "epoch": 0.12693, + "grad_norm": 1.1694215564141754, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 12693 + }, + { + "epoch": 0.12694, + "grad_norm": 1.2250507543257125, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 12694 + }, + { + "epoch": 0.12695, + "grad_norm": 1.205915660828806, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 12695 + }, + { + "epoch": 0.12696, + "grad_norm": 1.2956504501050932, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 12696 + }, + { + "epoch": 0.12697, + "grad_norm": 1.0210621754158544, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 12697 + }, + { + "epoch": 0.12698, + "grad_norm": 1.2599085606308953, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 12698 + }, + { + "epoch": 0.12699, + "grad_norm": 1.2736017966213373, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 12699 + }, + { + "epoch": 0.127, + "grad_norm": 1.069125521043092, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 12700 + }, + { + "epoch": 0.12701, + "grad_norm": 1.1512390118258113, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 12701 + }, + { + "epoch": 0.12702, + "grad_norm": 1.2300830756231789, + "learning_rate": 0.003, + "loss": 4.024, + "step": 12702 + }, + { + "epoch": 0.12703, + "grad_norm": 1.3934301726307217, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 12703 + }, + { + "epoch": 0.12704, + "grad_norm": 1.0242474018081544, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 12704 + }, + { + "epoch": 0.12705, + "grad_norm": 1.331495197064204, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12705 + }, + { + "epoch": 0.12706, + "grad_norm": 1.1882006045530609, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 12706 + }, + { + "epoch": 0.12707, + "grad_norm": 1.2945057105662945, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 12707 + }, + { + "epoch": 0.12708, + "grad_norm": 1.1382027832201964, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 12708 + }, + { + "epoch": 0.12709, + "grad_norm": 1.4619644889251888, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 12709 + }, + { + "epoch": 0.1271, + "grad_norm": 0.8937466524403883, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 12710 + }, + { + "epoch": 0.12711, + "grad_norm": 1.0864349938474558, + "learning_rate": 0.003, + "loss": 4.012, + "step": 12711 + }, + { + "epoch": 0.12712, + "grad_norm": 1.5497917839594966, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 12712 + }, + { + "epoch": 0.12713, + "grad_norm": 1.183683442872549, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 12713 + }, + { + "epoch": 0.12714, + "grad_norm": 1.3295079655733, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 12714 + }, + { + "epoch": 0.12715, + "grad_norm": 0.9693033253470243, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 12715 + }, + { + "epoch": 0.12716, + "grad_norm": 1.3432193381885342, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 12716 + }, + { + "epoch": 0.12717, + "grad_norm": 1.128924336563658, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 12717 + }, + { + "epoch": 0.12718, + "grad_norm": 1.2072489511705236, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 12718 + }, + { + "epoch": 0.12719, + "grad_norm": 1.2789377715316639, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 12719 + }, + { + "epoch": 0.1272, + "grad_norm": 1.1420326588382799, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 12720 + }, + { + "epoch": 0.12721, + "grad_norm": 1.1654815751718461, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 12721 + }, + { + "epoch": 0.12722, + "grad_norm": 1.4278123071241624, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 12722 + }, + { + "epoch": 0.12723, + "grad_norm": 1.0738923650659709, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 12723 + }, + { + "epoch": 0.12724, + "grad_norm": 1.3422297872267799, + "learning_rate": 0.003, + "loss": 4.027, + "step": 12724 + }, + { + "epoch": 0.12725, + "grad_norm": 0.9996258275269408, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 12725 + }, + { + "epoch": 0.12726, + "grad_norm": 1.310472670505521, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 12726 + }, + { + "epoch": 0.12727, + "grad_norm": 0.9766609124996919, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 12727 + }, + { + "epoch": 0.12728, + "grad_norm": 1.3007863463493645, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 12728 + }, + { + "epoch": 0.12729, + "grad_norm": 1.249819730679953, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 12729 + }, + { + "epoch": 0.1273, + "grad_norm": 1.0103597361587262, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 12730 + }, + { + "epoch": 0.12731, + "grad_norm": 1.3138143045183632, + "learning_rate": 0.003, + "loss": 4.003, + "step": 12731 + }, + { + "epoch": 0.12732, + "grad_norm": 1.1308795357386041, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 12732 + }, + { + "epoch": 0.12733, + "grad_norm": 1.008945679356882, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 12733 + }, + { + "epoch": 0.12734, + "grad_norm": 1.3302092880616652, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 12734 + }, + { + "epoch": 0.12735, + "grad_norm": 1.264541944097203, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 12735 + }, + { + "epoch": 0.12736, + "grad_norm": 1.1314818993430544, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 12736 + }, + { + "epoch": 0.12737, + "grad_norm": 1.2516386048861714, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 12737 + }, + { + "epoch": 0.12738, + "grad_norm": 1.131131341631946, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 12738 + }, + { + "epoch": 0.12739, + "grad_norm": 1.156530685800842, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 12739 + }, + { + "epoch": 0.1274, + "grad_norm": 1.2694955448613015, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 12740 + }, + { + "epoch": 0.12741, + "grad_norm": 1.1550920436159788, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 12741 + }, + { + "epoch": 0.12742, + "grad_norm": 1.3406628844910542, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 12742 + }, + { + "epoch": 0.12743, + "grad_norm": 1.0580149423492489, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 12743 + }, + { + "epoch": 0.12744, + "grad_norm": 1.5671618568755727, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 12744 + }, + { + "epoch": 0.12745, + "grad_norm": 0.9666293075162874, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 12745 + }, + { + "epoch": 0.12746, + "grad_norm": 1.2172568887775441, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 12746 + }, + { + "epoch": 0.12747, + "grad_norm": 0.9342543557121035, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 12747 + }, + { + "epoch": 0.12748, + "grad_norm": 1.3005333446109317, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 12748 + }, + { + "epoch": 0.12749, + "grad_norm": 1.2021938544731907, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 12749 + }, + { + "epoch": 0.1275, + "grad_norm": 1.2152856263128113, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 12750 + }, + { + "epoch": 0.12751, + "grad_norm": 1.0638444261787958, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 12751 + }, + { + "epoch": 0.12752, + "grad_norm": 1.3389818000295493, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 12752 + }, + { + "epoch": 0.12753, + "grad_norm": 1.270707387228433, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12753 + }, + { + "epoch": 0.12754, + "grad_norm": 1.13608709343689, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 12754 + }, + { + "epoch": 0.12755, + "grad_norm": 1.3147802356213878, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 12755 + }, + { + "epoch": 0.12756, + "grad_norm": 1.2108384992780414, + "learning_rate": 0.003, + "loss": 4.037, + "step": 12756 + }, + { + "epoch": 0.12757, + "grad_norm": 1.0753953018228892, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 12757 + }, + { + "epoch": 0.12758, + "grad_norm": 1.7678007404208398, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 12758 + }, + { + "epoch": 0.12759, + "grad_norm": 0.8688622010901259, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 12759 + }, + { + "epoch": 0.1276, + "grad_norm": 1.2116409016348908, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 12760 + }, + { + "epoch": 0.12761, + "grad_norm": 1.254787253319999, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 12761 + }, + { + "epoch": 0.12762, + "grad_norm": 1.1585454805799762, + "learning_rate": 0.003, + "loss": 3.994, + "step": 12762 + }, + { + "epoch": 0.12763, + "grad_norm": 1.513644571729462, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 12763 + }, + { + "epoch": 0.12764, + "grad_norm": 0.9027207563494971, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 12764 + }, + { + "epoch": 0.12765, + "grad_norm": 1.317176094864067, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 12765 + }, + { + "epoch": 0.12766, + "grad_norm": 1.117017075660199, + "learning_rate": 0.003, + "loss": 4.038, + "step": 12766 + }, + { + "epoch": 0.12767, + "grad_norm": 1.1414308429206412, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 12767 + }, + { + "epoch": 0.12768, + "grad_norm": 1.2669417304006838, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 12768 + }, + { + "epoch": 0.12769, + "grad_norm": 0.9801818112259263, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 12769 + }, + { + "epoch": 0.1277, + "grad_norm": 1.2263495100618897, + "learning_rate": 0.003, + "loss": 3.993, + "step": 12770 + }, + { + "epoch": 0.12771, + "grad_norm": 1.1259193359926396, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 12771 + }, + { + "epoch": 0.12772, + "grad_norm": 1.2606850167055357, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 12772 + }, + { + "epoch": 0.12773, + "grad_norm": 1.1914645637275918, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 12773 + }, + { + "epoch": 0.12774, + "grad_norm": 1.110448610605845, + "learning_rate": 0.003, + "loss": 3.978, + "step": 12774 + }, + { + "epoch": 0.12775, + "grad_norm": 1.4574492731231232, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 12775 + }, + { + "epoch": 0.12776, + "grad_norm": 1.2136472323541958, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 12776 + }, + { + "epoch": 0.12777, + "grad_norm": 1.1822253563491354, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 12777 + }, + { + "epoch": 0.12778, + "grad_norm": 1.0520996981570394, + "learning_rate": 0.003, + "loss": 4.02, + "step": 12778 + }, + { + "epoch": 0.12779, + "grad_norm": 1.2831106263336869, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 12779 + }, + { + "epoch": 0.1278, + "grad_norm": 1.0560172383499893, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12780 + }, + { + "epoch": 0.12781, + "grad_norm": 1.258889909849877, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12781 + }, + { + "epoch": 0.12782, + "grad_norm": 1.2827369883028994, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12782 + }, + { + "epoch": 0.12783, + "grad_norm": 1.2007328133659945, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 12783 + }, + { + "epoch": 0.12784, + "grad_norm": 1.2077859948776573, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 12784 + }, + { + "epoch": 0.12785, + "grad_norm": 1.1480481310658968, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 12785 + }, + { + "epoch": 0.12786, + "grad_norm": 1.2382022064067508, + "learning_rate": 0.003, + "loss": 4.026, + "step": 12786 + }, + { + "epoch": 0.12787, + "grad_norm": 0.9824667158119392, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 12787 + }, + { + "epoch": 0.12788, + "grad_norm": 1.1508760952959747, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12788 + }, + { + "epoch": 0.12789, + "grad_norm": 1.1160820774172333, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 12789 + }, + { + "epoch": 0.1279, + "grad_norm": 1.2883683988031678, + "learning_rate": 0.003, + "loss": 4.047, + "step": 12790 + }, + { + "epoch": 0.12791, + "grad_norm": 1.333631695602577, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 12791 + }, + { + "epoch": 0.12792, + "grad_norm": 1.069325969160836, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 12792 + }, + { + "epoch": 0.12793, + "grad_norm": 1.3669166797853736, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 12793 + }, + { + "epoch": 0.12794, + "grad_norm": 1.1649933110212536, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 12794 + }, + { + "epoch": 0.12795, + "grad_norm": 1.330830664464028, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 12795 + }, + { + "epoch": 0.12796, + "grad_norm": 1.1422954086481265, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 12796 + }, + { + "epoch": 0.12797, + "grad_norm": 1.5008484587331534, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 12797 + }, + { + "epoch": 0.12798, + "grad_norm": 1.0499892932988404, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 12798 + }, + { + "epoch": 0.12799, + "grad_norm": 1.1323962859092773, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 12799 + }, + { + "epoch": 0.128, + "grad_norm": 1.0851630911927252, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 12800 + }, + { + "epoch": 0.12801, + "grad_norm": 1.5752077380787768, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 12801 + }, + { + "epoch": 0.12802, + "grad_norm": 1.1017740945533272, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 12802 + }, + { + "epoch": 0.12803, + "grad_norm": 1.1667952025678203, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 12803 + }, + { + "epoch": 0.12804, + "grad_norm": 1.0195581846750794, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 12804 + }, + { + "epoch": 0.12805, + "grad_norm": 1.2081431236473223, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 12805 + }, + { + "epoch": 0.12806, + "grad_norm": 0.9719602406726462, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 12806 + }, + { + "epoch": 0.12807, + "grad_norm": 1.299116883426473, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 12807 + }, + { + "epoch": 0.12808, + "grad_norm": 0.9210572627915543, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 12808 + }, + { + "epoch": 0.12809, + "grad_norm": 1.0176288806302762, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 12809 + }, + { + "epoch": 0.1281, + "grad_norm": 1.179455454339618, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 12810 + }, + { + "epoch": 0.12811, + "grad_norm": 1.125011783055268, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 12811 + }, + { + "epoch": 0.12812, + "grad_norm": 1.2823431624618908, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 12812 + }, + { + "epoch": 0.12813, + "grad_norm": 1.0853685286518326, + "learning_rate": 0.003, + "loss": 4.036, + "step": 12813 + }, + { + "epoch": 0.12814, + "grad_norm": 1.3098030352639953, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 12814 + }, + { + "epoch": 0.12815, + "grad_norm": 1.4009011882707822, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 12815 + }, + { + "epoch": 0.12816, + "grad_norm": 1.0624113021821955, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12816 + }, + { + "epoch": 0.12817, + "grad_norm": 1.5273471712819569, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 12817 + }, + { + "epoch": 0.12818, + "grad_norm": 1.1622392449946617, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 12818 + }, + { + "epoch": 0.12819, + "grad_norm": 1.0737180829548525, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 12819 + }, + { + "epoch": 0.1282, + "grad_norm": 1.5028275338727308, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 12820 + }, + { + "epoch": 0.12821, + "grad_norm": 0.9820781601781966, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 12821 + }, + { + "epoch": 0.12822, + "grad_norm": 1.189653346367063, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 12822 + }, + { + "epoch": 0.12823, + "grad_norm": 1.305750372775048, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 12823 + }, + { + "epoch": 0.12824, + "grad_norm": 1.1085853553524894, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 12824 + }, + { + "epoch": 0.12825, + "grad_norm": 1.3521052760748702, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 12825 + }, + { + "epoch": 0.12826, + "grad_norm": 1.1339361194151034, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 12826 + }, + { + "epoch": 0.12827, + "grad_norm": 1.2919587450515737, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 12827 + }, + { + "epoch": 0.12828, + "grad_norm": 1.1162216951237776, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 12828 + }, + { + "epoch": 0.12829, + "grad_norm": 1.1035553365879642, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 12829 + }, + { + "epoch": 0.1283, + "grad_norm": 1.418334221315941, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 12830 + }, + { + "epoch": 0.12831, + "grad_norm": 0.827587958149206, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 12831 + }, + { + "epoch": 0.12832, + "grad_norm": 0.9941792427885942, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 12832 + }, + { + "epoch": 0.12833, + "grad_norm": 1.3415460248468394, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 12833 + }, + { + "epoch": 0.12834, + "grad_norm": 1.5454290749754083, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 12834 + }, + { + "epoch": 0.12835, + "grad_norm": 0.966935009632508, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12835 + }, + { + "epoch": 0.12836, + "grad_norm": 1.3383965177659312, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 12836 + }, + { + "epoch": 0.12837, + "grad_norm": 1.214439584671967, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 12837 + }, + { + "epoch": 0.12838, + "grad_norm": 1.2522112846423572, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 12838 + }, + { + "epoch": 0.12839, + "grad_norm": 1.114227336414836, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 12839 + }, + { + "epoch": 0.1284, + "grad_norm": 1.3933317053817007, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 12840 + }, + { + "epoch": 0.12841, + "grad_norm": 0.9084725468732007, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 12841 + }, + { + "epoch": 0.12842, + "grad_norm": 1.3593545948818664, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 12842 + }, + { + "epoch": 0.12843, + "grad_norm": 1.05750875955245, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 12843 + }, + { + "epoch": 0.12844, + "grad_norm": 1.2623165134390104, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 12844 + }, + { + "epoch": 0.12845, + "grad_norm": 1.215934308443997, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 12845 + }, + { + "epoch": 0.12846, + "grad_norm": 1.2534991068204633, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 12846 + }, + { + "epoch": 0.12847, + "grad_norm": 1.2460794845808603, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 12847 + }, + { + "epoch": 0.12848, + "grad_norm": 1.1933751931734735, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 12848 + }, + { + "epoch": 0.12849, + "grad_norm": 1.166245408826738, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 12849 + }, + { + "epoch": 0.1285, + "grad_norm": 1.1464404870864988, + "learning_rate": 0.003, + "loss": 4.003, + "step": 12850 + }, + { + "epoch": 0.12851, + "grad_norm": 1.043965368615515, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 12851 + }, + { + "epoch": 0.12852, + "grad_norm": 1.6275648542340913, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 12852 + }, + { + "epoch": 0.12853, + "grad_norm": 0.9973713390567429, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 12853 + }, + { + "epoch": 0.12854, + "grad_norm": 1.3145129655582115, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 12854 + }, + { + "epoch": 0.12855, + "grad_norm": 0.9255561273968774, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 12855 + }, + { + "epoch": 0.12856, + "grad_norm": 1.123206468742272, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 12856 + }, + { + "epoch": 0.12857, + "grad_norm": 1.165459338103236, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 12857 + }, + { + "epoch": 0.12858, + "grad_norm": 1.1367968392625207, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 12858 + }, + { + "epoch": 0.12859, + "grad_norm": 0.9960241599180506, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 12859 + }, + { + "epoch": 0.1286, + "grad_norm": 1.3943574392367513, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12860 + }, + { + "epoch": 0.12861, + "grad_norm": 1.0839718347871652, + "learning_rate": 0.003, + "loss": 4.029, + "step": 12861 + }, + { + "epoch": 0.12862, + "grad_norm": 1.450534960996423, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 12862 + }, + { + "epoch": 0.12863, + "grad_norm": 1.2124543848819562, + "learning_rate": 0.003, + "loss": 4.039, + "step": 12863 + }, + { + "epoch": 0.12864, + "grad_norm": 1.0730704444607297, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 12864 + }, + { + "epoch": 0.12865, + "grad_norm": 1.416501422991841, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 12865 + }, + { + "epoch": 0.12866, + "grad_norm": 1.2864051589898484, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 12866 + }, + { + "epoch": 0.12867, + "grad_norm": 1.3012900075342926, + "learning_rate": 0.003, + "loss": 4.007, + "step": 12867 + }, + { + "epoch": 0.12868, + "grad_norm": 1.2316271420711697, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 12868 + }, + { + "epoch": 0.12869, + "grad_norm": 1.0559148720019054, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 12869 + }, + { + "epoch": 0.1287, + "grad_norm": 1.2076505185344402, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 12870 + }, + { + "epoch": 0.12871, + "grad_norm": 1.0351257024278024, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 12871 + }, + { + "epoch": 0.12872, + "grad_norm": 1.4946297034159481, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 12872 + }, + { + "epoch": 0.12873, + "grad_norm": 1.0444761583654851, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 12873 + }, + { + "epoch": 0.12874, + "grad_norm": 1.4111365525307598, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 12874 + }, + { + "epoch": 0.12875, + "grad_norm": 0.92305991229768, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 12875 + }, + { + "epoch": 0.12876, + "grad_norm": 1.181940558244761, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 12876 + }, + { + "epoch": 0.12877, + "grad_norm": 1.3588258265079936, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 12877 + }, + { + "epoch": 0.12878, + "grad_norm": 1.1435193997564612, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 12878 + }, + { + "epoch": 0.12879, + "grad_norm": 1.1931087094596353, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 12879 + }, + { + "epoch": 0.1288, + "grad_norm": 1.0300788873313722, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 12880 + }, + { + "epoch": 0.12881, + "grad_norm": 1.2086016131275303, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 12881 + }, + { + "epoch": 0.12882, + "grad_norm": 1.0888737394809258, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 12882 + }, + { + "epoch": 0.12883, + "grad_norm": 1.2692961787043513, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12883 + }, + { + "epoch": 0.12884, + "grad_norm": 0.906168836593753, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 12884 + }, + { + "epoch": 0.12885, + "grad_norm": 1.0714132168907917, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 12885 + }, + { + "epoch": 0.12886, + "grad_norm": 1.384801008569906, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 12886 + }, + { + "epoch": 0.12887, + "grad_norm": 1.0782291140817764, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 12887 + }, + { + "epoch": 0.12888, + "grad_norm": 1.2316065046066518, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 12888 + }, + { + "epoch": 0.12889, + "grad_norm": 1.2766952978758708, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 12889 + }, + { + "epoch": 0.1289, + "grad_norm": 1.14895904230218, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 12890 + }, + { + "epoch": 0.12891, + "grad_norm": 1.1935543230517358, + "learning_rate": 0.003, + "loss": 4.008, + "step": 12891 + }, + { + "epoch": 0.12892, + "grad_norm": 1.0722825731155168, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 12892 + }, + { + "epoch": 0.12893, + "grad_norm": 1.1006109032398521, + "learning_rate": 0.003, + "loss": 4.066, + "step": 12893 + }, + { + "epoch": 0.12894, + "grad_norm": 1.1575330038873581, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 12894 + }, + { + "epoch": 0.12895, + "grad_norm": 1.4045037649238294, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 12895 + }, + { + "epoch": 0.12896, + "grad_norm": 1.2226745359152975, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 12896 + }, + { + "epoch": 0.12897, + "grad_norm": 0.9810855290951486, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 12897 + }, + { + "epoch": 0.12898, + "grad_norm": 1.4321113882827925, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 12898 + }, + { + "epoch": 0.12899, + "grad_norm": 0.9279563463843422, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 12899 + }, + { + "epoch": 0.129, + "grad_norm": 1.0112843463751415, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 12900 + }, + { + "epoch": 0.12901, + "grad_norm": 1.258312102527465, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 12901 + }, + { + "epoch": 0.12902, + "grad_norm": 1.2599438181290643, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 12902 + }, + { + "epoch": 0.12903, + "grad_norm": 1.3168883187604847, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 12903 + }, + { + "epoch": 0.12904, + "grad_norm": 1.158850205061437, + "learning_rate": 0.003, + "loss": 3.99, + "step": 12904 + }, + { + "epoch": 0.12905, + "grad_norm": 1.2534304093157755, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 12905 + }, + { + "epoch": 0.12906, + "grad_norm": 1.2279071281854934, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 12906 + }, + { + "epoch": 0.12907, + "grad_norm": 1.3135077430669198, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 12907 + }, + { + "epoch": 0.12908, + "grad_norm": 1.1340062936234097, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 12908 + }, + { + "epoch": 0.12909, + "grad_norm": 1.2105010880042733, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 12909 + }, + { + "epoch": 0.1291, + "grad_norm": 1.035585642600207, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 12910 + }, + { + "epoch": 0.12911, + "grad_norm": 1.2677876907556864, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 12911 + }, + { + "epoch": 0.12912, + "grad_norm": 1.1098026814924011, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 12912 + }, + { + "epoch": 0.12913, + "grad_norm": 1.3247567294192373, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12913 + }, + { + "epoch": 0.12914, + "grad_norm": 1.087353870991555, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 12914 + }, + { + "epoch": 0.12915, + "grad_norm": 1.3235097791685535, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 12915 + }, + { + "epoch": 0.12916, + "grad_norm": 1.0418896283143477, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 12916 + }, + { + "epoch": 0.12917, + "grad_norm": 1.430243661406629, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 12917 + }, + { + "epoch": 0.12918, + "grad_norm": 1.244820526694616, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 12918 + }, + { + "epoch": 0.12919, + "grad_norm": 1.0491507724949716, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 12919 + }, + { + "epoch": 0.1292, + "grad_norm": 1.7564519294617458, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 12920 + }, + { + "epoch": 0.12921, + "grad_norm": 0.9365160750153148, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 12921 + }, + { + "epoch": 0.12922, + "grad_norm": 1.2570769167876839, + "learning_rate": 0.003, + "loss": 4.035, + "step": 12922 + }, + { + "epoch": 0.12923, + "grad_norm": 1.1340761136316975, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 12923 + }, + { + "epoch": 0.12924, + "grad_norm": 1.1839216294414823, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 12924 + }, + { + "epoch": 0.12925, + "grad_norm": 1.2177068177432462, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 12925 + }, + { + "epoch": 0.12926, + "grad_norm": 1.3158018270836227, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 12926 + }, + { + "epoch": 0.12927, + "grad_norm": 1.0917985819220148, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 12927 + }, + { + "epoch": 0.12928, + "grad_norm": 1.2406607657891981, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 12928 + }, + { + "epoch": 0.12929, + "grad_norm": 1.2452029167820236, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 12929 + }, + { + "epoch": 0.1293, + "grad_norm": 1.0704826750375576, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 12930 + }, + { + "epoch": 0.12931, + "grad_norm": 1.1059105445383641, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12931 + }, + { + "epoch": 0.12932, + "grad_norm": 1.2032685945299755, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 12932 + }, + { + "epoch": 0.12933, + "grad_norm": 1.139197150088746, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 12933 + }, + { + "epoch": 0.12934, + "grad_norm": 1.1274527096083895, + "learning_rate": 0.003, + "loss": 4.03, + "step": 12934 + }, + { + "epoch": 0.12935, + "grad_norm": 1.507321804546948, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 12935 + }, + { + "epoch": 0.12936, + "grad_norm": 0.975323122122746, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 12936 + }, + { + "epoch": 0.12937, + "grad_norm": 1.3623939109449617, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 12937 + }, + { + "epoch": 0.12938, + "grad_norm": 1.057288137364664, + "learning_rate": 0.003, + "loss": 4.04, + "step": 12938 + }, + { + "epoch": 0.12939, + "grad_norm": 1.3259203102570047, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 12939 + }, + { + "epoch": 0.1294, + "grad_norm": 1.110701965074797, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 12940 + }, + { + "epoch": 0.12941, + "grad_norm": 1.287210392735098, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 12941 + }, + { + "epoch": 0.12942, + "grad_norm": 1.1518747016361184, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 12942 + }, + { + "epoch": 0.12943, + "grad_norm": 1.1900719699662106, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 12943 + }, + { + "epoch": 0.12944, + "grad_norm": 1.154771928089284, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 12944 + }, + { + "epoch": 0.12945, + "grad_norm": 1.1482591004768885, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 12945 + }, + { + "epoch": 0.12946, + "grad_norm": 1.0932439918983072, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 12946 + }, + { + "epoch": 0.12947, + "grad_norm": 1.2097249296318278, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 12947 + }, + { + "epoch": 0.12948, + "grad_norm": 1.2986622397477905, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 12948 + }, + { + "epoch": 0.12949, + "grad_norm": 1.3825525145439725, + "learning_rate": 0.003, + "loss": 4.022, + "step": 12949 + }, + { + "epoch": 0.1295, + "grad_norm": 1.1758601196064937, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 12950 + }, + { + "epoch": 0.12951, + "grad_norm": 1.3591353151221124, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 12951 + }, + { + "epoch": 0.12952, + "grad_norm": 0.9131650280317642, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 12952 + }, + { + "epoch": 0.12953, + "grad_norm": 1.1014974217693525, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 12953 + }, + { + "epoch": 0.12954, + "grad_norm": 1.3026270117165117, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 12954 + }, + { + "epoch": 0.12955, + "grad_norm": 1.2136899689336957, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 12955 + }, + { + "epoch": 0.12956, + "grad_norm": 1.1081746289222236, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 12956 + }, + { + "epoch": 0.12957, + "grad_norm": 1.3558413742243949, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 12957 + }, + { + "epoch": 0.12958, + "grad_norm": 1.2985873728581943, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 12958 + }, + { + "epoch": 0.12959, + "grad_norm": 1.0915415838285472, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 12959 + }, + { + "epoch": 0.1296, + "grad_norm": 1.2271488870813643, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12960 + }, + { + "epoch": 0.12961, + "grad_norm": 1.0660449140643198, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 12961 + }, + { + "epoch": 0.12962, + "grad_norm": 1.2713070418720303, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 12962 + }, + { + "epoch": 0.12963, + "grad_norm": 1.1140722977923936, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12963 + }, + { + "epoch": 0.12964, + "grad_norm": 1.4334928816984343, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12964 + }, + { + "epoch": 0.12965, + "grad_norm": 1.0798644543572762, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 12965 + }, + { + "epoch": 0.12966, + "grad_norm": 1.3576702711558502, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 12966 + }, + { + "epoch": 0.12967, + "grad_norm": 0.9565988586229004, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 12967 + }, + { + "epoch": 0.12968, + "grad_norm": 1.356676054202947, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 12968 + }, + { + "epoch": 0.12969, + "grad_norm": 1.1836635707869203, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 12969 + }, + { + "epoch": 0.1297, + "grad_norm": 1.1187549800265795, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 12970 + }, + { + "epoch": 0.12971, + "grad_norm": 1.1920675764967605, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 12971 + }, + { + "epoch": 0.12972, + "grad_norm": 1.1969678957485745, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 12972 + }, + { + "epoch": 0.12973, + "grad_norm": 1.1830367876527692, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 12973 + }, + { + "epoch": 0.12974, + "grad_norm": 1.094989770680786, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 12974 + }, + { + "epoch": 0.12975, + "grad_norm": 1.2852262825752716, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 12975 + }, + { + "epoch": 0.12976, + "grad_norm": 1.2513488888113358, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 12976 + }, + { + "epoch": 0.12977, + "grad_norm": 1.160455662683476, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 12977 + }, + { + "epoch": 0.12978, + "grad_norm": 1.1863743123144554, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 12978 + }, + { + "epoch": 0.12979, + "grad_norm": 0.9523514728095331, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 12979 + }, + { + "epoch": 0.1298, + "grad_norm": 1.2746725183316685, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 12980 + }, + { + "epoch": 0.12981, + "grad_norm": 1.2113630031469498, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 12981 + }, + { + "epoch": 0.12982, + "grad_norm": 1.4401568049490077, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 12982 + }, + { + "epoch": 0.12983, + "grad_norm": 1.0589139916757784, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 12983 + }, + { + "epoch": 0.12984, + "grad_norm": 1.3111421740836104, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 12984 + }, + { + "epoch": 0.12985, + "grad_norm": 1.0916263279984209, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 12985 + }, + { + "epoch": 0.12986, + "grad_norm": 1.3150320490843068, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 12986 + }, + { + "epoch": 0.12987, + "grad_norm": 1.1622979768305781, + "learning_rate": 0.003, + "loss": 4.029, + "step": 12987 + }, + { + "epoch": 0.12988, + "grad_norm": 1.1795490087421945, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 12988 + }, + { + "epoch": 0.12989, + "grad_norm": 1.2023289153396604, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 12989 + }, + { + "epoch": 0.1299, + "grad_norm": 1.2904015849325297, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 12990 + }, + { + "epoch": 0.12991, + "grad_norm": 1.1788631395030478, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 12991 + }, + { + "epoch": 0.12992, + "grad_norm": 1.1325099422571487, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 12992 + }, + { + "epoch": 0.12993, + "grad_norm": 1.2285007962804115, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 12993 + }, + { + "epoch": 0.12994, + "grad_norm": 1.1110998831328704, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 12994 + }, + { + "epoch": 0.12995, + "grad_norm": 1.345079656718258, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 12995 + }, + { + "epoch": 0.12996, + "grad_norm": 1.3307329461460866, + "learning_rate": 0.003, + "loss": 3.995, + "step": 12996 + }, + { + "epoch": 0.12997, + "grad_norm": 1.1226943235816225, + "learning_rate": 0.003, + "loss": 4.027, + "step": 12997 + }, + { + "epoch": 0.12998, + "grad_norm": 1.3592512691421177, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 12998 + }, + { + "epoch": 0.12999, + "grad_norm": 1.0450028186646942, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 12999 + }, + { + "epoch": 0.13, + "grad_norm": 1.1867332569892648, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 13000 + }, + { + "epoch": 0.13001, + "grad_norm": 1.23639273204252, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 13001 + }, + { + "epoch": 0.13002, + "grad_norm": 1.1831666531849891, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 13002 + }, + { + "epoch": 0.13003, + "grad_norm": 1.2823332562222987, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 13003 + }, + { + "epoch": 0.13004, + "grad_norm": 1.191483777431203, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 13004 + }, + { + "epoch": 0.13005, + "grad_norm": 1.1804479129189784, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 13005 + }, + { + "epoch": 0.13006, + "grad_norm": 1.0660348758813278, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 13006 + }, + { + "epoch": 0.13007, + "grad_norm": 1.1998590835283771, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 13007 + }, + { + "epoch": 0.13008, + "grad_norm": 1.211992067571615, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 13008 + }, + { + "epoch": 0.13009, + "grad_norm": 1.0715092102241917, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 13009 + }, + { + "epoch": 0.1301, + "grad_norm": 1.2311289310569815, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 13010 + }, + { + "epoch": 0.13011, + "grad_norm": 1.1482130599826008, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 13011 + }, + { + "epoch": 0.13012, + "grad_norm": 1.419970438601161, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 13012 + }, + { + "epoch": 0.13013, + "grad_norm": 1.289807647099006, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 13013 + }, + { + "epoch": 0.13014, + "grad_norm": 1.3265352693337469, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 13014 + }, + { + "epoch": 0.13015, + "grad_norm": 1.3345475639012292, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 13015 + }, + { + "epoch": 0.13016, + "grad_norm": 1.0368363723912866, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 13016 + }, + { + "epoch": 0.13017, + "grad_norm": 1.2877451811752048, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 13017 + }, + { + "epoch": 0.13018, + "grad_norm": 1.012908939390002, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 13018 + }, + { + "epoch": 0.13019, + "grad_norm": 1.3703226825238353, + "learning_rate": 0.003, + "loss": 4.043, + "step": 13019 + }, + { + "epoch": 0.1302, + "grad_norm": 1.176604438017887, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 13020 + }, + { + "epoch": 0.13021, + "grad_norm": 1.1122871690777643, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 13021 + }, + { + "epoch": 0.13022, + "grad_norm": 1.2974145454477173, + "learning_rate": 0.003, + "loss": 4.041, + "step": 13022 + }, + { + "epoch": 0.13023, + "grad_norm": 1.2352521460749524, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 13023 + }, + { + "epoch": 0.13024, + "grad_norm": 1.168606954054629, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 13024 + }, + { + "epoch": 0.13025, + "grad_norm": 1.2499960934819396, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 13025 + }, + { + "epoch": 0.13026, + "grad_norm": 1.3152707080427926, + "learning_rate": 0.003, + "loss": 4.039, + "step": 13026 + }, + { + "epoch": 0.13027, + "grad_norm": 1.1778120535765901, + "learning_rate": 0.003, + "loss": 4.002, + "step": 13027 + }, + { + "epoch": 0.13028, + "grad_norm": 1.2079691050764194, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 13028 + }, + { + "epoch": 0.13029, + "grad_norm": 1.1071644576188255, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 13029 + }, + { + "epoch": 0.1303, + "grad_norm": 1.3864312407201933, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 13030 + }, + { + "epoch": 0.13031, + "grad_norm": 1.1125770889548465, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 13031 + }, + { + "epoch": 0.13032, + "grad_norm": 1.3413211527215434, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 13032 + }, + { + "epoch": 0.13033, + "grad_norm": 1.06880186736233, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 13033 + }, + { + "epoch": 0.13034, + "grad_norm": 1.2728892725094454, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 13034 + }, + { + "epoch": 0.13035, + "grad_norm": 1.0651150503240214, + "learning_rate": 0.003, + "loss": 4.038, + "step": 13035 + }, + { + "epoch": 0.13036, + "grad_norm": 1.1830519127271366, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13036 + }, + { + "epoch": 0.13037, + "grad_norm": 1.1141922529152608, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13037 + }, + { + "epoch": 0.13038, + "grad_norm": 1.3343469496900424, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 13038 + }, + { + "epoch": 0.13039, + "grad_norm": 1.134356082858109, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 13039 + }, + { + "epoch": 0.1304, + "grad_norm": 1.0985262524667785, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 13040 + }, + { + "epoch": 0.13041, + "grad_norm": 1.4877606344783099, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 13041 + }, + { + "epoch": 0.13042, + "grad_norm": 1.1653508801524972, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 13042 + }, + { + "epoch": 0.13043, + "grad_norm": 1.1754164423686313, + "learning_rate": 0.003, + "loss": 4.029, + "step": 13043 + }, + { + "epoch": 0.13044, + "grad_norm": 1.3302602473205494, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 13044 + }, + { + "epoch": 0.13045, + "grad_norm": 1.0904545751149235, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 13045 + }, + { + "epoch": 0.13046, + "grad_norm": 1.3172464854441723, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 13046 + }, + { + "epoch": 0.13047, + "grad_norm": 1.00886913655674, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 13047 + }, + { + "epoch": 0.13048, + "grad_norm": 1.244453787403923, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 13048 + }, + { + "epoch": 0.13049, + "grad_norm": 0.9503210626593559, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 13049 + }, + { + "epoch": 0.1305, + "grad_norm": 1.1082209982210707, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 13050 + }, + { + "epoch": 0.13051, + "grad_norm": 1.3825316783754058, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 13051 + }, + { + "epoch": 0.13052, + "grad_norm": 1.1065415073757985, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 13052 + }, + { + "epoch": 0.13053, + "grad_norm": 1.2282454715331608, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 13053 + }, + { + "epoch": 0.13054, + "grad_norm": 1.2066879253376723, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 13054 + }, + { + "epoch": 0.13055, + "grad_norm": 0.9873407819705948, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 13055 + }, + { + "epoch": 0.13056, + "grad_norm": 1.3135267445481453, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 13056 + }, + { + "epoch": 0.13057, + "grad_norm": 1.0202575371173688, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 13057 + }, + { + "epoch": 0.13058, + "grad_norm": 1.331106710918081, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 13058 + }, + { + "epoch": 0.13059, + "grad_norm": 1.1490975344521086, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 13059 + }, + { + "epoch": 0.1306, + "grad_norm": 1.3554628358227565, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 13060 + }, + { + "epoch": 0.13061, + "grad_norm": 1.0613734741931096, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 13061 + }, + { + "epoch": 0.13062, + "grad_norm": 1.1786998069131467, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 13062 + }, + { + "epoch": 0.13063, + "grad_norm": 1.122364934303279, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 13063 + }, + { + "epoch": 0.13064, + "grad_norm": 1.245717571216681, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 13064 + }, + { + "epoch": 0.13065, + "grad_norm": 1.2300096300862102, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13065 + }, + { + "epoch": 0.13066, + "grad_norm": 1.177243977511725, + "learning_rate": 0.003, + "loss": 4.021, + "step": 13066 + }, + { + "epoch": 0.13067, + "grad_norm": 1.121267875399439, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 13067 + }, + { + "epoch": 0.13068, + "grad_norm": 1.4458148439390073, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 13068 + }, + { + "epoch": 0.13069, + "grad_norm": 1.0655288352193661, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 13069 + }, + { + "epoch": 0.1307, + "grad_norm": 1.4673778899232879, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 13070 + }, + { + "epoch": 0.13071, + "grad_norm": 1.0077870922762957, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13071 + }, + { + "epoch": 0.13072, + "grad_norm": 1.2477720049192509, + "learning_rate": 0.003, + "loss": 4.016, + "step": 13072 + }, + { + "epoch": 0.13073, + "grad_norm": 1.2295747164214397, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 13073 + }, + { + "epoch": 0.13074, + "grad_norm": 1.2677829946765515, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 13074 + }, + { + "epoch": 0.13075, + "grad_norm": 1.3998043440354033, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 13075 + }, + { + "epoch": 0.13076, + "grad_norm": 1.0540236803053258, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 13076 + }, + { + "epoch": 0.13077, + "grad_norm": 1.2607216048499503, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 13077 + }, + { + "epoch": 0.13078, + "grad_norm": 1.0932345623866864, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 13078 + }, + { + "epoch": 0.13079, + "grad_norm": 1.2051592865391518, + "learning_rate": 0.003, + "loss": 3.991, + "step": 13079 + }, + { + "epoch": 0.1308, + "grad_norm": 1.0491942575755948, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 13080 + }, + { + "epoch": 0.13081, + "grad_norm": 1.3062916574774783, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 13081 + }, + { + "epoch": 0.13082, + "grad_norm": 1.1321098357352237, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 13082 + }, + { + "epoch": 0.13083, + "grad_norm": 1.3665990418676934, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13083 + }, + { + "epoch": 0.13084, + "grad_norm": 0.9978809201908994, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 13084 + }, + { + "epoch": 0.13085, + "grad_norm": 1.373453319634236, + "learning_rate": 0.003, + "loss": 4.037, + "step": 13085 + }, + { + "epoch": 0.13086, + "grad_norm": 0.8948839986805347, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 13086 + }, + { + "epoch": 0.13087, + "grad_norm": 1.1553678542467876, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 13087 + }, + { + "epoch": 0.13088, + "grad_norm": 1.228440641943863, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 13088 + }, + { + "epoch": 0.13089, + "grad_norm": 1.3675739443552888, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 13089 + }, + { + "epoch": 0.1309, + "grad_norm": 1.0849909250003047, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 13090 + }, + { + "epoch": 0.13091, + "grad_norm": 1.2724323504464439, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 13091 + }, + { + "epoch": 0.13092, + "grad_norm": 1.1154728415628237, + "learning_rate": 0.003, + "loss": 4.026, + "step": 13092 + }, + { + "epoch": 0.13093, + "grad_norm": 1.3266592579188943, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 13093 + }, + { + "epoch": 0.13094, + "grad_norm": 1.0431979252943413, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 13094 + }, + { + "epoch": 0.13095, + "grad_norm": 1.2587443347677894, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13095 + }, + { + "epoch": 0.13096, + "grad_norm": 1.0518579515346653, + "learning_rate": 0.003, + "loss": 4.021, + "step": 13096 + }, + { + "epoch": 0.13097, + "grad_norm": 1.182379232046843, + "learning_rate": 0.003, + "loss": 4.046, + "step": 13097 + }, + { + "epoch": 0.13098, + "grad_norm": 1.2441723313218485, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 13098 + }, + { + "epoch": 0.13099, + "grad_norm": 1.1745868481759494, + "learning_rate": 0.003, + "loss": 4.031, + "step": 13099 + }, + { + "epoch": 0.131, + "grad_norm": 1.2466331476420034, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 13100 + }, + { + "epoch": 0.13101, + "grad_norm": 1.111399149137611, + "learning_rate": 0.003, + "loss": 4.023, + "step": 13101 + }, + { + "epoch": 0.13102, + "grad_norm": 1.3801332479480273, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 13102 + }, + { + "epoch": 0.13103, + "grad_norm": 1.0319688794090096, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 13103 + }, + { + "epoch": 0.13104, + "grad_norm": 1.352104538176443, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 13104 + }, + { + "epoch": 0.13105, + "grad_norm": 0.9840982994262835, + "learning_rate": 0.003, + "loss": 4.012, + "step": 13105 + }, + { + "epoch": 0.13106, + "grad_norm": 1.2720508957564811, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 13106 + }, + { + "epoch": 0.13107, + "grad_norm": 1.3582344819108931, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 13107 + }, + { + "epoch": 0.13108, + "grad_norm": 1.263872749869849, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 13108 + }, + { + "epoch": 0.13109, + "grad_norm": 1.3146328515584427, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 13109 + }, + { + "epoch": 0.1311, + "grad_norm": 1.1142588293129503, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 13110 + }, + { + "epoch": 0.13111, + "grad_norm": 1.2948103531876165, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 13111 + }, + { + "epoch": 0.13112, + "grad_norm": 1.2131004643968781, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 13112 + }, + { + "epoch": 0.13113, + "grad_norm": 1.10642883210753, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 13113 + }, + { + "epoch": 0.13114, + "grad_norm": 1.2028397006048634, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 13114 + }, + { + "epoch": 0.13115, + "grad_norm": 0.9699969253696146, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 13115 + }, + { + "epoch": 0.13116, + "grad_norm": 1.244963431503124, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 13116 + }, + { + "epoch": 0.13117, + "grad_norm": 1.0532682402073905, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 13117 + }, + { + "epoch": 0.13118, + "grad_norm": 1.1743842407713163, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 13118 + }, + { + "epoch": 0.13119, + "grad_norm": 1.1241278412152182, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 13119 + }, + { + "epoch": 0.1312, + "grad_norm": 1.2176277731495335, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 13120 + }, + { + "epoch": 0.13121, + "grad_norm": 1.2384334922856706, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13121 + }, + { + "epoch": 0.13122, + "grad_norm": 1.3559900193532226, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 13122 + }, + { + "epoch": 0.13123, + "grad_norm": 1.0763654633297999, + "learning_rate": 0.003, + "loss": 4.013, + "step": 13123 + }, + { + "epoch": 0.13124, + "grad_norm": 1.1916683197228783, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 13124 + }, + { + "epoch": 0.13125, + "grad_norm": 1.227531132739174, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 13125 + }, + { + "epoch": 0.13126, + "grad_norm": 1.0960208761266186, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 13126 + }, + { + "epoch": 0.13127, + "grad_norm": 1.2337908088108465, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 13127 + }, + { + "epoch": 0.13128, + "grad_norm": 1.1628289901082414, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 13128 + }, + { + "epoch": 0.13129, + "grad_norm": 1.6903240617545117, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 13129 + }, + { + "epoch": 0.1313, + "grad_norm": 1.0718491893109336, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 13130 + }, + { + "epoch": 0.13131, + "grad_norm": 1.389295634459969, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 13131 + }, + { + "epoch": 0.13132, + "grad_norm": 0.9675178513023274, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 13132 + }, + { + "epoch": 0.13133, + "grad_norm": 1.2463845794263217, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 13133 + }, + { + "epoch": 0.13134, + "grad_norm": 1.303025274762317, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 13134 + }, + { + "epoch": 0.13135, + "grad_norm": 1.1598289958201813, + "learning_rate": 0.003, + "loss": 4.02, + "step": 13135 + }, + { + "epoch": 0.13136, + "grad_norm": 1.3105925553909337, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 13136 + }, + { + "epoch": 0.13137, + "grad_norm": 1.1781013394534319, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 13137 + }, + { + "epoch": 0.13138, + "grad_norm": 1.1794371671889319, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 13138 + }, + { + "epoch": 0.13139, + "grad_norm": 1.2986139091265898, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 13139 + }, + { + "epoch": 0.1314, + "grad_norm": 0.989722674097084, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 13140 + }, + { + "epoch": 0.13141, + "grad_norm": 1.3343673770114424, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 13141 + }, + { + "epoch": 0.13142, + "grad_norm": 1.1470785700967723, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 13142 + }, + { + "epoch": 0.13143, + "grad_norm": 1.1589060396306303, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 13143 + }, + { + "epoch": 0.13144, + "grad_norm": 1.0582465386323356, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 13144 + }, + { + "epoch": 0.13145, + "grad_norm": 1.3672232216600506, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 13145 + }, + { + "epoch": 0.13146, + "grad_norm": 1.1879858947806368, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 13146 + }, + { + "epoch": 0.13147, + "grad_norm": 1.20843180027465, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 13147 + }, + { + "epoch": 0.13148, + "grad_norm": 1.3432584402500107, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 13148 + }, + { + "epoch": 0.13149, + "grad_norm": 0.9095709723571204, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13149 + }, + { + "epoch": 0.1315, + "grad_norm": 1.057646553778702, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 13150 + }, + { + "epoch": 0.13151, + "grad_norm": 1.3610467106606845, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 13151 + }, + { + "epoch": 0.13152, + "grad_norm": 1.0777210898069285, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 13152 + }, + { + "epoch": 0.13153, + "grad_norm": 1.3859619564262946, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 13153 + }, + { + "epoch": 0.13154, + "grad_norm": 1.0844304505512883, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 13154 + }, + { + "epoch": 0.13155, + "grad_norm": 1.2359660173968072, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 13155 + }, + { + "epoch": 0.13156, + "grad_norm": 1.3677030719934207, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 13156 + }, + { + "epoch": 0.13157, + "grad_norm": 0.9833401342037275, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 13157 + }, + { + "epoch": 0.13158, + "grad_norm": 1.1062735484981285, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 13158 + }, + { + "epoch": 0.13159, + "grad_norm": 1.1254248542885985, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 13159 + }, + { + "epoch": 0.1316, + "grad_norm": 1.1120092920802103, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 13160 + }, + { + "epoch": 0.13161, + "grad_norm": 1.1989205083321814, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 13161 + }, + { + "epoch": 0.13162, + "grad_norm": 1.1928171487306694, + "learning_rate": 0.003, + "loss": 4.029, + "step": 13162 + }, + { + "epoch": 0.13163, + "grad_norm": 1.4081902393551464, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 13163 + }, + { + "epoch": 0.13164, + "grad_norm": 1.1583999173018455, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 13164 + }, + { + "epoch": 0.13165, + "grad_norm": 1.3293599987195368, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 13165 + }, + { + "epoch": 0.13166, + "grad_norm": 1.042849775568771, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 13166 + }, + { + "epoch": 0.13167, + "grad_norm": 1.179654456448683, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 13167 + }, + { + "epoch": 0.13168, + "grad_norm": 1.0543994115678994, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 13168 + }, + { + "epoch": 0.13169, + "grad_norm": 1.3091653230611913, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 13169 + }, + { + "epoch": 0.1317, + "grad_norm": 1.1789551231881399, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 13170 + }, + { + "epoch": 0.13171, + "grad_norm": 1.2679235757864307, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 13171 + }, + { + "epoch": 0.13172, + "grad_norm": 1.0121238818707832, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13172 + }, + { + "epoch": 0.13173, + "grad_norm": 1.3674161299963243, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 13173 + }, + { + "epoch": 0.13174, + "grad_norm": 1.1333333512368313, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 13174 + }, + { + "epoch": 0.13175, + "grad_norm": 1.263735157123079, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 13175 + }, + { + "epoch": 0.13176, + "grad_norm": 1.4329249743192631, + "learning_rate": 0.003, + "loss": 4.022, + "step": 13176 + }, + { + "epoch": 0.13177, + "grad_norm": 1.113757810778901, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 13177 + }, + { + "epoch": 0.13178, + "grad_norm": 1.2033659807687154, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 13178 + }, + { + "epoch": 0.13179, + "grad_norm": 1.1629366378785435, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 13179 + }, + { + "epoch": 0.1318, + "grad_norm": 1.3602821104802627, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 13180 + }, + { + "epoch": 0.13181, + "grad_norm": 1.2445246941668198, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 13181 + }, + { + "epoch": 0.13182, + "grad_norm": 1.337209314159866, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 13182 + }, + { + "epoch": 0.13183, + "grad_norm": 0.9878572274233712, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 13183 + }, + { + "epoch": 0.13184, + "grad_norm": 1.0960981251343198, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 13184 + }, + { + "epoch": 0.13185, + "grad_norm": 1.3102949978898657, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 13185 + }, + { + "epoch": 0.13186, + "grad_norm": 1.205314079023528, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 13186 + }, + { + "epoch": 0.13187, + "grad_norm": 1.0255720665353423, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 13187 + }, + { + "epoch": 0.13188, + "grad_norm": 1.3262167257541027, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 13188 + }, + { + "epoch": 0.13189, + "grad_norm": 1.2168641109639486, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 13189 + }, + { + "epoch": 0.1319, + "grad_norm": 1.1889265011218615, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 13190 + }, + { + "epoch": 0.13191, + "grad_norm": 1.3805581899749395, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 13191 + }, + { + "epoch": 0.13192, + "grad_norm": 1.0413867655514306, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 13192 + }, + { + "epoch": 0.13193, + "grad_norm": 1.2492107648050148, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 13193 + }, + { + "epoch": 0.13194, + "grad_norm": 1.0030085267089808, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 13194 + }, + { + "epoch": 0.13195, + "grad_norm": 1.3532728033233967, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13195 + }, + { + "epoch": 0.13196, + "grad_norm": 1.3442955676133201, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 13196 + }, + { + "epoch": 0.13197, + "grad_norm": 1.1839299552626144, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 13197 + }, + { + "epoch": 0.13198, + "grad_norm": 1.0302625638882434, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 13198 + }, + { + "epoch": 0.13199, + "grad_norm": 1.2400658727503961, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 13199 + }, + { + "epoch": 0.132, + "grad_norm": 1.1317920385131712, + "learning_rate": 0.003, + "loss": 4.032, + "step": 13200 + }, + { + "epoch": 0.13201, + "grad_norm": 1.3070532346117103, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 13201 + }, + { + "epoch": 0.13202, + "grad_norm": 1.290994390815391, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 13202 + }, + { + "epoch": 0.13203, + "grad_norm": 1.3259193241549494, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 13203 + }, + { + "epoch": 0.13204, + "grad_norm": 1.1783424720493545, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 13204 + }, + { + "epoch": 0.13205, + "grad_norm": 1.170009083173674, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 13205 + }, + { + "epoch": 0.13206, + "grad_norm": 1.423622540381494, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13206 + }, + { + "epoch": 0.13207, + "grad_norm": 1.0283146612527605, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 13207 + }, + { + "epoch": 0.13208, + "grad_norm": 1.2998154192550682, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 13208 + }, + { + "epoch": 0.13209, + "grad_norm": 0.873572889486228, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 13209 + }, + { + "epoch": 0.1321, + "grad_norm": 1.2389240668521508, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 13210 + }, + { + "epoch": 0.13211, + "grad_norm": 1.5130364765500839, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 13211 + }, + { + "epoch": 0.13212, + "grad_norm": 1.3577792356335452, + "learning_rate": 0.003, + "loss": 4.025, + "step": 13212 + }, + { + "epoch": 0.13213, + "grad_norm": 0.8591220566534142, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 13213 + }, + { + "epoch": 0.13214, + "grad_norm": 0.9734006922172769, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 13214 + }, + { + "epoch": 0.13215, + "grad_norm": 1.3037427579729166, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 13215 + }, + { + "epoch": 0.13216, + "grad_norm": 1.0660492545368943, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 13216 + }, + { + "epoch": 0.13217, + "grad_norm": 1.3276671787509555, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 13217 + }, + { + "epoch": 0.13218, + "grad_norm": 0.9912861724784381, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 13218 + }, + { + "epoch": 0.13219, + "grad_norm": 1.1947328423567711, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 13219 + }, + { + "epoch": 0.1322, + "grad_norm": 1.1324627330357286, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 13220 + }, + { + "epoch": 0.13221, + "grad_norm": 1.2736609006663895, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 13221 + }, + { + "epoch": 0.13222, + "grad_norm": 1.2523335140850878, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 13222 + }, + { + "epoch": 0.13223, + "grad_norm": 1.125126287078147, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13223 + }, + { + "epoch": 0.13224, + "grad_norm": 1.3849418026736036, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 13224 + }, + { + "epoch": 0.13225, + "grad_norm": 1.1461342791536524, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 13225 + }, + { + "epoch": 0.13226, + "grad_norm": 1.4180290204833317, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 13226 + }, + { + "epoch": 0.13227, + "grad_norm": 1.014686375750144, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 13227 + }, + { + "epoch": 0.13228, + "grad_norm": 1.5037739153801797, + "learning_rate": 0.003, + "loss": 4.025, + "step": 13228 + }, + { + "epoch": 0.13229, + "grad_norm": 1.0446456028567737, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 13229 + }, + { + "epoch": 0.1323, + "grad_norm": 1.2431434834546604, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 13230 + }, + { + "epoch": 0.13231, + "grad_norm": 1.1595233786574235, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 13231 + }, + { + "epoch": 0.13232, + "grad_norm": 1.23700285851049, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 13232 + }, + { + "epoch": 0.13233, + "grad_norm": 1.2414313005494253, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 13233 + }, + { + "epoch": 0.13234, + "grad_norm": 1.4284272611351443, + "learning_rate": 0.003, + "loss": 4.012, + "step": 13234 + }, + { + "epoch": 0.13235, + "grad_norm": 1.1272287828969014, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 13235 + }, + { + "epoch": 0.13236, + "grad_norm": 1.0883267700976904, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 13236 + }, + { + "epoch": 0.13237, + "grad_norm": 1.3095687487861305, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 13237 + }, + { + "epoch": 0.13238, + "grad_norm": 1.178701415006733, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 13238 + }, + { + "epoch": 0.13239, + "grad_norm": 1.2027726294396186, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 13239 + }, + { + "epoch": 0.1324, + "grad_norm": 1.2377893040922567, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 13240 + }, + { + "epoch": 0.13241, + "grad_norm": 1.0267119343524598, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 13241 + }, + { + "epoch": 0.13242, + "grad_norm": 1.3012800497905883, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 13242 + }, + { + "epoch": 0.13243, + "grad_norm": 1.1614741672299007, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 13243 + }, + { + "epoch": 0.13244, + "grad_norm": 1.2023819031296115, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 13244 + }, + { + "epoch": 0.13245, + "grad_norm": 1.1969761782597237, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 13245 + }, + { + "epoch": 0.13246, + "grad_norm": 1.1736874650380753, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 13246 + }, + { + "epoch": 0.13247, + "grad_norm": 1.2217478030413764, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 13247 + }, + { + "epoch": 0.13248, + "grad_norm": 1.195487292425333, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 13248 + }, + { + "epoch": 0.13249, + "grad_norm": 1.3531113906627674, + "learning_rate": 0.003, + "loss": 4.004, + "step": 13249 + }, + { + "epoch": 0.1325, + "grad_norm": 1.1331994653407587, + "learning_rate": 0.003, + "loss": 4.018, + "step": 13250 + }, + { + "epoch": 0.13251, + "grad_norm": 1.3568915508653279, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 13251 + }, + { + "epoch": 0.13252, + "grad_norm": 1.1022967952260512, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 13252 + }, + { + "epoch": 0.13253, + "grad_norm": 1.4730317933527035, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 13253 + }, + { + "epoch": 0.13254, + "grad_norm": 1.1698133513210476, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 13254 + }, + { + "epoch": 0.13255, + "grad_norm": 1.3006938019160277, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13255 + }, + { + "epoch": 0.13256, + "grad_norm": 1.1089839963119312, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13256 + }, + { + "epoch": 0.13257, + "grad_norm": 1.2055086984398902, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 13257 + }, + { + "epoch": 0.13258, + "grad_norm": 1.0517577625863193, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 13258 + }, + { + "epoch": 0.13259, + "grad_norm": 1.0260203890418644, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 13259 + }, + { + "epoch": 0.1326, + "grad_norm": 1.1451620545525771, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 13260 + }, + { + "epoch": 0.13261, + "grad_norm": 1.3573920654937885, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 13261 + }, + { + "epoch": 0.13262, + "grad_norm": 1.2997613200686104, + "learning_rate": 0.003, + "loss": 4.011, + "step": 13262 + }, + { + "epoch": 0.13263, + "grad_norm": 1.162070533819352, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 13263 + }, + { + "epoch": 0.13264, + "grad_norm": 1.1860611604132822, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13264 + }, + { + "epoch": 0.13265, + "grad_norm": 1.155878268074166, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 13265 + }, + { + "epoch": 0.13266, + "grad_norm": 1.162897362742571, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 13266 + }, + { + "epoch": 0.13267, + "grad_norm": 1.2566274122158887, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 13267 + }, + { + "epoch": 0.13268, + "grad_norm": 1.0624053962002076, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 13268 + }, + { + "epoch": 0.13269, + "grad_norm": 1.146155331042088, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 13269 + }, + { + "epoch": 0.1327, + "grad_norm": 1.1523816823944344, + "learning_rate": 0.003, + "loss": 4.009, + "step": 13270 + }, + { + "epoch": 0.13271, + "grad_norm": 1.2822840784270244, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 13271 + }, + { + "epoch": 0.13272, + "grad_norm": 1.0786807154247118, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 13272 + }, + { + "epoch": 0.13273, + "grad_norm": 1.218673056995643, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 13273 + }, + { + "epoch": 0.13274, + "grad_norm": 1.3254738935758086, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 13274 + }, + { + "epoch": 0.13275, + "grad_norm": 1.2485461958782031, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 13275 + }, + { + "epoch": 0.13276, + "grad_norm": 1.0339646828439346, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 13276 + }, + { + "epoch": 0.13277, + "grad_norm": 1.2296936563599417, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 13277 + }, + { + "epoch": 0.13278, + "grad_norm": 1.1719359498015458, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 13278 + }, + { + "epoch": 0.13279, + "grad_norm": 1.2906893737489893, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 13279 + }, + { + "epoch": 0.1328, + "grad_norm": 0.9747260590672138, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 13280 + }, + { + "epoch": 0.13281, + "grad_norm": 1.3295126084209, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 13281 + }, + { + "epoch": 0.13282, + "grad_norm": 1.1543837502411551, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 13282 + }, + { + "epoch": 0.13283, + "grad_norm": 1.3388418810355132, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 13283 + }, + { + "epoch": 0.13284, + "grad_norm": 0.9952350561662376, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13284 + }, + { + "epoch": 0.13285, + "grad_norm": 1.4499501567801372, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13285 + }, + { + "epoch": 0.13286, + "grad_norm": 1.066590029594129, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 13286 + }, + { + "epoch": 0.13287, + "grad_norm": 1.388542523326155, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 13287 + }, + { + "epoch": 0.13288, + "grad_norm": 1.242952241814662, + "learning_rate": 0.003, + "loss": 4.029, + "step": 13288 + }, + { + "epoch": 0.13289, + "grad_norm": 1.2844622720087564, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 13289 + }, + { + "epoch": 0.1329, + "grad_norm": 1.1296034429914248, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 13290 + }, + { + "epoch": 0.13291, + "grad_norm": 1.1817733386973799, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 13291 + }, + { + "epoch": 0.13292, + "grad_norm": 1.227418218840078, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 13292 + }, + { + "epoch": 0.13293, + "grad_norm": 1.2934766559092665, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 13293 + }, + { + "epoch": 0.13294, + "grad_norm": 1.2324210240838256, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 13294 + }, + { + "epoch": 0.13295, + "grad_norm": 1.1190504319413743, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 13295 + }, + { + "epoch": 0.13296, + "grad_norm": 1.3831470071383796, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 13296 + }, + { + "epoch": 0.13297, + "grad_norm": 0.9024796111158896, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 13297 + }, + { + "epoch": 0.13298, + "grad_norm": 1.3130099662344958, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 13298 + }, + { + "epoch": 0.13299, + "grad_norm": 1.1678197485533626, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 13299 + }, + { + "epoch": 0.133, + "grad_norm": 1.2927678212090512, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 13300 + }, + { + "epoch": 0.13301, + "grad_norm": 1.1463633834307947, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 13301 + }, + { + "epoch": 0.13302, + "grad_norm": 1.3051173709107187, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 13302 + }, + { + "epoch": 0.13303, + "grad_norm": 1.136023830327464, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 13303 + }, + { + "epoch": 0.13304, + "grad_norm": 1.231542541977393, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 13304 + }, + { + "epoch": 0.13305, + "grad_norm": 1.0420078614674835, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 13305 + }, + { + "epoch": 0.13306, + "grad_norm": 1.4456939860953055, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 13306 + }, + { + "epoch": 0.13307, + "grad_norm": 0.9463696467956895, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 13307 + }, + { + "epoch": 0.13308, + "grad_norm": 1.372066712252764, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 13308 + }, + { + "epoch": 0.13309, + "grad_norm": 1.2982308263401467, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 13309 + }, + { + "epoch": 0.1331, + "grad_norm": 1.3021473169835807, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13310 + }, + { + "epoch": 0.13311, + "grad_norm": 1.1181520647256893, + "learning_rate": 0.003, + "loss": 4.005, + "step": 13311 + }, + { + "epoch": 0.13312, + "grad_norm": 1.1323773951622496, + "learning_rate": 0.003, + "loss": 4.03, + "step": 13312 + }, + { + "epoch": 0.13313, + "grad_norm": 1.2994672754854095, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 13313 + }, + { + "epoch": 0.13314, + "grad_norm": 1.1123813052260192, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13314 + }, + { + "epoch": 0.13315, + "grad_norm": 1.3356269811242296, + "learning_rate": 0.003, + "loss": 4.039, + "step": 13315 + }, + { + "epoch": 0.13316, + "grad_norm": 1.1307571245786385, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 13316 + }, + { + "epoch": 0.13317, + "grad_norm": 1.1289097193017215, + "learning_rate": 0.003, + "loss": 4.029, + "step": 13317 + }, + { + "epoch": 0.13318, + "grad_norm": 1.2008083845027793, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 13318 + }, + { + "epoch": 0.13319, + "grad_norm": 1.1306337491064893, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 13319 + }, + { + "epoch": 0.1332, + "grad_norm": 1.3679305192275055, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 13320 + }, + { + "epoch": 0.13321, + "grad_norm": 1.1711826264907252, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 13321 + }, + { + "epoch": 0.13322, + "grad_norm": 1.0966041565646945, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 13322 + }, + { + "epoch": 0.13323, + "grad_norm": 1.1551860383609442, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 13323 + }, + { + "epoch": 0.13324, + "grad_norm": 1.0761459065829688, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 13324 + }, + { + "epoch": 0.13325, + "grad_norm": 1.1855031922305135, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 13325 + }, + { + "epoch": 0.13326, + "grad_norm": 1.358087467790536, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 13326 + }, + { + "epoch": 0.13327, + "grad_norm": 1.5789407226797987, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 13327 + }, + { + "epoch": 0.13328, + "grad_norm": 1.0846902206425766, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 13328 + }, + { + "epoch": 0.13329, + "grad_norm": 1.166482834679772, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 13329 + }, + { + "epoch": 0.1333, + "grad_norm": 1.1676348442033728, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 13330 + }, + { + "epoch": 0.13331, + "grad_norm": 1.2387302200322514, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 13331 + }, + { + "epoch": 0.13332, + "grad_norm": 1.090022503966062, + "learning_rate": 0.003, + "loss": 4.016, + "step": 13332 + }, + { + "epoch": 0.13333, + "grad_norm": 1.1993557738278178, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 13333 + }, + { + "epoch": 0.13334, + "grad_norm": 1.149779889121198, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 13334 + }, + { + "epoch": 0.13335, + "grad_norm": 1.2404908878452907, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 13335 + }, + { + "epoch": 0.13336, + "grad_norm": 1.1097718936919116, + "learning_rate": 0.003, + "loss": 4.017, + "step": 13336 + }, + { + "epoch": 0.13337, + "grad_norm": 1.3923059951933716, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 13337 + }, + { + "epoch": 0.13338, + "grad_norm": 1.2188422973936779, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 13338 + }, + { + "epoch": 0.13339, + "grad_norm": 1.4947926128477245, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13339 + }, + { + "epoch": 0.1334, + "grad_norm": 1.135167360467982, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 13340 + }, + { + "epoch": 0.13341, + "grad_norm": 1.2957735877895509, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 13341 + }, + { + "epoch": 0.13342, + "grad_norm": 1.0937534826464805, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 13342 + }, + { + "epoch": 0.13343, + "grad_norm": 1.146491272433081, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 13343 + }, + { + "epoch": 0.13344, + "grad_norm": 1.4468045189337417, + "learning_rate": 0.003, + "loss": 4.062, + "step": 13344 + }, + { + "epoch": 0.13345, + "grad_norm": 1.2592429566456962, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 13345 + }, + { + "epoch": 0.13346, + "grad_norm": 1.2143728986915379, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 13346 + }, + { + "epoch": 0.13347, + "grad_norm": 1.1905439376198783, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 13347 + }, + { + "epoch": 0.13348, + "grad_norm": 1.1443391482742789, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 13348 + }, + { + "epoch": 0.13349, + "grad_norm": 1.246186743392689, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 13349 + }, + { + "epoch": 0.1335, + "grad_norm": 1.0806035133689187, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 13350 + }, + { + "epoch": 0.13351, + "grad_norm": 1.123461988084758, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 13351 + }, + { + "epoch": 0.13352, + "grad_norm": 1.1992586433931847, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 13352 + }, + { + "epoch": 0.13353, + "grad_norm": 1.2287700219343816, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13353 + }, + { + "epoch": 0.13354, + "grad_norm": 1.145713010144192, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 13354 + }, + { + "epoch": 0.13355, + "grad_norm": 1.3129522428667089, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 13355 + }, + { + "epoch": 0.13356, + "grad_norm": 1.1217225093906096, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13356 + }, + { + "epoch": 0.13357, + "grad_norm": 1.3610017363079503, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 13357 + }, + { + "epoch": 0.13358, + "grad_norm": 1.0659946872902204, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 13358 + }, + { + "epoch": 0.13359, + "grad_norm": 1.4579500688790328, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 13359 + }, + { + "epoch": 0.1336, + "grad_norm": 1.003934981079245, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 13360 + }, + { + "epoch": 0.13361, + "grad_norm": 1.1638155922468505, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 13361 + }, + { + "epoch": 0.13362, + "grad_norm": 1.0889598114731738, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 13362 + }, + { + "epoch": 0.13363, + "grad_norm": 1.3693945445048112, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 13363 + }, + { + "epoch": 0.13364, + "grad_norm": 1.0988775373011543, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 13364 + }, + { + "epoch": 0.13365, + "grad_norm": 1.4802228295419162, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 13365 + }, + { + "epoch": 0.13366, + "grad_norm": 1.1350414757360303, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13366 + }, + { + "epoch": 0.13367, + "grad_norm": 1.0970300311081125, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 13367 + }, + { + "epoch": 0.13368, + "grad_norm": 1.2364580439070338, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 13368 + }, + { + "epoch": 0.13369, + "grad_norm": 1.1393845065668011, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 13369 + }, + { + "epoch": 0.1337, + "grad_norm": 1.1533636621662187, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 13370 + }, + { + "epoch": 0.13371, + "grad_norm": 1.0816471655884832, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 13371 + }, + { + "epoch": 0.13372, + "grad_norm": 1.193792088943506, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 13372 + }, + { + "epoch": 0.13373, + "grad_norm": 1.1094174710891824, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 13373 + }, + { + "epoch": 0.13374, + "grad_norm": 1.2879292981648436, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 13374 + }, + { + "epoch": 0.13375, + "grad_norm": 1.1453786363334113, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 13375 + }, + { + "epoch": 0.13376, + "grad_norm": 1.3055465273095053, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 13376 + }, + { + "epoch": 0.13377, + "grad_norm": 1.1532149751861094, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 13377 + }, + { + "epoch": 0.13378, + "grad_norm": 1.2267252708924727, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 13378 + }, + { + "epoch": 0.13379, + "grad_norm": 1.2465260797365454, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 13379 + }, + { + "epoch": 0.1338, + "grad_norm": 1.1671338719435786, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 13380 + }, + { + "epoch": 0.13381, + "grad_norm": 1.2770301462955111, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 13381 + }, + { + "epoch": 0.13382, + "grad_norm": 1.3607247962128857, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 13382 + }, + { + "epoch": 0.13383, + "grad_norm": 1.036366389732506, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 13383 + }, + { + "epoch": 0.13384, + "grad_norm": 1.2065138074211352, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 13384 + }, + { + "epoch": 0.13385, + "grad_norm": 1.1857547899204484, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 13385 + }, + { + "epoch": 0.13386, + "grad_norm": 0.946051102788784, + "learning_rate": 0.003, + "loss": 4.023, + "step": 13386 + }, + { + "epoch": 0.13387, + "grad_norm": 1.1633476212241936, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 13387 + }, + { + "epoch": 0.13388, + "grad_norm": 1.3245812612445407, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 13388 + }, + { + "epoch": 0.13389, + "grad_norm": 1.134263149423251, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 13389 + }, + { + "epoch": 0.1339, + "grad_norm": 1.296265090606207, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 13390 + }, + { + "epoch": 0.13391, + "grad_norm": 1.1360413290115436, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 13391 + }, + { + "epoch": 0.13392, + "grad_norm": 1.291707542924984, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 13392 + }, + { + "epoch": 0.13393, + "grad_norm": 1.076626333520554, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 13393 + }, + { + "epoch": 0.13394, + "grad_norm": 1.2115304300049723, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 13394 + }, + { + "epoch": 0.13395, + "grad_norm": 1.126113520409547, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 13395 + }, + { + "epoch": 0.13396, + "grad_norm": 1.2100032511507974, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 13396 + }, + { + "epoch": 0.13397, + "grad_norm": 1.1413697699504317, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13397 + }, + { + "epoch": 0.13398, + "grad_norm": 1.1369134666653362, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 13398 + }, + { + "epoch": 0.13399, + "grad_norm": 1.141769147643521, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 13399 + }, + { + "epoch": 0.134, + "grad_norm": 1.351880588084034, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 13400 + }, + { + "epoch": 0.13401, + "grad_norm": 1.0694809056532624, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 13401 + }, + { + "epoch": 0.13402, + "grad_norm": 1.2462804234100007, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 13402 + }, + { + "epoch": 0.13403, + "grad_norm": 0.9550894770645565, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 13403 + }, + { + "epoch": 0.13404, + "grad_norm": 1.4004013738063046, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 13404 + }, + { + "epoch": 0.13405, + "grad_norm": 1.0182565499258522, + "learning_rate": 0.003, + "loss": 4.008, + "step": 13405 + }, + { + "epoch": 0.13406, + "grad_norm": 1.3389242782316388, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 13406 + }, + { + "epoch": 0.13407, + "grad_norm": 1.265463794551261, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 13407 + }, + { + "epoch": 0.13408, + "grad_norm": 1.334487167757332, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 13408 + }, + { + "epoch": 0.13409, + "grad_norm": 1.3576258794005125, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 13409 + }, + { + "epoch": 0.1341, + "grad_norm": 1.2184686143115655, + "learning_rate": 0.003, + "loss": 4.003, + "step": 13410 + }, + { + "epoch": 0.13411, + "grad_norm": 0.9916996275319055, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 13411 + }, + { + "epoch": 0.13412, + "grad_norm": 1.2216044612518433, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 13412 + }, + { + "epoch": 0.13413, + "grad_norm": 1.160783345128535, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 13413 + }, + { + "epoch": 0.13414, + "grad_norm": 1.1292722771811052, + "learning_rate": 0.003, + "loss": 4.042, + "step": 13414 + }, + { + "epoch": 0.13415, + "grad_norm": 1.126718355717055, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 13415 + }, + { + "epoch": 0.13416, + "grad_norm": 1.2476269931115544, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 13416 + }, + { + "epoch": 0.13417, + "grad_norm": 1.2048295152923227, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 13417 + }, + { + "epoch": 0.13418, + "grad_norm": 1.3428948792281432, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 13418 + }, + { + "epoch": 0.13419, + "grad_norm": 1.1181666148157166, + "learning_rate": 0.003, + "loss": 4.037, + "step": 13419 + }, + { + "epoch": 0.1342, + "grad_norm": 1.3215144632242453, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 13420 + }, + { + "epoch": 0.13421, + "grad_norm": 1.3097747341982395, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 13421 + }, + { + "epoch": 0.13422, + "grad_norm": 1.198201371504427, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 13422 + }, + { + "epoch": 0.13423, + "grad_norm": 1.2244440779414874, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 13423 + }, + { + "epoch": 0.13424, + "grad_norm": 0.9789279406920552, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 13424 + }, + { + "epoch": 0.13425, + "grad_norm": 1.419037471910816, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 13425 + }, + { + "epoch": 0.13426, + "grad_norm": 1.0024421860356587, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 13426 + }, + { + "epoch": 0.13427, + "grad_norm": 1.236195542106977, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 13427 + }, + { + "epoch": 0.13428, + "grad_norm": 1.2230106966663075, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 13428 + }, + { + "epoch": 0.13429, + "grad_norm": 1.2019836590046975, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 13429 + }, + { + "epoch": 0.1343, + "grad_norm": 1.4494027288332898, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 13430 + }, + { + "epoch": 0.13431, + "grad_norm": 1.192145526605227, + "learning_rate": 0.003, + "loss": 4.032, + "step": 13431 + }, + { + "epoch": 0.13432, + "grad_norm": 1.1248484191707095, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 13432 + }, + { + "epoch": 0.13433, + "grad_norm": 1.260785974024862, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 13433 + }, + { + "epoch": 0.13434, + "grad_norm": 1.0400060793363335, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 13434 + }, + { + "epoch": 0.13435, + "grad_norm": 1.0783745236421138, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 13435 + }, + { + "epoch": 0.13436, + "grad_norm": 1.165074049861491, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 13436 + }, + { + "epoch": 0.13437, + "grad_norm": 1.162590117493891, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 13437 + }, + { + "epoch": 0.13438, + "grad_norm": 1.0904670998164137, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 13438 + }, + { + "epoch": 0.13439, + "grad_norm": 1.3044658478058948, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 13439 + }, + { + "epoch": 0.1344, + "grad_norm": 1.2585749984772712, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 13440 + }, + { + "epoch": 0.13441, + "grad_norm": 1.1350577146238163, + "learning_rate": 0.003, + "loss": 4.004, + "step": 13441 + }, + { + "epoch": 0.13442, + "grad_norm": 1.2934824566661551, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 13442 + }, + { + "epoch": 0.13443, + "grad_norm": 1.1704384561374903, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 13443 + }, + { + "epoch": 0.13444, + "grad_norm": 1.2287067692197493, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 13444 + }, + { + "epoch": 0.13445, + "grad_norm": 1.052312126159789, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 13445 + }, + { + "epoch": 0.13446, + "grad_norm": 1.3351631871063565, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 13446 + }, + { + "epoch": 0.13447, + "grad_norm": 1.1456741213952397, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 13447 + }, + { + "epoch": 0.13448, + "grad_norm": 1.3178994702812536, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 13448 + }, + { + "epoch": 0.13449, + "grad_norm": 1.1376435500601898, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 13449 + }, + { + "epoch": 0.1345, + "grad_norm": 1.5577330564326097, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 13450 + }, + { + "epoch": 0.13451, + "grad_norm": 1.240799192639951, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 13451 + }, + { + "epoch": 0.13452, + "grad_norm": 1.3303368261452497, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 13452 + }, + { + "epoch": 0.13453, + "grad_norm": 0.9864499235298809, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 13453 + }, + { + "epoch": 0.13454, + "grad_norm": 1.3913306919812294, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 13454 + }, + { + "epoch": 0.13455, + "grad_norm": 0.9262772369745972, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 13455 + }, + { + "epoch": 0.13456, + "grad_norm": 1.1715238243590518, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 13456 + }, + { + "epoch": 0.13457, + "grad_norm": 1.007775791407712, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 13457 + }, + { + "epoch": 0.13458, + "grad_norm": 1.4373867057512402, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 13458 + }, + { + "epoch": 0.13459, + "grad_norm": 1.2087819192138398, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 13459 + }, + { + "epoch": 0.1346, + "grad_norm": 1.2761546089109206, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 13460 + }, + { + "epoch": 0.13461, + "grad_norm": 1.3195149048198116, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 13461 + }, + { + "epoch": 0.13462, + "grad_norm": 1.4298685154422084, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 13462 + }, + { + "epoch": 0.13463, + "grad_norm": 1.0232005094391101, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 13463 + }, + { + "epoch": 0.13464, + "grad_norm": 1.3554979757755057, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 13464 + }, + { + "epoch": 0.13465, + "grad_norm": 1.1332202658140447, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 13465 + }, + { + "epoch": 0.13466, + "grad_norm": 1.3885916310764128, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 13466 + }, + { + "epoch": 0.13467, + "grad_norm": 1.1009272857461163, + "learning_rate": 0.003, + "loss": 4.048, + "step": 13467 + }, + { + "epoch": 0.13468, + "grad_norm": 1.291927423311447, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 13468 + }, + { + "epoch": 0.13469, + "grad_norm": 1.2553231921644208, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 13469 + }, + { + "epoch": 0.1347, + "grad_norm": 1.5595510592158073, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 13470 + }, + { + "epoch": 0.13471, + "grad_norm": 1.4597265580853351, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 13471 + }, + { + "epoch": 0.13472, + "grad_norm": 1.06169793876242, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 13472 + }, + { + "epoch": 0.13473, + "grad_norm": 1.1661843523771311, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 13473 + }, + { + "epoch": 0.13474, + "grad_norm": 1.3141625224505493, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 13474 + }, + { + "epoch": 0.13475, + "grad_norm": 1.0191480284471333, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 13475 + }, + { + "epoch": 0.13476, + "grad_norm": 1.3593921109749711, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 13476 + }, + { + "epoch": 0.13477, + "grad_norm": 1.029868013350082, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 13477 + }, + { + "epoch": 0.13478, + "grad_norm": 1.220297263485053, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 13478 + }, + { + "epoch": 0.13479, + "grad_norm": 1.2216981759838992, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 13479 + }, + { + "epoch": 0.1348, + "grad_norm": 1.1588932875116496, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 13480 + }, + { + "epoch": 0.13481, + "grad_norm": 1.04814139318589, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 13481 + }, + { + "epoch": 0.13482, + "grad_norm": 1.27331759602881, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 13482 + }, + { + "epoch": 0.13483, + "grad_norm": 1.2024848177421057, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 13483 + }, + { + "epoch": 0.13484, + "grad_norm": 1.1441160988818182, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 13484 + }, + { + "epoch": 0.13485, + "grad_norm": 1.2492437430898804, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 13485 + }, + { + "epoch": 0.13486, + "grad_norm": 1.2067132902882773, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 13486 + }, + { + "epoch": 0.13487, + "grad_norm": 1.1942540105040946, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 13487 + }, + { + "epoch": 0.13488, + "grad_norm": 1.3019162946617349, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 13488 + }, + { + "epoch": 0.13489, + "grad_norm": 1.076361817110379, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 13489 + }, + { + "epoch": 0.1349, + "grad_norm": 1.0367542395991136, + "learning_rate": 0.003, + "loss": 4.019, + "step": 13490 + }, + { + "epoch": 0.13491, + "grad_norm": 1.1496630472442426, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 13491 + }, + { + "epoch": 0.13492, + "grad_norm": 1.4714537384829591, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13492 + }, + { + "epoch": 0.13493, + "grad_norm": 1.2095126352691563, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 13493 + }, + { + "epoch": 0.13494, + "grad_norm": 1.2689510126659043, + "learning_rate": 0.003, + "loss": 4.017, + "step": 13494 + }, + { + "epoch": 0.13495, + "grad_norm": 1.1657755327513644, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 13495 + }, + { + "epoch": 0.13496, + "grad_norm": 1.1697449439005412, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 13496 + }, + { + "epoch": 0.13497, + "grad_norm": 1.2346266760881337, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 13497 + }, + { + "epoch": 0.13498, + "grad_norm": 1.6136535135218877, + "learning_rate": 0.003, + "loss": 4.037, + "step": 13498 + }, + { + "epoch": 0.13499, + "grad_norm": 1.1067394699171362, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 13499 + }, + { + "epoch": 0.135, + "grad_norm": 1.266309341268125, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 13500 + }, + { + "epoch": 0.13501, + "grad_norm": 1.2728591837186685, + "learning_rate": 0.003, + "loss": 4.043, + "step": 13501 + }, + { + "epoch": 0.13502, + "grad_norm": 1.2093142337945428, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 13502 + }, + { + "epoch": 0.13503, + "grad_norm": 1.0891243791379945, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 13503 + }, + { + "epoch": 0.13504, + "grad_norm": 1.121546991957497, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 13504 + }, + { + "epoch": 0.13505, + "grad_norm": 1.290359557449459, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 13505 + }, + { + "epoch": 0.13506, + "grad_norm": 1.1792088578482778, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 13506 + }, + { + "epoch": 0.13507, + "grad_norm": 1.2456337380797515, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 13507 + }, + { + "epoch": 0.13508, + "grad_norm": 1.4020559997019646, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13508 + }, + { + "epoch": 0.13509, + "grad_norm": 1.1533109080710795, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 13509 + }, + { + "epoch": 0.1351, + "grad_norm": 1.2745718197348026, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 13510 + }, + { + "epoch": 0.13511, + "grad_norm": 1.3014225992516306, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13511 + }, + { + "epoch": 0.13512, + "grad_norm": 1.0258185120614944, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 13512 + }, + { + "epoch": 0.13513, + "grad_norm": 1.2987664783952992, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 13513 + }, + { + "epoch": 0.13514, + "grad_norm": 1.0301254210890254, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 13514 + }, + { + "epoch": 0.13515, + "grad_norm": 1.4063675226394983, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 13515 + }, + { + "epoch": 0.13516, + "grad_norm": 1.06062614177715, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 13516 + }, + { + "epoch": 0.13517, + "grad_norm": 1.2326757174346914, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 13517 + }, + { + "epoch": 0.13518, + "grad_norm": 0.9967612734492626, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 13518 + }, + { + "epoch": 0.13519, + "grad_norm": 1.3638494530839866, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 13519 + }, + { + "epoch": 0.1352, + "grad_norm": 0.9993900606979094, + "learning_rate": 0.003, + "loss": 3.997, + "step": 13520 + }, + { + "epoch": 0.13521, + "grad_norm": 1.2647391132461463, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 13521 + }, + { + "epoch": 0.13522, + "grad_norm": 1.076840147795485, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 13522 + }, + { + "epoch": 0.13523, + "grad_norm": 1.536667316815823, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 13523 + }, + { + "epoch": 0.13524, + "grad_norm": 1.1907360604879682, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 13524 + }, + { + "epoch": 0.13525, + "grad_norm": 1.1183691818173742, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 13525 + }, + { + "epoch": 0.13526, + "grad_norm": 1.1958633957563631, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 13526 + }, + { + "epoch": 0.13527, + "grad_norm": 1.1733058300297852, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 13527 + }, + { + "epoch": 0.13528, + "grad_norm": 1.2818225476163603, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 13528 + }, + { + "epoch": 0.13529, + "grad_norm": 1.40427771743178, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 13529 + }, + { + "epoch": 0.1353, + "grad_norm": 1.0559809394022808, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13530 + }, + { + "epoch": 0.13531, + "grad_norm": 1.2141375395885508, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 13531 + }, + { + "epoch": 0.13532, + "grad_norm": 1.1631699152071382, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 13532 + }, + { + "epoch": 0.13533, + "grad_norm": 1.1435788324807838, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 13533 + }, + { + "epoch": 0.13534, + "grad_norm": 1.0902001010477733, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 13534 + }, + { + "epoch": 0.13535, + "grad_norm": 1.3308338913254787, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 13535 + }, + { + "epoch": 0.13536, + "grad_norm": 1.1546192103612023, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13536 + }, + { + "epoch": 0.13537, + "grad_norm": 1.0527740548229891, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 13537 + }, + { + "epoch": 0.13538, + "grad_norm": 1.3938719152318257, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 13538 + }, + { + "epoch": 0.13539, + "grad_norm": 1.294181469325066, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 13539 + }, + { + "epoch": 0.1354, + "grad_norm": 1.3313605648250066, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 13540 + }, + { + "epoch": 0.13541, + "grad_norm": 1.1090694180026441, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 13541 + }, + { + "epoch": 0.13542, + "grad_norm": 1.2566169447470952, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 13542 + }, + { + "epoch": 0.13543, + "grad_norm": 1.1364640575819667, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 13543 + }, + { + "epoch": 0.13544, + "grad_norm": 1.238508893201822, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 13544 + }, + { + "epoch": 0.13545, + "grad_norm": 1.3426050575470458, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 13545 + }, + { + "epoch": 0.13546, + "grad_norm": 1.1224828710678576, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13546 + }, + { + "epoch": 0.13547, + "grad_norm": 1.4489054917469295, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 13547 + }, + { + "epoch": 0.13548, + "grad_norm": 0.9563686440940138, + "learning_rate": 0.003, + "loss": 4.027, + "step": 13548 + }, + { + "epoch": 0.13549, + "grad_norm": 1.2742319396351343, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 13549 + }, + { + "epoch": 0.1355, + "grad_norm": 1.2331965104642149, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 13550 + }, + { + "epoch": 0.13551, + "grad_norm": 1.1910295938637434, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 13551 + }, + { + "epoch": 0.13552, + "grad_norm": 1.2063895895243764, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 13552 + }, + { + "epoch": 0.13553, + "grad_norm": 1.3495585085551443, + "learning_rate": 0.003, + "loss": 4.058, + "step": 13553 + }, + { + "epoch": 0.13554, + "grad_norm": 1.3261178312946502, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 13554 + }, + { + "epoch": 0.13555, + "grad_norm": 1.2671382357999805, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 13555 + }, + { + "epoch": 0.13556, + "grad_norm": 1.013440546564239, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 13556 + }, + { + "epoch": 0.13557, + "grad_norm": 1.358405967910683, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 13557 + }, + { + "epoch": 0.13558, + "grad_norm": 0.9529425587072567, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 13558 + }, + { + "epoch": 0.13559, + "grad_norm": 1.3926067932351842, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 13559 + }, + { + "epoch": 0.1356, + "grad_norm": 1.1927999409657768, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 13560 + }, + { + "epoch": 0.13561, + "grad_norm": 1.2975418476490068, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 13561 + }, + { + "epoch": 0.13562, + "grad_norm": 1.268040128104862, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 13562 + }, + { + "epoch": 0.13563, + "grad_norm": 1.068965547878699, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 13563 + }, + { + "epoch": 0.13564, + "grad_norm": 1.2462949090924584, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 13564 + }, + { + "epoch": 0.13565, + "grad_norm": 1.0262326038063978, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 13565 + }, + { + "epoch": 0.13566, + "grad_norm": 1.241331896159747, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 13566 + }, + { + "epoch": 0.13567, + "grad_norm": 1.106533598252521, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 13567 + }, + { + "epoch": 0.13568, + "grad_norm": 1.2845913986916169, + "learning_rate": 0.003, + "loss": 4.022, + "step": 13568 + }, + { + "epoch": 0.13569, + "grad_norm": 1.0934200649882952, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 13569 + }, + { + "epoch": 0.1357, + "grad_norm": 1.3469028866488575, + "learning_rate": 0.003, + "loss": 4.032, + "step": 13570 + }, + { + "epoch": 0.13571, + "grad_norm": 1.1323505460539465, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 13571 + }, + { + "epoch": 0.13572, + "grad_norm": 1.3450317600710857, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 13572 + }, + { + "epoch": 0.13573, + "grad_norm": 1.217539370131012, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 13573 + }, + { + "epoch": 0.13574, + "grad_norm": 1.3619147322432303, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 13574 + }, + { + "epoch": 0.13575, + "grad_norm": 1.2943028459713686, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 13575 + }, + { + "epoch": 0.13576, + "grad_norm": 1.056776897622436, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 13576 + }, + { + "epoch": 0.13577, + "grad_norm": 1.372028195433907, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 13577 + }, + { + "epoch": 0.13578, + "grad_norm": 1.0412699789951014, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 13578 + }, + { + "epoch": 0.13579, + "grad_norm": 1.375918893580882, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 13579 + }, + { + "epoch": 0.1358, + "grad_norm": 1.0483388215012421, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13580 + }, + { + "epoch": 0.13581, + "grad_norm": 1.2980414345055213, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 13581 + }, + { + "epoch": 0.13582, + "grad_norm": 1.0353181140648553, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 13582 + }, + { + "epoch": 0.13583, + "grad_norm": 1.265469273405755, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 13583 + }, + { + "epoch": 0.13584, + "grad_norm": 1.0112357228896525, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 13584 + }, + { + "epoch": 0.13585, + "grad_norm": 1.343005759231974, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 13585 + }, + { + "epoch": 0.13586, + "grad_norm": 1.2175197887500093, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 13586 + }, + { + "epoch": 0.13587, + "grad_norm": 1.179478814557246, + "learning_rate": 0.003, + "loss": 4.007, + "step": 13587 + }, + { + "epoch": 0.13588, + "grad_norm": 1.307292876356786, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 13588 + }, + { + "epoch": 0.13589, + "grad_norm": 1.2162754598432555, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 13589 + }, + { + "epoch": 0.1359, + "grad_norm": 1.2257717680755864, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 13590 + }, + { + "epoch": 0.13591, + "grad_norm": 1.3668360738196097, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 13591 + }, + { + "epoch": 0.13592, + "grad_norm": 1.1463625785724558, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 13592 + }, + { + "epoch": 0.13593, + "grad_norm": 1.2555792951194609, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 13593 + }, + { + "epoch": 0.13594, + "grad_norm": 1.136425816545672, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 13594 + }, + { + "epoch": 0.13595, + "grad_norm": 1.3017380885036873, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 13595 + }, + { + "epoch": 0.13596, + "grad_norm": 1.2390442231201604, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 13596 + }, + { + "epoch": 0.13597, + "grad_norm": 1.0792428908460772, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 13597 + }, + { + "epoch": 0.13598, + "grad_norm": 1.1627665380886698, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 13598 + }, + { + "epoch": 0.13599, + "grad_norm": 1.0891261570324686, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 13599 + }, + { + "epoch": 0.136, + "grad_norm": 1.3290981851394779, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 13600 + }, + { + "epoch": 0.13601, + "grad_norm": 1.1191130692722162, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 13601 + }, + { + "epoch": 0.13602, + "grad_norm": 1.14018704398269, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 13602 + }, + { + "epoch": 0.13603, + "grad_norm": 1.4134013940831658, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 13603 + }, + { + "epoch": 0.13604, + "grad_norm": 0.9919395717353615, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 13604 + }, + { + "epoch": 0.13605, + "grad_norm": 1.4685076537988184, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 13605 + }, + { + "epoch": 0.13606, + "grad_norm": 1.3043209564442602, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 13606 + }, + { + "epoch": 0.13607, + "grad_norm": 1.2720661225646608, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 13607 + }, + { + "epoch": 0.13608, + "grad_norm": 1.1218482561071894, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 13608 + }, + { + "epoch": 0.13609, + "grad_norm": 1.1870233607488427, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 13609 + }, + { + "epoch": 0.1361, + "grad_norm": 1.4120658302714315, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 13610 + }, + { + "epoch": 0.13611, + "grad_norm": 1.068516624222778, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 13611 + }, + { + "epoch": 0.13612, + "grad_norm": 1.245146046438301, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 13612 + }, + { + "epoch": 0.13613, + "grad_norm": 1.145508954883413, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13613 + }, + { + "epoch": 0.13614, + "grad_norm": 1.055471911251265, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 13614 + }, + { + "epoch": 0.13615, + "grad_norm": 1.1450571637294198, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 13615 + }, + { + "epoch": 0.13616, + "grad_norm": 1.148179332102041, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 13616 + }, + { + "epoch": 0.13617, + "grad_norm": 1.2645193541346786, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 13617 + }, + { + "epoch": 0.13618, + "grad_norm": 1.3830179107987441, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 13618 + }, + { + "epoch": 0.13619, + "grad_norm": 1.1986654364411988, + "learning_rate": 0.003, + "loss": 4.036, + "step": 13619 + }, + { + "epoch": 0.1362, + "grad_norm": 1.1818480902296788, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 13620 + }, + { + "epoch": 0.13621, + "grad_norm": 1.1745157716112689, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 13621 + }, + { + "epoch": 0.13622, + "grad_norm": 1.1066332563589678, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13622 + }, + { + "epoch": 0.13623, + "grad_norm": 1.3645281692577584, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 13623 + }, + { + "epoch": 0.13624, + "grad_norm": 1.1929185751443674, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 13624 + }, + { + "epoch": 0.13625, + "grad_norm": 1.4248611420754493, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 13625 + }, + { + "epoch": 0.13626, + "grad_norm": 1.186299479007345, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 13626 + }, + { + "epoch": 0.13627, + "grad_norm": 1.0788810221878606, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 13627 + }, + { + "epoch": 0.13628, + "grad_norm": 1.2977698017293953, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 13628 + }, + { + "epoch": 0.13629, + "grad_norm": 1.07972580227147, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 13629 + }, + { + "epoch": 0.1363, + "grad_norm": 1.1688938042282193, + "learning_rate": 0.003, + "loss": 3.986, + "step": 13630 + }, + { + "epoch": 0.13631, + "grad_norm": 1.3352889662898282, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 13631 + }, + { + "epoch": 0.13632, + "grad_norm": 1.1365058041036933, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 13632 + }, + { + "epoch": 0.13633, + "grad_norm": 1.315315923471128, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 13633 + }, + { + "epoch": 0.13634, + "grad_norm": 1.0931804813692574, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 13634 + }, + { + "epoch": 0.13635, + "grad_norm": 1.264910695283981, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 13635 + }, + { + "epoch": 0.13636, + "grad_norm": 1.2708625560327722, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 13636 + }, + { + "epoch": 0.13637, + "grad_norm": 0.9990499280248041, + "learning_rate": 0.003, + "loss": 3.997, + "step": 13637 + }, + { + "epoch": 0.13638, + "grad_norm": 1.2961716625668893, + "learning_rate": 0.003, + "loss": 4.046, + "step": 13638 + }, + { + "epoch": 0.13639, + "grad_norm": 1.1163179620633081, + "learning_rate": 0.003, + "loss": 4.024, + "step": 13639 + }, + { + "epoch": 0.1364, + "grad_norm": 1.1076994011515442, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 13640 + }, + { + "epoch": 0.13641, + "grad_norm": 1.1762842182555493, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 13641 + }, + { + "epoch": 0.13642, + "grad_norm": 1.1464394748013567, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 13642 + }, + { + "epoch": 0.13643, + "grad_norm": 1.2598077685915399, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 13643 + }, + { + "epoch": 0.13644, + "grad_norm": 1.060741487532991, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 13644 + }, + { + "epoch": 0.13645, + "grad_norm": 1.377870157370045, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 13645 + }, + { + "epoch": 0.13646, + "grad_norm": 1.1109566861432008, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 13646 + }, + { + "epoch": 0.13647, + "grad_norm": 1.3433279023666778, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 13647 + }, + { + "epoch": 0.13648, + "grad_norm": 0.9491736753233733, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 13648 + }, + { + "epoch": 0.13649, + "grad_norm": 1.5014629142589342, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 13649 + }, + { + "epoch": 0.1365, + "grad_norm": 1.2241394783719237, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 13650 + }, + { + "epoch": 0.13651, + "grad_norm": 1.4139447872011477, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 13651 + }, + { + "epoch": 0.13652, + "grad_norm": 0.996084137685544, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 13652 + }, + { + "epoch": 0.13653, + "grad_norm": 1.0902948797409415, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 13653 + }, + { + "epoch": 0.13654, + "grad_norm": 1.3141932106860696, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 13654 + }, + { + "epoch": 0.13655, + "grad_norm": 1.1740235304999693, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 13655 + }, + { + "epoch": 0.13656, + "grad_norm": 1.2292046645803103, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 13656 + }, + { + "epoch": 0.13657, + "grad_norm": 1.0944277948772945, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 13657 + }, + { + "epoch": 0.13658, + "grad_norm": 1.212064236516267, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 13658 + }, + { + "epoch": 0.13659, + "grad_norm": 1.2888223611027119, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 13659 + }, + { + "epoch": 0.1366, + "grad_norm": 1.2746851628935965, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 13660 + }, + { + "epoch": 0.13661, + "grad_norm": 1.4234278610089575, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13661 + }, + { + "epoch": 0.13662, + "grad_norm": 1.2116319288225803, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13662 + }, + { + "epoch": 0.13663, + "grad_norm": 1.328145092229361, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 13663 + }, + { + "epoch": 0.13664, + "grad_norm": 0.9959449839391749, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 13664 + }, + { + "epoch": 0.13665, + "grad_norm": 1.2867223826779501, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 13665 + }, + { + "epoch": 0.13666, + "grad_norm": 0.9853202985653796, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 13666 + }, + { + "epoch": 0.13667, + "grad_norm": 1.3798038152578214, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 13667 + }, + { + "epoch": 0.13668, + "grad_norm": 1.2428031701336075, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 13668 + }, + { + "epoch": 0.13669, + "grad_norm": 1.0698514700923212, + "learning_rate": 0.003, + "loss": 4.015, + "step": 13669 + }, + { + "epoch": 0.1367, + "grad_norm": 1.3737478426089011, + "learning_rate": 0.003, + "loss": 4.039, + "step": 13670 + }, + { + "epoch": 0.13671, + "grad_norm": 1.122164008438741, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 13671 + }, + { + "epoch": 0.13672, + "grad_norm": 1.5307147958375853, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 13672 + }, + { + "epoch": 0.13673, + "grad_norm": 0.899868551945437, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 13673 + }, + { + "epoch": 0.13674, + "grad_norm": 1.2097351095429887, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 13674 + }, + { + "epoch": 0.13675, + "grad_norm": 1.149552133151396, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 13675 + }, + { + "epoch": 0.13676, + "grad_norm": 1.5380813479804742, + "learning_rate": 0.003, + "loss": 3.994, + "step": 13676 + }, + { + "epoch": 0.13677, + "grad_norm": 1.1197435050673568, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 13677 + }, + { + "epoch": 0.13678, + "grad_norm": 1.2370751158278261, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 13678 + }, + { + "epoch": 0.13679, + "grad_norm": 1.37577427018495, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 13679 + }, + { + "epoch": 0.1368, + "grad_norm": 1.2362491627572139, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 13680 + }, + { + "epoch": 0.13681, + "grad_norm": 1.133991783462781, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 13681 + }, + { + "epoch": 0.13682, + "grad_norm": 1.3599758993204916, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 13682 + }, + { + "epoch": 0.13683, + "grad_norm": 1.0764047723814174, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13683 + }, + { + "epoch": 0.13684, + "grad_norm": 1.2192253420881678, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 13684 + }, + { + "epoch": 0.13685, + "grad_norm": 1.082565952772552, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 13685 + }, + { + "epoch": 0.13686, + "grad_norm": 1.1530607333400629, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 13686 + }, + { + "epoch": 0.13687, + "grad_norm": 1.1444553141251446, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 13687 + }, + { + "epoch": 0.13688, + "grad_norm": 1.1769917803189258, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13688 + }, + { + "epoch": 0.13689, + "grad_norm": 1.054751783887151, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 13689 + }, + { + "epoch": 0.1369, + "grad_norm": 1.5601870027824791, + "learning_rate": 0.003, + "loss": 4.043, + "step": 13690 + }, + { + "epoch": 0.13691, + "grad_norm": 1.093268599921564, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 13691 + }, + { + "epoch": 0.13692, + "grad_norm": 1.1917534985211347, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 13692 + }, + { + "epoch": 0.13693, + "grad_norm": 1.2994734700544726, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13693 + }, + { + "epoch": 0.13694, + "grad_norm": 1.1547766676263558, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 13694 + }, + { + "epoch": 0.13695, + "grad_norm": 1.1635361967495015, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 13695 + }, + { + "epoch": 0.13696, + "grad_norm": 1.2507602619889238, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 13696 + }, + { + "epoch": 0.13697, + "grad_norm": 0.9315906290639528, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 13697 + }, + { + "epoch": 0.13698, + "grad_norm": 1.1915880289336969, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 13698 + }, + { + "epoch": 0.13699, + "grad_norm": 1.2447972729649426, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 13699 + }, + { + "epoch": 0.137, + "grad_norm": 1.115048622483683, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 13700 + }, + { + "epoch": 0.13701, + "grad_norm": 1.1805289485345356, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 13701 + }, + { + "epoch": 0.13702, + "grad_norm": 1.0326553481841696, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 13702 + }, + { + "epoch": 0.13703, + "grad_norm": 1.1744583181917279, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 13703 + }, + { + "epoch": 0.13704, + "grad_norm": 1.1175079166697401, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13704 + }, + { + "epoch": 0.13705, + "grad_norm": 1.224709949818222, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 13705 + }, + { + "epoch": 0.13706, + "grad_norm": 1.36848019240307, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 13706 + }, + { + "epoch": 0.13707, + "grad_norm": 1.1390307172209935, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 13707 + }, + { + "epoch": 0.13708, + "grad_norm": 1.090064422114458, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13708 + }, + { + "epoch": 0.13709, + "grad_norm": 1.2237984657350607, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 13709 + }, + { + "epoch": 0.1371, + "grad_norm": 1.168599884032333, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 13710 + }, + { + "epoch": 0.13711, + "grad_norm": 1.6158106815844115, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 13711 + }, + { + "epoch": 0.13712, + "grad_norm": 0.9296745648114974, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 13712 + }, + { + "epoch": 0.13713, + "grad_norm": 1.184943380430527, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 13713 + }, + { + "epoch": 0.13714, + "grad_norm": 1.2657594217766623, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 13714 + }, + { + "epoch": 0.13715, + "grad_norm": 1.2599122393116169, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 13715 + }, + { + "epoch": 0.13716, + "grad_norm": 1.1314866178955236, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 13716 + }, + { + "epoch": 0.13717, + "grad_norm": 1.1669368270819687, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 13717 + }, + { + "epoch": 0.13718, + "grad_norm": 1.3826944955399298, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 13718 + }, + { + "epoch": 0.13719, + "grad_norm": 1.2228088691409713, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 13719 + }, + { + "epoch": 0.1372, + "grad_norm": 1.1745635287793246, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 13720 + }, + { + "epoch": 0.13721, + "grad_norm": 1.1651313503452432, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 13721 + }, + { + "epoch": 0.13722, + "grad_norm": 1.1640986926150954, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 13722 + }, + { + "epoch": 0.13723, + "grad_norm": 1.349857413894611, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 13723 + }, + { + "epoch": 0.13724, + "grad_norm": 1.0418103753447379, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 13724 + }, + { + "epoch": 0.13725, + "grad_norm": 1.420188473153875, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 13725 + }, + { + "epoch": 0.13726, + "grad_norm": 1.0133290806749176, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 13726 + }, + { + "epoch": 0.13727, + "grad_norm": 1.425719934360195, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 13727 + }, + { + "epoch": 0.13728, + "grad_norm": 0.9872080493101567, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 13728 + }, + { + "epoch": 0.13729, + "grad_norm": 1.412401336069546, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 13729 + }, + { + "epoch": 0.1373, + "grad_norm": 1.1511414874159784, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 13730 + }, + { + "epoch": 0.13731, + "grad_norm": 1.4826903861671865, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 13731 + }, + { + "epoch": 0.13732, + "grad_norm": 1.0612799057595916, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 13732 + }, + { + "epoch": 0.13733, + "grad_norm": 1.1130211645705217, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 13733 + }, + { + "epoch": 0.13734, + "grad_norm": 1.1992034429309264, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 13734 + }, + { + "epoch": 0.13735, + "grad_norm": 1.1869985348367593, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 13735 + }, + { + "epoch": 0.13736, + "grad_norm": 1.0581702676513116, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13736 + }, + { + "epoch": 0.13737, + "grad_norm": 1.2556745480544078, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 13737 + }, + { + "epoch": 0.13738, + "grad_norm": 1.0967254187692979, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 13738 + }, + { + "epoch": 0.13739, + "grad_norm": 1.2665754205291968, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 13739 + }, + { + "epoch": 0.1374, + "grad_norm": 0.8686616512028059, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 13740 + }, + { + "epoch": 0.13741, + "grad_norm": 1.0704742058175276, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 13741 + }, + { + "epoch": 0.13742, + "grad_norm": 1.479336455824835, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 13742 + }, + { + "epoch": 0.13743, + "grad_norm": 1.1989083802465799, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 13743 + }, + { + "epoch": 0.13744, + "grad_norm": 1.2784732801027345, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 13744 + }, + { + "epoch": 0.13745, + "grad_norm": 1.0593265535350147, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 13745 + }, + { + "epoch": 0.13746, + "grad_norm": 1.3268605067423047, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 13746 + }, + { + "epoch": 0.13747, + "grad_norm": 1.073961143923257, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 13747 + }, + { + "epoch": 0.13748, + "grad_norm": 1.3522000440774327, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 13748 + }, + { + "epoch": 0.13749, + "grad_norm": 0.8957601239200287, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 13749 + }, + { + "epoch": 0.1375, + "grad_norm": 1.3417390194539345, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 13750 + }, + { + "epoch": 0.13751, + "grad_norm": 1.5273003202455322, + "learning_rate": 0.003, + "loss": 4.056, + "step": 13751 + }, + { + "epoch": 0.13752, + "grad_norm": 0.9674239717893509, + "learning_rate": 0.003, + "loss": 3.989, + "step": 13752 + }, + { + "epoch": 0.13753, + "grad_norm": 1.331551362321103, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 13753 + }, + { + "epoch": 0.13754, + "grad_norm": 0.9727047760605053, + "learning_rate": 0.003, + "loss": 4.029, + "step": 13754 + }, + { + "epoch": 0.13755, + "grad_norm": 1.3971142255392022, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 13755 + }, + { + "epoch": 0.13756, + "grad_norm": 1.144141970264651, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 13756 + }, + { + "epoch": 0.13757, + "grad_norm": 1.570052296011186, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 13757 + }, + { + "epoch": 0.13758, + "grad_norm": 1.2885569367110294, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 13758 + }, + { + "epoch": 0.13759, + "grad_norm": 1.2012167993907945, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 13759 + }, + { + "epoch": 0.1376, + "grad_norm": 1.1971123342918735, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 13760 + }, + { + "epoch": 0.13761, + "grad_norm": 1.2711990626473382, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 13761 + }, + { + "epoch": 0.13762, + "grad_norm": 1.2563892597273907, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 13762 + }, + { + "epoch": 0.13763, + "grad_norm": 1.3999239237162613, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 13763 + }, + { + "epoch": 0.13764, + "grad_norm": 0.9286033918747508, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 13764 + }, + { + "epoch": 0.13765, + "grad_norm": 1.229986753069097, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 13765 + }, + { + "epoch": 0.13766, + "grad_norm": 1.2910295749295635, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 13766 + }, + { + "epoch": 0.13767, + "grad_norm": 1.2404265094621845, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 13767 + }, + { + "epoch": 0.13768, + "grad_norm": 1.380749421877674, + "learning_rate": 0.003, + "loss": 4.03, + "step": 13768 + }, + { + "epoch": 0.13769, + "grad_norm": 1.055616536431569, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 13769 + }, + { + "epoch": 0.1377, + "grad_norm": 1.261496186434932, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 13770 + }, + { + "epoch": 0.13771, + "grad_norm": 1.078420340221653, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 13771 + }, + { + "epoch": 0.13772, + "grad_norm": 1.1886135398725044, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 13772 + }, + { + "epoch": 0.13773, + "grad_norm": 1.1260744620890015, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 13773 + }, + { + "epoch": 0.13774, + "grad_norm": 1.3720426378978656, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 13774 + }, + { + "epoch": 0.13775, + "grad_norm": 1.1216764323554513, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 13775 + }, + { + "epoch": 0.13776, + "grad_norm": 1.393869507420362, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 13776 + }, + { + "epoch": 0.13777, + "grad_norm": 1.2412187589578358, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 13777 + }, + { + "epoch": 0.13778, + "grad_norm": 1.3400166349526852, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 13778 + }, + { + "epoch": 0.13779, + "grad_norm": 1.1264589087670016, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 13779 + }, + { + "epoch": 0.1378, + "grad_norm": 1.4068617953668512, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 13780 + }, + { + "epoch": 0.13781, + "grad_norm": 1.163379495883538, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 13781 + }, + { + "epoch": 0.13782, + "grad_norm": 1.0635912857489516, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 13782 + }, + { + "epoch": 0.13783, + "grad_norm": 1.295939159374199, + "learning_rate": 0.003, + "loss": 4.059, + "step": 13783 + }, + { + "epoch": 0.13784, + "grad_norm": 1.0123106600447564, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 13784 + }, + { + "epoch": 0.13785, + "grad_norm": 1.1793198715769542, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 13785 + }, + { + "epoch": 0.13786, + "grad_norm": 1.3681108391101406, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 13786 + }, + { + "epoch": 0.13787, + "grad_norm": 1.0806468901634552, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 13787 + }, + { + "epoch": 0.13788, + "grad_norm": 1.4124903568965343, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 13788 + }, + { + "epoch": 0.13789, + "grad_norm": 0.9941383912137073, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 13789 + }, + { + "epoch": 0.1379, + "grad_norm": 1.4690114949842519, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 13790 + }, + { + "epoch": 0.13791, + "grad_norm": 0.9399050464965364, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 13791 + }, + { + "epoch": 0.13792, + "grad_norm": 1.0929317392269118, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 13792 + }, + { + "epoch": 0.13793, + "grad_norm": 1.1524550818415649, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 13793 + }, + { + "epoch": 0.13794, + "grad_norm": 1.4067300730540004, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 13794 + }, + { + "epoch": 0.13795, + "grad_norm": 0.8852232954524153, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 13795 + }, + { + "epoch": 0.13796, + "grad_norm": 1.34857853723551, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 13796 + }, + { + "epoch": 0.13797, + "grad_norm": 1.417437703995536, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 13797 + }, + { + "epoch": 0.13798, + "grad_norm": 1.2875740904930375, + "learning_rate": 0.003, + "loss": 4.019, + "step": 13798 + }, + { + "epoch": 0.13799, + "grad_norm": 1.1346468457962757, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 13799 + }, + { + "epoch": 0.138, + "grad_norm": 1.1202318261674071, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 13800 + }, + { + "epoch": 0.13801, + "grad_norm": 1.2250164759666067, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 13801 + }, + { + "epoch": 0.13802, + "grad_norm": 1.273665428202938, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 13802 + }, + { + "epoch": 0.13803, + "grad_norm": 1.0505705506231746, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 13803 + }, + { + "epoch": 0.13804, + "grad_norm": 1.396502131917135, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 13804 + }, + { + "epoch": 0.13805, + "grad_norm": 1.040820881985539, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 13805 + }, + { + "epoch": 0.13806, + "grad_norm": 1.3201950992023033, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 13806 + }, + { + "epoch": 0.13807, + "grad_norm": 1.0861147573091232, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13807 + }, + { + "epoch": 0.13808, + "grad_norm": 1.2454784265931447, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 13808 + }, + { + "epoch": 0.13809, + "grad_norm": 1.0460082186264419, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 13809 + }, + { + "epoch": 0.1381, + "grad_norm": 1.4013112005182495, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 13810 + }, + { + "epoch": 0.13811, + "grad_norm": 1.4702898465350844, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 13811 + }, + { + "epoch": 0.13812, + "grad_norm": 1.1119162188653258, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 13812 + }, + { + "epoch": 0.13813, + "grad_norm": 1.3278272093657792, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13813 + }, + { + "epoch": 0.13814, + "grad_norm": 1.2921164651858392, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 13814 + }, + { + "epoch": 0.13815, + "grad_norm": 1.070492395176577, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 13815 + }, + { + "epoch": 0.13816, + "grad_norm": 1.1857974334956447, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 13816 + }, + { + "epoch": 0.13817, + "grad_norm": 1.2491079518196415, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 13817 + }, + { + "epoch": 0.13818, + "grad_norm": 1.0503690419199725, + "learning_rate": 0.003, + "loss": 3.995, + "step": 13818 + }, + { + "epoch": 0.13819, + "grad_norm": 1.213579214314059, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 13819 + }, + { + "epoch": 0.1382, + "grad_norm": 1.2256209285990405, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 13820 + }, + { + "epoch": 0.13821, + "grad_norm": 1.1823879754528905, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 13821 + }, + { + "epoch": 0.13822, + "grad_norm": 1.1448906824952305, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 13822 + }, + { + "epoch": 0.13823, + "grad_norm": 1.472628687368256, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 13823 + }, + { + "epoch": 0.13824, + "grad_norm": 1.0233371215440514, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 13824 + }, + { + "epoch": 0.13825, + "grad_norm": 1.4748736420222301, + "learning_rate": 0.003, + "loss": 4.011, + "step": 13825 + }, + { + "epoch": 0.13826, + "grad_norm": 1.134890953235084, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 13826 + }, + { + "epoch": 0.13827, + "grad_norm": 1.1542926288725004, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 13827 + }, + { + "epoch": 0.13828, + "grad_norm": 1.0772783402004467, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 13828 + }, + { + "epoch": 0.13829, + "grad_norm": 1.398030801311642, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 13829 + }, + { + "epoch": 0.1383, + "grad_norm": 1.3276117884935643, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13830 + }, + { + "epoch": 0.13831, + "grad_norm": 0.9571093429384878, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 13831 + }, + { + "epoch": 0.13832, + "grad_norm": 1.4892711039285325, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 13832 + }, + { + "epoch": 0.13833, + "grad_norm": 1.150716669767256, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 13833 + }, + { + "epoch": 0.13834, + "grad_norm": 1.0530121395710959, + "learning_rate": 0.003, + "loss": 4.017, + "step": 13834 + }, + { + "epoch": 0.13835, + "grad_norm": 1.1494600804013582, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 13835 + }, + { + "epoch": 0.13836, + "grad_norm": 1.2083708079434357, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 13836 + }, + { + "epoch": 0.13837, + "grad_norm": 1.1715556568975776, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 13837 + }, + { + "epoch": 0.13838, + "grad_norm": 1.0801918163617674, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 13838 + }, + { + "epoch": 0.13839, + "grad_norm": 1.1347098154169728, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 13839 + }, + { + "epoch": 0.1384, + "grad_norm": 1.3958779067058595, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 13840 + }, + { + "epoch": 0.13841, + "grad_norm": 1.006181864114487, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 13841 + }, + { + "epoch": 0.13842, + "grad_norm": 1.272894091173234, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13842 + }, + { + "epoch": 0.13843, + "grad_norm": 1.2013856009795578, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 13843 + }, + { + "epoch": 0.13844, + "grad_norm": 1.1246714118232903, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 13844 + }, + { + "epoch": 0.13845, + "grad_norm": 1.322509334891701, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 13845 + }, + { + "epoch": 0.13846, + "grad_norm": 1.1521055168429188, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 13846 + }, + { + "epoch": 0.13847, + "grad_norm": 1.2478833693518745, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 13847 + }, + { + "epoch": 0.13848, + "grad_norm": 1.3016325633154966, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13848 + }, + { + "epoch": 0.13849, + "grad_norm": 1.2280057760265102, + "learning_rate": 0.003, + "loss": 4.017, + "step": 13849 + }, + { + "epoch": 0.1385, + "grad_norm": 1.0772662000472513, + "learning_rate": 0.003, + "loss": 4.013, + "step": 13850 + }, + { + "epoch": 0.13851, + "grad_norm": 1.3768658155376017, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 13851 + }, + { + "epoch": 0.13852, + "grad_norm": 1.010429514973437, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13852 + }, + { + "epoch": 0.13853, + "grad_norm": 1.3832349644979869, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 13853 + }, + { + "epoch": 0.13854, + "grad_norm": 1.0641415255798539, + "learning_rate": 0.003, + "loss": 4.027, + "step": 13854 + }, + { + "epoch": 0.13855, + "grad_norm": 1.4794813909654045, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 13855 + }, + { + "epoch": 0.13856, + "grad_norm": 1.1183826799442813, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 13856 + }, + { + "epoch": 0.13857, + "grad_norm": 1.23748079145574, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 13857 + }, + { + "epoch": 0.13858, + "grad_norm": 1.114360873205981, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 13858 + }, + { + "epoch": 0.13859, + "grad_norm": 1.2406829264267383, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 13859 + }, + { + "epoch": 0.1386, + "grad_norm": 1.1466333335232666, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 13860 + }, + { + "epoch": 0.13861, + "grad_norm": 1.301601988392719, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13861 + }, + { + "epoch": 0.13862, + "grad_norm": 1.0949167558716029, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 13862 + }, + { + "epoch": 0.13863, + "grad_norm": 0.9560402898539208, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 13863 + }, + { + "epoch": 0.13864, + "grad_norm": 1.2825507857461917, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 13864 + }, + { + "epoch": 0.13865, + "grad_norm": 1.178410218571524, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 13865 + }, + { + "epoch": 0.13866, + "grad_norm": 1.174055726164993, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 13866 + }, + { + "epoch": 0.13867, + "grad_norm": 1.4536410253305663, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 13867 + }, + { + "epoch": 0.13868, + "grad_norm": 0.9062474618794782, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 13868 + }, + { + "epoch": 0.13869, + "grad_norm": 1.0994109586504763, + "learning_rate": 0.003, + "loss": 4.0, + "step": 13869 + }, + { + "epoch": 0.1387, + "grad_norm": 1.1734072957099282, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 13870 + }, + { + "epoch": 0.13871, + "grad_norm": 1.3104684470724604, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 13871 + }, + { + "epoch": 0.13872, + "grad_norm": 1.0379320482503833, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 13872 + }, + { + "epoch": 0.13873, + "grad_norm": 1.293243056137157, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 13873 + }, + { + "epoch": 0.13874, + "grad_norm": 1.0469175605557832, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 13874 + }, + { + "epoch": 0.13875, + "grad_norm": 1.0299822183811806, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 13875 + }, + { + "epoch": 0.13876, + "grad_norm": 1.2130323863458037, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 13876 + }, + { + "epoch": 0.13877, + "grad_norm": 1.1859891666047682, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 13877 + }, + { + "epoch": 0.13878, + "grad_norm": 1.2489914141355798, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 13878 + }, + { + "epoch": 0.13879, + "grad_norm": 1.243883938349891, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 13879 + }, + { + "epoch": 0.1388, + "grad_norm": 1.2098306819415796, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 13880 + }, + { + "epoch": 0.13881, + "grad_norm": 1.1157535130031828, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 13881 + }, + { + "epoch": 0.13882, + "grad_norm": 1.309752718257075, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 13882 + }, + { + "epoch": 0.13883, + "grad_norm": 1.1834101954134257, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 13883 + }, + { + "epoch": 0.13884, + "grad_norm": 1.3622147587914692, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 13884 + }, + { + "epoch": 0.13885, + "grad_norm": 1.0817660002621556, + "learning_rate": 0.003, + "loss": 4.055, + "step": 13885 + }, + { + "epoch": 0.13886, + "grad_norm": 1.220724330378295, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 13886 + }, + { + "epoch": 0.13887, + "grad_norm": 1.0857253756268257, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 13887 + }, + { + "epoch": 0.13888, + "grad_norm": 1.2881188044988716, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 13888 + }, + { + "epoch": 0.13889, + "grad_norm": 1.3277501652331098, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 13889 + }, + { + "epoch": 0.1389, + "grad_norm": 1.2847936050254294, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 13890 + }, + { + "epoch": 0.13891, + "grad_norm": 1.10509747370125, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 13891 + }, + { + "epoch": 0.13892, + "grad_norm": 1.3355982750193622, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 13892 + }, + { + "epoch": 0.13893, + "grad_norm": 1.199772978665531, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 13893 + }, + { + "epoch": 0.13894, + "grad_norm": 1.1919544025083375, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 13894 + }, + { + "epoch": 0.13895, + "grad_norm": 1.1791968684172016, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 13895 + }, + { + "epoch": 0.13896, + "grad_norm": 1.045813585726014, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 13896 + }, + { + "epoch": 0.13897, + "grad_norm": 1.464421185361696, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 13897 + }, + { + "epoch": 0.13898, + "grad_norm": 0.9830880963597627, + "learning_rate": 0.003, + "loss": 3.992, + "step": 13898 + }, + { + "epoch": 0.13899, + "grad_norm": 1.4356757886313305, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 13899 + }, + { + "epoch": 0.139, + "grad_norm": 1.3762264672777118, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 13900 + }, + { + "epoch": 0.13901, + "grad_norm": 1.4262850064756332, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 13901 + }, + { + "epoch": 0.13902, + "grad_norm": 1.1067642729379954, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13902 + }, + { + "epoch": 0.13903, + "grad_norm": 1.2662437461045917, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 13903 + }, + { + "epoch": 0.13904, + "grad_norm": 1.0311275415899397, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 13904 + }, + { + "epoch": 0.13905, + "grad_norm": 1.1936049630810572, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 13905 + }, + { + "epoch": 0.13906, + "grad_norm": 1.052760904049721, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 13906 + }, + { + "epoch": 0.13907, + "grad_norm": 1.3029101585399128, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 13907 + }, + { + "epoch": 0.13908, + "grad_norm": 1.2335774415370064, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 13908 + }, + { + "epoch": 0.13909, + "grad_norm": 1.3275677632729739, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 13909 + }, + { + "epoch": 0.1391, + "grad_norm": 1.1297632688079993, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 13910 + }, + { + "epoch": 0.13911, + "grad_norm": 1.2722418665779391, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 13911 + }, + { + "epoch": 0.13912, + "grad_norm": 1.0805464020930946, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 13912 + }, + { + "epoch": 0.13913, + "grad_norm": 1.340313135441908, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 13913 + }, + { + "epoch": 0.13914, + "grad_norm": 0.9922681051920051, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 13914 + }, + { + "epoch": 0.13915, + "grad_norm": 1.3705530383268818, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 13915 + }, + { + "epoch": 0.13916, + "grad_norm": 0.9267537104518053, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 13916 + }, + { + "epoch": 0.13917, + "grad_norm": 1.1780767482538883, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 13917 + }, + { + "epoch": 0.13918, + "grad_norm": 1.294385111083149, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 13918 + }, + { + "epoch": 0.13919, + "grad_norm": 0.9510832470643007, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 13919 + }, + { + "epoch": 0.1392, + "grad_norm": 1.1967730551592841, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 13920 + }, + { + "epoch": 0.13921, + "grad_norm": 1.241002135891913, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 13921 + }, + { + "epoch": 0.13922, + "grad_norm": 1.0521373234826998, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 13922 + }, + { + "epoch": 0.13923, + "grad_norm": 1.3016596425010858, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13923 + }, + { + "epoch": 0.13924, + "grad_norm": 1.3241790790815522, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 13924 + }, + { + "epoch": 0.13925, + "grad_norm": 1.2397807904648384, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 13925 + }, + { + "epoch": 0.13926, + "grad_norm": 1.5988946043532852, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 13926 + }, + { + "epoch": 0.13927, + "grad_norm": 1.043076253026414, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 13927 + }, + { + "epoch": 0.13928, + "grad_norm": 1.3983772820072078, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 13928 + }, + { + "epoch": 0.13929, + "grad_norm": 1.3822214057025675, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 13929 + }, + { + "epoch": 0.1393, + "grad_norm": 1.0932817291788222, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 13930 + }, + { + "epoch": 0.13931, + "grad_norm": 1.2862918829184873, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 13931 + }, + { + "epoch": 0.13932, + "grad_norm": 1.2821414397282036, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 13932 + }, + { + "epoch": 0.13933, + "grad_norm": 1.2089220976008554, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 13933 + }, + { + "epoch": 0.13934, + "grad_norm": 1.2285477929917308, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 13934 + }, + { + "epoch": 0.13935, + "grad_norm": 1.0331335294069954, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13935 + }, + { + "epoch": 0.13936, + "grad_norm": 1.228393740053751, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 13936 + }, + { + "epoch": 0.13937, + "grad_norm": 1.2194970798916072, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 13937 + }, + { + "epoch": 0.13938, + "grad_norm": 1.1371455385108602, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 13938 + }, + { + "epoch": 0.13939, + "grad_norm": 1.1464202944234065, + "learning_rate": 0.003, + "loss": 4.026, + "step": 13939 + }, + { + "epoch": 0.1394, + "grad_norm": 1.209782506620369, + "learning_rate": 0.003, + "loss": 3.997, + "step": 13940 + }, + { + "epoch": 0.13941, + "grad_norm": 1.2437706017362091, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 13941 + }, + { + "epoch": 0.13942, + "grad_norm": 0.9606867500748588, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 13942 + }, + { + "epoch": 0.13943, + "grad_norm": 1.330782043238364, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 13943 + }, + { + "epoch": 0.13944, + "grad_norm": 1.0944235490095693, + "learning_rate": 0.003, + "loss": 4.033, + "step": 13944 + }, + { + "epoch": 0.13945, + "grad_norm": 1.3837713427805216, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 13945 + }, + { + "epoch": 0.13946, + "grad_norm": 1.1283957430188176, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 13946 + }, + { + "epoch": 0.13947, + "grad_norm": 1.3073033270119225, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 13947 + }, + { + "epoch": 0.13948, + "grad_norm": 1.0356700644168169, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 13948 + }, + { + "epoch": 0.13949, + "grad_norm": 1.2827850931174634, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 13949 + }, + { + "epoch": 0.1395, + "grad_norm": 0.9681245671456082, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 13950 + }, + { + "epoch": 0.13951, + "grad_norm": 1.6241769253303424, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 13951 + }, + { + "epoch": 0.13952, + "grad_norm": 1.0930959372215059, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13952 + }, + { + "epoch": 0.13953, + "grad_norm": 1.2948262234227748, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 13953 + }, + { + "epoch": 0.13954, + "grad_norm": 1.0006578557828247, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 13954 + }, + { + "epoch": 0.13955, + "grad_norm": 1.4064078111835758, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 13955 + }, + { + "epoch": 0.13956, + "grad_norm": 1.086412035581726, + "learning_rate": 0.003, + "loss": 4.03, + "step": 13956 + }, + { + "epoch": 0.13957, + "grad_norm": 1.4309866455033866, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 13957 + }, + { + "epoch": 0.13958, + "grad_norm": 0.9741802566465895, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 13958 + }, + { + "epoch": 0.13959, + "grad_norm": 1.403639528165463, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 13959 + }, + { + "epoch": 0.1396, + "grad_norm": 1.2468214735098933, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 13960 + }, + { + "epoch": 0.13961, + "grad_norm": 1.2613812556591117, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 13961 + }, + { + "epoch": 0.13962, + "grad_norm": 1.1023642989666198, + "learning_rate": 0.003, + "loss": 4.028, + "step": 13962 + }, + { + "epoch": 0.13963, + "grad_norm": 1.2135199740752762, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 13963 + }, + { + "epoch": 0.13964, + "grad_norm": 1.0943237528793914, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 13964 + }, + { + "epoch": 0.13965, + "grad_norm": 1.3984602769716774, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 13965 + }, + { + "epoch": 0.13966, + "grad_norm": 1.1833554431013096, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 13966 + }, + { + "epoch": 0.13967, + "grad_norm": 1.311230921590294, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 13967 + }, + { + "epoch": 0.13968, + "grad_norm": 0.997931625907764, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 13968 + }, + { + "epoch": 0.13969, + "grad_norm": 1.6613948429954382, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 13969 + }, + { + "epoch": 0.1397, + "grad_norm": 1.1706001548186498, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 13970 + }, + { + "epoch": 0.13971, + "grad_norm": 1.3455729323827497, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 13971 + }, + { + "epoch": 0.13972, + "grad_norm": 1.0071348841464436, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 13972 + }, + { + "epoch": 0.13973, + "grad_norm": 1.382779050748766, + "learning_rate": 0.003, + "loss": 4.049, + "step": 13973 + }, + { + "epoch": 0.13974, + "grad_norm": 1.22731538379118, + "learning_rate": 0.003, + "loss": 4.034, + "step": 13974 + }, + { + "epoch": 0.13975, + "grad_norm": 1.381444376592699, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 13975 + }, + { + "epoch": 0.13976, + "grad_norm": 1.334391128566007, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 13976 + }, + { + "epoch": 0.13977, + "grad_norm": 1.112083982151318, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 13977 + }, + { + "epoch": 0.13978, + "grad_norm": 1.0883189299965499, + "learning_rate": 0.003, + "loss": 4.076, + "step": 13978 + }, + { + "epoch": 0.13979, + "grad_norm": 1.3200277147595922, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 13979 + }, + { + "epoch": 0.1398, + "grad_norm": 1.3044645984561822, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 13980 + }, + { + "epoch": 0.13981, + "grad_norm": 1.168169695750215, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 13981 + }, + { + "epoch": 0.13982, + "grad_norm": 1.3016765422844125, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 13982 + }, + { + "epoch": 0.13983, + "grad_norm": 1.1304122952759246, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 13983 + }, + { + "epoch": 0.13984, + "grad_norm": 1.0634672602319772, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 13984 + }, + { + "epoch": 0.13985, + "grad_norm": 1.1666668593758922, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 13985 + }, + { + "epoch": 0.13986, + "grad_norm": 1.1577098649486288, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 13986 + }, + { + "epoch": 0.13987, + "grad_norm": 1.2543916406092464, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 13987 + }, + { + "epoch": 0.13988, + "grad_norm": 1.2350106643993233, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 13988 + }, + { + "epoch": 0.13989, + "grad_norm": 1.142275915177779, + "learning_rate": 0.003, + "loss": 4.009, + "step": 13989 + }, + { + "epoch": 0.1399, + "grad_norm": 1.2704798136615096, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 13990 + }, + { + "epoch": 0.13991, + "grad_norm": 0.9015601618647462, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 13991 + }, + { + "epoch": 0.13992, + "grad_norm": 0.9856059615225615, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 13992 + }, + { + "epoch": 0.13993, + "grad_norm": 1.465937483460713, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 13993 + }, + { + "epoch": 0.13994, + "grad_norm": 1.2082285117015399, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 13994 + }, + { + "epoch": 0.13995, + "grad_norm": 1.4099211609051017, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 13995 + }, + { + "epoch": 0.13996, + "grad_norm": 1.1942828382556108, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13996 + }, + { + "epoch": 0.13997, + "grad_norm": 1.0453877453839826, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13997 + }, + { + "epoch": 0.13998, + "grad_norm": 1.3971265611083508, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 13998 + }, + { + "epoch": 0.13999, + "grad_norm": 1.0312978267883302, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 13999 + }, + { + "epoch": 0.14, + "grad_norm": 1.3522912522542927, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 14000 + }, + { + "epoch": 0.14001, + "grad_norm": 1.2606310031280883, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 14001 + }, + { + "epoch": 0.14002, + "grad_norm": 1.1408868714728893, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 14002 + }, + { + "epoch": 0.14003, + "grad_norm": 1.2117998216576167, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 14003 + }, + { + "epoch": 0.14004, + "grad_norm": 1.1914649510372295, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 14004 + }, + { + "epoch": 0.14005, + "grad_norm": 1.13158563813848, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14005 + }, + { + "epoch": 0.14006, + "grad_norm": 1.1789590777846468, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 14006 + }, + { + "epoch": 0.14007, + "grad_norm": 1.2482102753530098, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 14007 + }, + { + "epoch": 0.14008, + "grad_norm": 1.3000865301090239, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 14008 + }, + { + "epoch": 0.14009, + "grad_norm": 1.0317882634647473, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 14009 + }, + { + "epoch": 0.1401, + "grad_norm": 1.2888815577614365, + "learning_rate": 0.003, + "loss": 4.031, + "step": 14010 + }, + { + "epoch": 0.14011, + "grad_norm": 0.9507500621278664, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 14011 + }, + { + "epoch": 0.14012, + "grad_norm": 1.410728391771078, + "learning_rate": 0.003, + "loss": 4.026, + "step": 14012 + }, + { + "epoch": 0.14013, + "grad_norm": 1.223325656820055, + "learning_rate": 0.003, + "loss": 4.011, + "step": 14013 + }, + { + "epoch": 0.14014, + "grad_norm": 1.110814995418571, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 14014 + }, + { + "epoch": 0.14015, + "grad_norm": 1.0921648099988692, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 14015 + }, + { + "epoch": 0.14016, + "grad_norm": 1.3727424125567884, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 14016 + }, + { + "epoch": 0.14017, + "grad_norm": 1.2203621215869773, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 14017 + }, + { + "epoch": 0.14018, + "grad_norm": 1.3990427406197736, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 14018 + }, + { + "epoch": 0.14019, + "grad_norm": 1.2834847469903492, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 14019 + }, + { + "epoch": 0.1402, + "grad_norm": 1.2203999381013084, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 14020 + }, + { + "epoch": 0.14021, + "grad_norm": 1.1537817679896496, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 14021 + }, + { + "epoch": 0.14022, + "grad_norm": 1.1840094298462875, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 14022 + }, + { + "epoch": 0.14023, + "grad_norm": 1.2514685746782503, + "learning_rate": 0.003, + "loss": 4.008, + "step": 14023 + }, + { + "epoch": 0.14024, + "grad_norm": 1.2572710511999712, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 14024 + }, + { + "epoch": 0.14025, + "grad_norm": 1.1529838978875182, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 14025 + }, + { + "epoch": 0.14026, + "grad_norm": 1.1196849447629258, + "learning_rate": 0.003, + "loss": 4.023, + "step": 14026 + }, + { + "epoch": 0.14027, + "grad_norm": 1.4120002031767034, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 14027 + }, + { + "epoch": 0.14028, + "grad_norm": 1.2252743403550672, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 14028 + }, + { + "epoch": 0.14029, + "grad_norm": 1.0560936086888912, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 14029 + }, + { + "epoch": 0.1403, + "grad_norm": 1.1071915022634027, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 14030 + }, + { + "epoch": 0.14031, + "grad_norm": 1.1260075699825944, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 14031 + }, + { + "epoch": 0.14032, + "grad_norm": 1.2044444637907508, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14032 + }, + { + "epoch": 0.14033, + "grad_norm": 1.0727155556561392, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 14033 + }, + { + "epoch": 0.14034, + "grad_norm": 1.32512443537798, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 14034 + }, + { + "epoch": 0.14035, + "grad_norm": 0.9462166052486559, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 14035 + }, + { + "epoch": 0.14036, + "grad_norm": 1.4978733792333327, + "learning_rate": 0.003, + "loss": 4.013, + "step": 14036 + }, + { + "epoch": 0.14037, + "grad_norm": 1.2740393697478154, + "learning_rate": 0.003, + "loss": 4.054, + "step": 14037 + }, + { + "epoch": 0.14038, + "grad_norm": 1.212192612661939, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 14038 + }, + { + "epoch": 0.14039, + "grad_norm": 1.2651133818216853, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 14039 + }, + { + "epoch": 0.1404, + "grad_norm": 1.159014019843782, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 14040 + }, + { + "epoch": 0.14041, + "grad_norm": 1.356334413419729, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 14041 + }, + { + "epoch": 0.14042, + "grad_norm": 1.1790065129947735, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 14042 + }, + { + "epoch": 0.14043, + "grad_norm": 1.3024594270807575, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 14043 + }, + { + "epoch": 0.14044, + "grad_norm": 1.1063212035382364, + "learning_rate": 0.003, + "loss": 3.985, + "step": 14044 + }, + { + "epoch": 0.14045, + "grad_norm": 1.358797191328306, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 14045 + }, + { + "epoch": 0.14046, + "grad_norm": 1.0003323125137746, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 14046 + }, + { + "epoch": 0.14047, + "grad_norm": 1.3248096853707327, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 14047 + }, + { + "epoch": 0.14048, + "grad_norm": 1.1217841806694424, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 14048 + }, + { + "epoch": 0.14049, + "grad_norm": 1.233918077507105, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 14049 + }, + { + "epoch": 0.1405, + "grad_norm": 1.1216017192561671, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 14050 + }, + { + "epoch": 0.14051, + "grad_norm": 1.108493840663453, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 14051 + }, + { + "epoch": 0.14052, + "grad_norm": 1.1763668314782794, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 14052 + }, + { + "epoch": 0.14053, + "grad_norm": 1.2001181755948613, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 14053 + }, + { + "epoch": 0.14054, + "grad_norm": 1.1838910215006067, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14054 + }, + { + "epoch": 0.14055, + "grad_norm": 1.3650656646670034, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 14055 + }, + { + "epoch": 0.14056, + "grad_norm": 1.236392593269088, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 14056 + }, + { + "epoch": 0.14057, + "grad_norm": 1.3761101064022911, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 14057 + }, + { + "epoch": 0.14058, + "grad_norm": 0.8473911571718556, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 14058 + }, + { + "epoch": 0.14059, + "grad_norm": 1.1611075233336143, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 14059 + }, + { + "epoch": 0.1406, + "grad_norm": 1.1457197339194993, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 14060 + }, + { + "epoch": 0.14061, + "grad_norm": 1.1708080471302844, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 14061 + }, + { + "epoch": 0.14062, + "grad_norm": 1.264785307429114, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 14062 + }, + { + "epoch": 0.14063, + "grad_norm": 1.4168821067309096, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14063 + }, + { + "epoch": 0.14064, + "grad_norm": 1.2688748501802554, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 14064 + }, + { + "epoch": 0.14065, + "grad_norm": 1.27489076739789, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 14065 + }, + { + "epoch": 0.14066, + "grad_norm": 1.2205632502308121, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 14066 + }, + { + "epoch": 0.14067, + "grad_norm": 1.0459763075762, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 14067 + }, + { + "epoch": 0.14068, + "grad_norm": 1.1345110724118248, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 14068 + }, + { + "epoch": 0.14069, + "grad_norm": 1.1143860163445822, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 14069 + }, + { + "epoch": 0.1407, + "grad_norm": 1.5410578205758152, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 14070 + }, + { + "epoch": 0.14071, + "grad_norm": 1.0829045851135513, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 14071 + }, + { + "epoch": 0.14072, + "grad_norm": 1.4511526715068985, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 14072 + }, + { + "epoch": 0.14073, + "grad_norm": 1.099805202133277, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14073 + }, + { + "epoch": 0.14074, + "grad_norm": 1.328312415775265, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 14074 + }, + { + "epoch": 0.14075, + "grad_norm": 1.3574759342501346, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 14075 + }, + { + "epoch": 0.14076, + "grad_norm": 1.2040648367123625, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14076 + }, + { + "epoch": 0.14077, + "grad_norm": 1.2356253648930262, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 14077 + }, + { + "epoch": 0.14078, + "grad_norm": 1.055788559715171, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14078 + }, + { + "epoch": 0.14079, + "grad_norm": 1.5561946686609198, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 14079 + }, + { + "epoch": 0.1408, + "grad_norm": 0.8650739053143781, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 14080 + }, + { + "epoch": 0.14081, + "grad_norm": 1.0303599602027143, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 14081 + }, + { + "epoch": 0.14082, + "grad_norm": 1.267751441529696, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 14082 + }, + { + "epoch": 0.14083, + "grad_norm": 0.9239122241956662, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 14083 + }, + { + "epoch": 0.14084, + "grad_norm": 1.0079857265874606, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 14084 + }, + { + "epoch": 0.14085, + "grad_norm": 1.4466701881203794, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 14085 + }, + { + "epoch": 0.14086, + "grad_norm": 1.1307165327501238, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 14086 + }, + { + "epoch": 0.14087, + "grad_norm": 1.2546536674308166, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 14087 + }, + { + "epoch": 0.14088, + "grad_norm": 1.1903476498333216, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 14088 + }, + { + "epoch": 0.14089, + "grad_norm": 1.3578091755598753, + "learning_rate": 0.003, + "loss": 4.018, + "step": 14089 + }, + { + "epoch": 0.1409, + "grad_norm": 1.268021700510964, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 14090 + }, + { + "epoch": 0.14091, + "grad_norm": 1.2047706355464072, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 14091 + }, + { + "epoch": 0.14092, + "grad_norm": 1.083127062472936, + "learning_rate": 0.003, + "loss": 4.032, + "step": 14092 + }, + { + "epoch": 0.14093, + "grad_norm": 1.418971172036768, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 14093 + }, + { + "epoch": 0.14094, + "grad_norm": 1.1484741827481468, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 14094 + }, + { + "epoch": 0.14095, + "grad_norm": 1.1705700080017016, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 14095 + }, + { + "epoch": 0.14096, + "grad_norm": 1.5220333740977865, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 14096 + }, + { + "epoch": 0.14097, + "grad_norm": 1.1069309793723154, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 14097 + }, + { + "epoch": 0.14098, + "grad_norm": 1.4591901998283383, + "learning_rate": 0.003, + "loss": 4.036, + "step": 14098 + }, + { + "epoch": 0.14099, + "grad_norm": 1.0732646509735886, + "learning_rate": 0.003, + "loss": 4.037, + "step": 14099 + }, + { + "epoch": 0.141, + "grad_norm": 1.2309199191771107, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 14100 + }, + { + "epoch": 0.14101, + "grad_norm": 1.1137438686043557, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 14101 + }, + { + "epoch": 0.14102, + "grad_norm": 1.3116072735141207, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 14102 + }, + { + "epoch": 0.14103, + "grad_norm": 1.3072196104614588, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 14103 + }, + { + "epoch": 0.14104, + "grad_norm": 1.0177560091617586, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 14104 + }, + { + "epoch": 0.14105, + "grad_norm": 1.3464912000778624, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 14105 + }, + { + "epoch": 0.14106, + "grad_norm": 1.0337998967026791, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 14106 + }, + { + "epoch": 0.14107, + "grad_norm": 1.5122607325580328, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 14107 + }, + { + "epoch": 0.14108, + "grad_norm": 1.2956936810229631, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 14108 + }, + { + "epoch": 0.14109, + "grad_norm": 1.3139702931394983, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 14109 + }, + { + "epoch": 0.1411, + "grad_norm": 1.0048453641489445, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 14110 + }, + { + "epoch": 0.14111, + "grad_norm": 1.430901945492829, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 14111 + }, + { + "epoch": 0.14112, + "grad_norm": 1.0367673762169347, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 14112 + }, + { + "epoch": 0.14113, + "grad_norm": 1.2892316542366529, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 14113 + }, + { + "epoch": 0.14114, + "grad_norm": 1.1071539287465246, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 14114 + }, + { + "epoch": 0.14115, + "grad_norm": 1.314574543696847, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 14115 + }, + { + "epoch": 0.14116, + "grad_norm": 1.0897407150869276, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 14116 + }, + { + "epoch": 0.14117, + "grad_norm": 1.1950700957763445, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14117 + }, + { + "epoch": 0.14118, + "grad_norm": 1.1412532223785143, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 14118 + }, + { + "epoch": 0.14119, + "grad_norm": 1.444097546450158, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 14119 + }, + { + "epoch": 0.1412, + "grad_norm": 1.1772444722146265, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 14120 + }, + { + "epoch": 0.14121, + "grad_norm": 1.2618437643898595, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 14121 + }, + { + "epoch": 0.14122, + "grad_norm": 1.0852284065931241, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 14122 + }, + { + "epoch": 0.14123, + "grad_norm": 1.2536516985760147, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 14123 + }, + { + "epoch": 0.14124, + "grad_norm": 1.1957565476061842, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 14124 + }, + { + "epoch": 0.14125, + "grad_norm": 1.266299712143771, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 14125 + }, + { + "epoch": 0.14126, + "grad_norm": 1.1179825213214207, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 14126 + }, + { + "epoch": 0.14127, + "grad_norm": 1.4270749872421038, + "learning_rate": 0.003, + "loss": 4.024, + "step": 14127 + }, + { + "epoch": 0.14128, + "grad_norm": 1.0854846432711502, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 14128 + }, + { + "epoch": 0.14129, + "grad_norm": 1.2293105496627963, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 14129 + }, + { + "epoch": 0.1413, + "grad_norm": 1.1580430778715787, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 14130 + }, + { + "epoch": 0.14131, + "grad_norm": 1.185949013161452, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 14131 + }, + { + "epoch": 0.14132, + "grad_norm": 1.1314620909176543, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 14132 + }, + { + "epoch": 0.14133, + "grad_norm": 1.2079450778243612, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 14133 + }, + { + "epoch": 0.14134, + "grad_norm": 0.9944973687931271, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 14134 + }, + { + "epoch": 0.14135, + "grad_norm": 1.3767494365485264, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 14135 + }, + { + "epoch": 0.14136, + "grad_norm": 1.0695325114096956, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 14136 + }, + { + "epoch": 0.14137, + "grad_norm": 1.4026688622739356, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 14137 + }, + { + "epoch": 0.14138, + "grad_norm": 1.1898668414843263, + "learning_rate": 0.003, + "loss": 4.021, + "step": 14138 + }, + { + "epoch": 0.14139, + "grad_norm": 1.144411276936079, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 14139 + }, + { + "epoch": 0.1414, + "grad_norm": 1.2266112224678802, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 14140 + }, + { + "epoch": 0.14141, + "grad_norm": 1.1860770803289438, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 14141 + }, + { + "epoch": 0.14142, + "grad_norm": 1.1660925207735835, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 14142 + }, + { + "epoch": 0.14143, + "grad_norm": 1.3955618961272698, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 14143 + }, + { + "epoch": 0.14144, + "grad_norm": 1.1286725867294176, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 14144 + }, + { + "epoch": 0.14145, + "grad_norm": 1.3666449864554138, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 14145 + }, + { + "epoch": 0.14146, + "grad_norm": 1.0899359311217849, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 14146 + }, + { + "epoch": 0.14147, + "grad_norm": 1.3076162503734543, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 14147 + }, + { + "epoch": 0.14148, + "grad_norm": 0.9456838894699942, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 14148 + }, + { + "epoch": 0.14149, + "grad_norm": 1.1458933750085323, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 14149 + }, + { + "epoch": 0.1415, + "grad_norm": 1.3026803164766048, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 14150 + }, + { + "epoch": 0.14151, + "grad_norm": 1.1936052509529032, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14151 + }, + { + "epoch": 0.14152, + "grad_norm": 1.53961674491896, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14152 + }, + { + "epoch": 0.14153, + "grad_norm": 1.0355640859393733, + "learning_rate": 0.003, + "loss": 3.987, + "step": 14153 + }, + { + "epoch": 0.14154, + "grad_norm": 1.2626192828203087, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 14154 + }, + { + "epoch": 0.14155, + "grad_norm": 1.364612527361059, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 14155 + }, + { + "epoch": 0.14156, + "grad_norm": 0.9904204490104055, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 14156 + }, + { + "epoch": 0.14157, + "grad_norm": 1.4606046724678607, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 14157 + }, + { + "epoch": 0.14158, + "grad_norm": 1.1412628005227667, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 14158 + }, + { + "epoch": 0.14159, + "grad_norm": 1.2566170966440469, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 14159 + }, + { + "epoch": 0.1416, + "grad_norm": 1.1942427192863938, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 14160 + }, + { + "epoch": 0.14161, + "grad_norm": 1.2747119258046609, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 14161 + }, + { + "epoch": 0.14162, + "grad_norm": 1.0707115414199597, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 14162 + }, + { + "epoch": 0.14163, + "grad_norm": 1.0897594650738252, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 14163 + }, + { + "epoch": 0.14164, + "grad_norm": 1.178469957198732, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 14164 + }, + { + "epoch": 0.14165, + "grad_norm": 1.0783093439173759, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 14165 + }, + { + "epoch": 0.14166, + "grad_norm": 1.3314651178177561, + "learning_rate": 0.003, + "loss": 4.04, + "step": 14166 + }, + { + "epoch": 0.14167, + "grad_norm": 1.0843718324605727, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 14167 + }, + { + "epoch": 0.14168, + "grad_norm": 1.2189229582966636, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14168 + }, + { + "epoch": 0.14169, + "grad_norm": 1.1528598161429937, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 14169 + }, + { + "epoch": 0.1417, + "grad_norm": 1.360734180095269, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14170 + }, + { + "epoch": 0.14171, + "grad_norm": 1.0189899619006824, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 14171 + }, + { + "epoch": 0.14172, + "grad_norm": 1.2409046032040727, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 14172 + }, + { + "epoch": 0.14173, + "grad_norm": 1.1134978895708785, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 14173 + }, + { + "epoch": 0.14174, + "grad_norm": 1.428086065445783, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 14174 + }, + { + "epoch": 0.14175, + "grad_norm": 1.293778397312492, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 14175 + }, + { + "epoch": 0.14176, + "grad_norm": 1.2644397446378224, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 14176 + }, + { + "epoch": 0.14177, + "grad_norm": 0.9565849380097393, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 14177 + }, + { + "epoch": 0.14178, + "grad_norm": 1.3059769568691229, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 14178 + }, + { + "epoch": 0.14179, + "grad_norm": 1.0491674393087167, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 14179 + }, + { + "epoch": 0.1418, + "grad_norm": 1.1180949361288677, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 14180 + }, + { + "epoch": 0.14181, + "grad_norm": 1.1157322008472714, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 14181 + }, + { + "epoch": 0.14182, + "grad_norm": 1.3179222987926962, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 14182 + }, + { + "epoch": 0.14183, + "grad_norm": 1.2403172237457192, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 14183 + }, + { + "epoch": 0.14184, + "grad_norm": 1.336720325007302, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 14184 + }, + { + "epoch": 0.14185, + "grad_norm": 1.100527077534773, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 14185 + }, + { + "epoch": 0.14186, + "grad_norm": 1.2033854975669016, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 14186 + }, + { + "epoch": 0.14187, + "grad_norm": 1.2187189628320798, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 14187 + }, + { + "epoch": 0.14188, + "grad_norm": 1.0200176822347935, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 14188 + }, + { + "epoch": 0.14189, + "grad_norm": 1.5073251661283125, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 14189 + }, + { + "epoch": 0.1419, + "grad_norm": 1.0512805060783437, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14190 + }, + { + "epoch": 0.14191, + "grad_norm": 1.167991755103892, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 14191 + }, + { + "epoch": 0.14192, + "grad_norm": 1.1772584721220622, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 14192 + }, + { + "epoch": 0.14193, + "grad_norm": 1.1976833915329468, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 14193 + }, + { + "epoch": 0.14194, + "grad_norm": 1.190152113141832, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 14194 + }, + { + "epoch": 0.14195, + "grad_norm": 1.1151335344861781, + "learning_rate": 0.003, + "loss": 4.032, + "step": 14195 + }, + { + "epoch": 0.14196, + "grad_norm": 1.2354467070065152, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 14196 + }, + { + "epoch": 0.14197, + "grad_norm": 1.0970972906795948, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 14197 + }, + { + "epoch": 0.14198, + "grad_norm": 1.177775059299803, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14198 + }, + { + "epoch": 0.14199, + "grad_norm": 1.144446470023133, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14199 + }, + { + "epoch": 0.142, + "grad_norm": 1.3620451445274342, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 14200 + }, + { + "epoch": 0.14201, + "grad_norm": 1.1392223147475815, + "learning_rate": 0.003, + "loss": 3.986, + "step": 14201 + }, + { + "epoch": 0.14202, + "grad_norm": 1.3754184187622027, + "learning_rate": 0.003, + "loss": 4.03, + "step": 14202 + }, + { + "epoch": 0.14203, + "grad_norm": 1.048321822711271, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 14203 + }, + { + "epoch": 0.14204, + "grad_norm": 1.3892998384218211, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 14204 + }, + { + "epoch": 0.14205, + "grad_norm": 1.051400001412997, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 14205 + }, + { + "epoch": 0.14206, + "grad_norm": 1.325490620508042, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 14206 + }, + { + "epoch": 0.14207, + "grad_norm": 1.13493940310957, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 14207 + }, + { + "epoch": 0.14208, + "grad_norm": 1.367807284383825, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 14208 + }, + { + "epoch": 0.14209, + "grad_norm": 1.1294605234034383, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 14209 + }, + { + "epoch": 0.1421, + "grad_norm": 1.343495245675508, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 14210 + }, + { + "epoch": 0.14211, + "grad_norm": 1.0038877220611284, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 14211 + }, + { + "epoch": 0.14212, + "grad_norm": 1.534988054028867, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 14212 + }, + { + "epoch": 0.14213, + "grad_norm": 1.0070828660250137, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 14213 + }, + { + "epoch": 0.14214, + "grad_norm": 1.5597998232138048, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 14214 + }, + { + "epoch": 0.14215, + "grad_norm": 1.0651385458745226, + "learning_rate": 0.003, + "loss": 4.011, + "step": 14215 + }, + { + "epoch": 0.14216, + "grad_norm": 1.3348080877703172, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 14216 + }, + { + "epoch": 0.14217, + "grad_norm": 0.9499056849838106, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 14217 + }, + { + "epoch": 0.14218, + "grad_norm": 1.1418179918085627, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 14218 + }, + { + "epoch": 0.14219, + "grad_norm": 1.3743036275552343, + "learning_rate": 0.003, + "loss": 4.016, + "step": 14219 + }, + { + "epoch": 0.1422, + "grad_norm": 1.0214575397901615, + "learning_rate": 0.003, + "loss": 3.999, + "step": 14220 + }, + { + "epoch": 0.14221, + "grad_norm": 1.5581151608740837, + "learning_rate": 0.003, + "loss": 4.015, + "step": 14221 + }, + { + "epoch": 0.14222, + "grad_norm": 1.027127027705112, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 14222 + }, + { + "epoch": 0.14223, + "grad_norm": 1.4981227018638756, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 14223 + }, + { + "epoch": 0.14224, + "grad_norm": 1.008107220993946, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 14224 + }, + { + "epoch": 0.14225, + "grad_norm": 1.3732868322500928, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14225 + }, + { + "epoch": 0.14226, + "grad_norm": 1.0437936957982215, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 14226 + }, + { + "epoch": 0.14227, + "grad_norm": 1.2731748595843275, + "learning_rate": 0.003, + "loss": 4.025, + "step": 14227 + }, + { + "epoch": 0.14228, + "grad_norm": 1.3649598712722133, + "learning_rate": 0.003, + "loss": 4.016, + "step": 14228 + }, + { + "epoch": 0.14229, + "grad_norm": 1.1752436368583978, + "learning_rate": 0.003, + "loss": 4.018, + "step": 14229 + }, + { + "epoch": 0.1423, + "grad_norm": 1.2731387940688244, + "learning_rate": 0.003, + "loss": 4.004, + "step": 14230 + }, + { + "epoch": 0.14231, + "grad_norm": 0.9888217773791853, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 14231 + }, + { + "epoch": 0.14232, + "grad_norm": 1.2419409956986254, + "learning_rate": 0.003, + "loss": 4.017, + "step": 14232 + }, + { + "epoch": 0.14233, + "grad_norm": 1.1208665504365474, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 14233 + }, + { + "epoch": 0.14234, + "grad_norm": 1.3934997178845823, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 14234 + }, + { + "epoch": 0.14235, + "grad_norm": 0.9939585794166588, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14235 + }, + { + "epoch": 0.14236, + "grad_norm": 1.466049674388318, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 14236 + }, + { + "epoch": 0.14237, + "grad_norm": 1.1516383833090853, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 14237 + }, + { + "epoch": 0.14238, + "grad_norm": 1.1905120240269704, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 14238 + }, + { + "epoch": 0.14239, + "grad_norm": 1.320357055484459, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 14239 + }, + { + "epoch": 0.1424, + "grad_norm": 1.1739371270086971, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 14240 + }, + { + "epoch": 0.14241, + "grad_norm": 1.1953821818276693, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 14241 + }, + { + "epoch": 0.14242, + "grad_norm": 1.049876836420036, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 14242 + }, + { + "epoch": 0.14243, + "grad_norm": 1.0654099565056714, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 14243 + }, + { + "epoch": 0.14244, + "grad_norm": 1.3029073713304329, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14244 + }, + { + "epoch": 0.14245, + "grad_norm": 1.256556933511795, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 14245 + }, + { + "epoch": 0.14246, + "grad_norm": 1.2772979759682255, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 14246 + }, + { + "epoch": 0.14247, + "grad_norm": 1.2552451182459947, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 14247 + }, + { + "epoch": 0.14248, + "grad_norm": 0.9939790102981942, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 14248 + }, + { + "epoch": 0.14249, + "grad_norm": 1.3178538598529268, + "learning_rate": 0.003, + "loss": 3.977, + "step": 14249 + }, + { + "epoch": 0.1425, + "grad_norm": 1.1601483906677303, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 14250 + }, + { + "epoch": 0.14251, + "grad_norm": 1.221185680742918, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 14251 + }, + { + "epoch": 0.14252, + "grad_norm": 1.2474809124796564, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 14252 + }, + { + "epoch": 0.14253, + "grad_norm": 1.1992344745787196, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 14253 + }, + { + "epoch": 0.14254, + "grad_norm": 0.9132225327909732, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 14254 + }, + { + "epoch": 0.14255, + "grad_norm": 1.1811406392830759, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14255 + }, + { + "epoch": 0.14256, + "grad_norm": 1.3453499901407837, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 14256 + }, + { + "epoch": 0.14257, + "grad_norm": 1.1500286543886105, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 14257 + }, + { + "epoch": 0.14258, + "grad_norm": 1.4664633371094062, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 14258 + }, + { + "epoch": 0.14259, + "grad_norm": 1.119114866847188, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14259 + }, + { + "epoch": 0.1426, + "grad_norm": 1.3478008305817595, + "learning_rate": 0.003, + "loss": 4.011, + "step": 14260 + }, + { + "epoch": 0.14261, + "grad_norm": 0.9832585037994765, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14261 + }, + { + "epoch": 0.14262, + "grad_norm": 1.1548328309543094, + "learning_rate": 0.003, + "loss": 3.99, + "step": 14262 + }, + { + "epoch": 0.14263, + "grad_norm": 1.1540288232997642, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14263 + }, + { + "epoch": 0.14264, + "grad_norm": 1.186729047593634, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 14264 + }, + { + "epoch": 0.14265, + "grad_norm": 1.1810656564347743, + "learning_rate": 0.003, + "loss": 4.035, + "step": 14265 + }, + { + "epoch": 0.14266, + "grad_norm": 1.4770408871743086, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 14266 + }, + { + "epoch": 0.14267, + "grad_norm": 0.9938931214161333, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 14267 + }, + { + "epoch": 0.14268, + "grad_norm": 1.3149471709837792, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 14268 + }, + { + "epoch": 0.14269, + "grad_norm": 1.2328045329778488, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 14269 + }, + { + "epoch": 0.1427, + "grad_norm": 1.1250321693454324, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 14270 + }, + { + "epoch": 0.14271, + "grad_norm": 0.9889732485401035, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 14271 + }, + { + "epoch": 0.14272, + "grad_norm": 1.1476898981247932, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 14272 + }, + { + "epoch": 0.14273, + "grad_norm": 1.1025789465267635, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 14273 + }, + { + "epoch": 0.14274, + "grad_norm": 1.122764833640579, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 14274 + }, + { + "epoch": 0.14275, + "grad_norm": 1.147038214357953, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 14275 + }, + { + "epoch": 0.14276, + "grad_norm": 1.2162589652144875, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 14276 + }, + { + "epoch": 0.14277, + "grad_norm": 1.0566978742450486, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 14277 + }, + { + "epoch": 0.14278, + "grad_norm": 1.7169108876688428, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 14278 + }, + { + "epoch": 0.14279, + "grad_norm": 0.876042071141214, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 14279 + }, + { + "epoch": 0.1428, + "grad_norm": 1.1084611799228745, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 14280 + }, + { + "epoch": 0.14281, + "grad_norm": 1.2709943034324318, + "learning_rate": 0.003, + "loss": 4.028, + "step": 14281 + }, + { + "epoch": 0.14282, + "grad_norm": 1.3812282195884515, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 14282 + }, + { + "epoch": 0.14283, + "grad_norm": 1.1566429269433423, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 14283 + }, + { + "epoch": 0.14284, + "grad_norm": 1.2835084133322305, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 14284 + }, + { + "epoch": 0.14285, + "grad_norm": 1.055340094724688, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 14285 + }, + { + "epoch": 0.14286, + "grad_norm": 1.4105259998316546, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 14286 + }, + { + "epoch": 0.14287, + "grad_norm": 1.2550811667078823, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 14287 + }, + { + "epoch": 0.14288, + "grad_norm": 1.1132410163567108, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 14288 + }, + { + "epoch": 0.14289, + "grad_norm": 1.501973582017856, + "learning_rate": 0.003, + "loss": 4.017, + "step": 14289 + }, + { + "epoch": 0.1429, + "grad_norm": 0.9719422279166505, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 14290 + }, + { + "epoch": 0.14291, + "grad_norm": 1.3825012805536943, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 14291 + }, + { + "epoch": 0.14292, + "grad_norm": 1.1099027929597218, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14292 + }, + { + "epoch": 0.14293, + "grad_norm": 1.3461577718551276, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 14293 + }, + { + "epoch": 0.14294, + "grad_norm": 1.24973576329806, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 14294 + }, + { + "epoch": 0.14295, + "grad_norm": 1.4312573908271133, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 14295 + }, + { + "epoch": 0.14296, + "grad_norm": 0.9116815765076999, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14296 + }, + { + "epoch": 0.14297, + "grad_norm": 1.1882497269250194, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 14297 + }, + { + "epoch": 0.14298, + "grad_norm": 1.243886949272355, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 14298 + }, + { + "epoch": 0.14299, + "grad_norm": 1.3152173645237428, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14299 + }, + { + "epoch": 0.143, + "grad_norm": 1.0919137431492918, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 14300 + }, + { + "epoch": 0.14301, + "grad_norm": 1.3685064657521364, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 14301 + }, + { + "epoch": 0.14302, + "grad_norm": 1.0110048864429841, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 14302 + }, + { + "epoch": 0.14303, + "grad_norm": 1.4776803140572186, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 14303 + }, + { + "epoch": 0.14304, + "grad_norm": 1.0986579942741568, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 14304 + }, + { + "epoch": 0.14305, + "grad_norm": 1.1158216374588756, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 14305 + }, + { + "epoch": 0.14306, + "grad_norm": 1.2946086944401998, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 14306 + }, + { + "epoch": 0.14307, + "grad_norm": 1.1641566971726678, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 14307 + }, + { + "epoch": 0.14308, + "grad_norm": 1.208708792057949, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 14308 + }, + { + "epoch": 0.14309, + "grad_norm": 1.2861984661320631, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 14309 + }, + { + "epoch": 0.1431, + "grad_norm": 1.1096737366335085, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 14310 + }, + { + "epoch": 0.14311, + "grad_norm": 1.332880293436986, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 14311 + }, + { + "epoch": 0.14312, + "grad_norm": 1.1899327349214601, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14312 + }, + { + "epoch": 0.14313, + "grad_norm": 1.585214015131889, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 14313 + }, + { + "epoch": 0.14314, + "grad_norm": 0.7871645445442891, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 14314 + }, + { + "epoch": 0.14315, + "grad_norm": 1.0506442944194525, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 14315 + }, + { + "epoch": 0.14316, + "grad_norm": 1.524745255403895, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14316 + }, + { + "epoch": 0.14317, + "grad_norm": 1.074288517670019, + "learning_rate": 0.003, + "loss": 4.016, + "step": 14317 + }, + { + "epoch": 0.14318, + "grad_norm": 1.5176426081616132, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 14318 + }, + { + "epoch": 0.14319, + "grad_norm": 1.004949697833761, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 14319 + }, + { + "epoch": 0.1432, + "grad_norm": 1.3267022711433671, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 14320 + }, + { + "epoch": 0.14321, + "grad_norm": 1.0043666421766428, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 14321 + }, + { + "epoch": 0.14322, + "grad_norm": 1.23646610606199, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 14322 + }, + { + "epoch": 0.14323, + "grad_norm": 1.2695267039105103, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 14323 + }, + { + "epoch": 0.14324, + "grad_norm": 1.091829883598905, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 14324 + }, + { + "epoch": 0.14325, + "grad_norm": 1.3614936043265236, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 14325 + }, + { + "epoch": 0.14326, + "grad_norm": 1.1215395898553848, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 14326 + }, + { + "epoch": 0.14327, + "grad_norm": 1.2154310116743614, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 14327 + }, + { + "epoch": 0.14328, + "grad_norm": 1.1889107742210372, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 14328 + }, + { + "epoch": 0.14329, + "grad_norm": 1.1507054007861035, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 14329 + }, + { + "epoch": 0.1433, + "grad_norm": 1.330111633581526, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 14330 + }, + { + "epoch": 0.14331, + "grad_norm": 1.0768442491212724, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 14331 + }, + { + "epoch": 0.14332, + "grad_norm": 1.239911147068525, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 14332 + }, + { + "epoch": 0.14333, + "grad_norm": 1.323571559451598, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 14333 + }, + { + "epoch": 0.14334, + "grad_norm": 1.1376701926449524, + "learning_rate": 0.003, + "loss": 4.017, + "step": 14334 + }, + { + "epoch": 0.14335, + "grad_norm": 1.3340901235718514, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 14335 + }, + { + "epoch": 0.14336, + "grad_norm": 1.2686228303701885, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 14336 + }, + { + "epoch": 0.14337, + "grad_norm": 1.0467394749911172, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 14337 + }, + { + "epoch": 0.14338, + "grad_norm": 1.216880189998311, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 14338 + }, + { + "epoch": 0.14339, + "grad_norm": 1.1505835170743297, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 14339 + }, + { + "epoch": 0.1434, + "grad_norm": 1.1715423738629298, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 14340 + }, + { + "epoch": 0.14341, + "grad_norm": 1.3591295789133988, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 14341 + }, + { + "epoch": 0.14342, + "grad_norm": 1.0802743236414356, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14342 + }, + { + "epoch": 0.14343, + "grad_norm": 1.2314655572568123, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 14343 + }, + { + "epoch": 0.14344, + "grad_norm": 1.2170050499547125, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 14344 + }, + { + "epoch": 0.14345, + "grad_norm": 1.1764760173694266, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 14345 + }, + { + "epoch": 0.14346, + "grad_norm": 1.4899350031793006, + "learning_rate": 0.003, + "loss": 4.025, + "step": 14346 + }, + { + "epoch": 0.14347, + "grad_norm": 1.0758576003187106, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 14347 + }, + { + "epoch": 0.14348, + "grad_norm": 1.3635564326619183, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 14348 + }, + { + "epoch": 0.14349, + "grad_norm": 1.2235786275907428, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 14349 + }, + { + "epoch": 0.1435, + "grad_norm": 1.2436695980903918, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 14350 + }, + { + "epoch": 0.14351, + "grad_norm": 1.210672505276866, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14351 + }, + { + "epoch": 0.14352, + "grad_norm": 1.1545045673205454, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 14352 + }, + { + "epoch": 0.14353, + "grad_norm": 1.3725191826934457, + "learning_rate": 0.003, + "loss": 4.014, + "step": 14353 + }, + { + "epoch": 0.14354, + "grad_norm": 1.1897111683673434, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 14354 + }, + { + "epoch": 0.14355, + "grad_norm": 1.5096476297328265, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 14355 + }, + { + "epoch": 0.14356, + "grad_norm": 0.8717003987565817, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 14356 + }, + { + "epoch": 0.14357, + "grad_norm": 1.0792518544905656, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 14357 + }, + { + "epoch": 0.14358, + "grad_norm": 1.351721600532738, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 14358 + }, + { + "epoch": 0.14359, + "grad_norm": 0.996693627500812, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14359 + }, + { + "epoch": 0.1436, + "grad_norm": 1.2646311347446006, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 14360 + }, + { + "epoch": 0.14361, + "grad_norm": 1.0228518544107583, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 14361 + }, + { + "epoch": 0.14362, + "grad_norm": 1.256227677652288, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 14362 + }, + { + "epoch": 0.14363, + "grad_norm": 1.166195059307722, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 14363 + }, + { + "epoch": 0.14364, + "grad_norm": 1.172896560370709, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 14364 + }, + { + "epoch": 0.14365, + "grad_norm": 1.2194603487156364, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 14365 + }, + { + "epoch": 0.14366, + "grad_norm": 1.1945393859702242, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14366 + }, + { + "epoch": 0.14367, + "grad_norm": 1.1653822712601563, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 14367 + }, + { + "epoch": 0.14368, + "grad_norm": 1.2726517892617004, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 14368 + }, + { + "epoch": 0.14369, + "grad_norm": 1.244704823359052, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 14369 + }, + { + "epoch": 0.1437, + "grad_norm": 1.2172796802879724, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 14370 + }, + { + "epoch": 0.14371, + "grad_norm": 1.2603191200469879, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 14371 + }, + { + "epoch": 0.14372, + "grad_norm": 1.0718879428269645, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 14372 + }, + { + "epoch": 0.14373, + "grad_norm": 1.3330073165146854, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 14373 + }, + { + "epoch": 0.14374, + "grad_norm": 1.2012912275507692, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 14374 + }, + { + "epoch": 0.14375, + "grad_norm": 1.1908696738600753, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 14375 + }, + { + "epoch": 0.14376, + "grad_norm": 1.3507919992492639, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 14376 + }, + { + "epoch": 0.14377, + "grad_norm": 1.1733220329176508, + "learning_rate": 0.003, + "loss": 4.013, + "step": 14377 + }, + { + "epoch": 0.14378, + "grad_norm": 1.363297551963369, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14378 + }, + { + "epoch": 0.14379, + "grad_norm": 1.0313070601521575, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 14379 + }, + { + "epoch": 0.1438, + "grad_norm": 1.3397131378092488, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 14380 + }, + { + "epoch": 0.14381, + "grad_norm": 1.0121858969636834, + "learning_rate": 0.003, + "loss": 3.997, + "step": 14381 + }, + { + "epoch": 0.14382, + "grad_norm": 1.390088716518865, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 14382 + }, + { + "epoch": 0.14383, + "grad_norm": 1.1129271392471018, + "learning_rate": 0.003, + "loss": 4.02, + "step": 14383 + }, + { + "epoch": 0.14384, + "grad_norm": 1.2375419522985815, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 14384 + }, + { + "epoch": 0.14385, + "grad_norm": 1.3673457599885632, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 14385 + }, + { + "epoch": 0.14386, + "grad_norm": 0.9916084434443762, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 14386 + }, + { + "epoch": 0.14387, + "grad_norm": 1.4002355526911558, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 14387 + }, + { + "epoch": 0.14388, + "grad_norm": 1.1109558133777058, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 14388 + }, + { + "epoch": 0.14389, + "grad_norm": 1.3626692795113289, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 14389 + }, + { + "epoch": 0.1439, + "grad_norm": 1.2460630032368645, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14390 + }, + { + "epoch": 0.14391, + "grad_norm": 1.2243928195308638, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 14391 + }, + { + "epoch": 0.14392, + "grad_norm": 1.1784102432358567, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 14392 + }, + { + "epoch": 0.14393, + "grad_norm": 1.4939716515347823, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 14393 + }, + { + "epoch": 0.14394, + "grad_norm": 1.1498274365748753, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 14394 + }, + { + "epoch": 0.14395, + "grad_norm": 1.200810908936847, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 14395 + }, + { + "epoch": 0.14396, + "grad_norm": 1.0745855059534282, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 14396 + }, + { + "epoch": 0.14397, + "grad_norm": 1.277116387315723, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 14397 + }, + { + "epoch": 0.14398, + "grad_norm": 0.9419150034212755, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 14398 + }, + { + "epoch": 0.14399, + "grad_norm": 1.3295125544631583, + "learning_rate": 0.003, + "loss": 3.999, + "step": 14399 + }, + { + "epoch": 0.144, + "grad_norm": 1.1989912070727897, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 14400 + }, + { + "epoch": 0.14401, + "grad_norm": 1.3189184717411828, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 14401 + }, + { + "epoch": 0.14402, + "grad_norm": 1.0816437140157895, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 14402 + }, + { + "epoch": 0.14403, + "grad_norm": 1.4360315372365975, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 14403 + }, + { + "epoch": 0.14404, + "grad_norm": 1.2193549764527278, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 14404 + }, + { + "epoch": 0.14405, + "grad_norm": 1.0823921330138107, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 14405 + }, + { + "epoch": 0.14406, + "grad_norm": 1.1963637207195172, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 14406 + }, + { + "epoch": 0.14407, + "grad_norm": 1.297169350319569, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 14407 + }, + { + "epoch": 0.14408, + "grad_norm": 1.275499530630152, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 14408 + }, + { + "epoch": 0.14409, + "grad_norm": 1.228439110366136, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 14409 + }, + { + "epoch": 0.1441, + "grad_norm": 1.5688924175322991, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 14410 + }, + { + "epoch": 0.14411, + "grad_norm": 1.0120584218587096, + "learning_rate": 0.003, + "loss": 4.002, + "step": 14411 + }, + { + "epoch": 0.14412, + "grad_norm": 1.3011317791543677, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 14412 + }, + { + "epoch": 0.14413, + "grad_norm": 1.0668175565997582, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14413 + }, + { + "epoch": 0.14414, + "grad_norm": 1.5931538460060055, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 14414 + }, + { + "epoch": 0.14415, + "grad_norm": 1.078989729152812, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 14415 + }, + { + "epoch": 0.14416, + "grad_norm": 1.1701151551911053, + "learning_rate": 0.003, + "loss": 3.99, + "step": 14416 + }, + { + "epoch": 0.14417, + "grad_norm": 1.0128987161588474, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 14417 + }, + { + "epoch": 0.14418, + "grad_norm": 1.3226158471972884, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 14418 + }, + { + "epoch": 0.14419, + "grad_norm": 1.0391012761182818, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14419 + }, + { + "epoch": 0.1442, + "grad_norm": 1.2995035977172058, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 14420 + }, + { + "epoch": 0.14421, + "grad_norm": 1.1675231453686032, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 14421 + }, + { + "epoch": 0.14422, + "grad_norm": 1.1990662827938074, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 14422 + }, + { + "epoch": 0.14423, + "grad_norm": 1.3341843292074833, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 14423 + }, + { + "epoch": 0.14424, + "grad_norm": 1.255184302280917, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 14424 + }, + { + "epoch": 0.14425, + "grad_norm": 1.213102381507609, + "learning_rate": 0.003, + "loss": 3.992, + "step": 14425 + }, + { + "epoch": 0.14426, + "grad_norm": 1.460646676879946, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 14426 + }, + { + "epoch": 0.14427, + "grad_norm": 1.0654397567379355, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14427 + }, + { + "epoch": 0.14428, + "grad_norm": 1.4763855575417097, + "learning_rate": 0.003, + "loss": 4.009, + "step": 14428 + }, + { + "epoch": 0.14429, + "grad_norm": 1.236018760227688, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 14429 + }, + { + "epoch": 0.1443, + "grad_norm": 1.1701305528800572, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 14430 + }, + { + "epoch": 0.14431, + "grad_norm": 1.0277455869468615, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 14431 + }, + { + "epoch": 0.14432, + "grad_norm": 1.264236895061662, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 14432 + }, + { + "epoch": 0.14433, + "grad_norm": 1.1450179472542663, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14433 + }, + { + "epoch": 0.14434, + "grad_norm": 1.3260135156923754, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 14434 + }, + { + "epoch": 0.14435, + "grad_norm": 1.0950989832118134, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 14435 + }, + { + "epoch": 0.14436, + "grad_norm": 1.3891352119599327, + "learning_rate": 0.003, + "loss": 4.046, + "step": 14436 + }, + { + "epoch": 0.14437, + "grad_norm": 1.051389794925024, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 14437 + }, + { + "epoch": 0.14438, + "grad_norm": 1.4529679895972711, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 14438 + }, + { + "epoch": 0.14439, + "grad_norm": 0.9878120297750842, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 14439 + }, + { + "epoch": 0.1444, + "grad_norm": 1.3500863354125725, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 14440 + }, + { + "epoch": 0.14441, + "grad_norm": 1.0391767489442227, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 14441 + }, + { + "epoch": 0.14442, + "grad_norm": 1.3341558927499608, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 14442 + }, + { + "epoch": 0.14443, + "grad_norm": 1.2367517579258847, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 14443 + }, + { + "epoch": 0.14444, + "grad_norm": 0.999685303003077, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 14444 + }, + { + "epoch": 0.14445, + "grad_norm": 1.637842531994357, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 14445 + }, + { + "epoch": 0.14446, + "grad_norm": 1.1935103420326467, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 14446 + }, + { + "epoch": 0.14447, + "grad_norm": 1.3946879258079155, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 14447 + }, + { + "epoch": 0.14448, + "grad_norm": 1.001298245674135, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 14448 + }, + { + "epoch": 0.14449, + "grad_norm": 1.273338157626047, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 14449 + }, + { + "epoch": 0.1445, + "grad_norm": 1.0845018678469514, + "learning_rate": 0.003, + "loss": 4.033, + "step": 14450 + }, + { + "epoch": 0.14451, + "grad_norm": 1.5101736035199864, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14451 + }, + { + "epoch": 0.14452, + "grad_norm": 1.0905745218691874, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 14452 + }, + { + "epoch": 0.14453, + "grad_norm": 1.3842347273258648, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14453 + }, + { + "epoch": 0.14454, + "grad_norm": 1.2922298534952803, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 14454 + }, + { + "epoch": 0.14455, + "grad_norm": 1.0925683960156223, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 14455 + }, + { + "epoch": 0.14456, + "grad_norm": 1.3559823940041604, + "learning_rate": 0.003, + "loss": 3.989, + "step": 14456 + }, + { + "epoch": 0.14457, + "grad_norm": 1.2658248605229905, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 14457 + }, + { + "epoch": 0.14458, + "grad_norm": 1.0732730407104027, + "learning_rate": 0.003, + "loss": 4.003, + "step": 14458 + }, + { + "epoch": 0.14459, + "grad_norm": 1.160826402475991, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 14459 + }, + { + "epoch": 0.1446, + "grad_norm": 1.2237950140205267, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 14460 + }, + { + "epoch": 0.14461, + "grad_norm": 1.1535901728274076, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 14461 + }, + { + "epoch": 0.14462, + "grad_norm": 1.3750677627530832, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 14462 + }, + { + "epoch": 0.14463, + "grad_norm": 1.1796103564127114, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 14463 + }, + { + "epoch": 0.14464, + "grad_norm": 1.422757421374509, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 14464 + }, + { + "epoch": 0.14465, + "grad_norm": 1.1045263307117212, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 14465 + }, + { + "epoch": 0.14466, + "grad_norm": 1.180752551862371, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14466 + }, + { + "epoch": 0.14467, + "grad_norm": 1.2050557387864302, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 14467 + }, + { + "epoch": 0.14468, + "grad_norm": 1.0681333646184352, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 14468 + }, + { + "epoch": 0.14469, + "grad_norm": 1.3246381531360234, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 14469 + }, + { + "epoch": 0.1447, + "grad_norm": 1.1523402047220455, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 14470 + }, + { + "epoch": 0.14471, + "grad_norm": 1.2207650779108505, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 14471 + }, + { + "epoch": 0.14472, + "grad_norm": 1.1728700266594385, + "learning_rate": 0.003, + "loss": 4.009, + "step": 14472 + }, + { + "epoch": 0.14473, + "grad_norm": 1.250042755849908, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 14473 + }, + { + "epoch": 0.14474, + "grad_norm": 1.3570314137605364, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 14474 + }, + { + "epoch": 0.14475, + "grad_norm": 1.0777052849486088, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 14475 + }, + { + "epoch": 0.14476, + "grad_norm": 1.2889017126743465, + "learning_rate": 0.003, + "loss": 4.043, + "step": 14476 + }, + { + "epoch": 0.14477, + "grad_norm": 1.0346286013190817, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 14477 + }, + { + "epoch": 0.14478, + "grad_norm": 1.1620056393435554, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 14478 + }, + { + "epoch": 0.14479, + "grad_norm": 1.484373566585143, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 14479 + }, + { + "epoch": 0.1448, + "grad_norm": 1.1443086993542309, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 14480 + }, + { + "epoch": 0.14481, + "grad_norm": 1.2698849354218817, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 14481 + }, + { + "epoch": 0.14482, + "grad_norm": 1.0087013208617437, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 14482 + }, + { + "epoch": 0.14483, + "grad_norm": 1.4960437345417261, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 14483 + }, + { + "epoch": 0.14484, + "grad_norm": 1.0081903193331918, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 14484 + }, + { + "epoch": 0.14485, + "grad_norm": 1.3618954479781298, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14485 + }, + { + "epoch": 0.14486, + "grad_norm": 1.174742735761288, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 14486 + }, + { + "epoch": 0.14487, + "grad_norm": 1.3257975675308202, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 14487 + }, + { + "epoch": 0.14488, + "grad_norm": 1.290576379160918, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 14488 + }, + { + "epoch": 0.14489, + "grad_norm": 1.1448041897333847, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 14489 + }, + { + "epoch": 0.1449, + "grad_norm": 1.318156067960972, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 14490 + }, + { + "epoch": 0.14491, + "grad_norm": 1.0646783634293895, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 14491 + }, + { + "epoch": 0.14492, + "grad_norm": 1.32623979475304, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 14492 + }, + { + "epoch": 0.14493, + "grad_norm": 1.1016488012244479, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 14493 + }, + { + "epoch": 0.14494, + "grad_norm": 1.3909336423726262, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 14494 + }, + { + "epoch": 0.14495, + "grad_norm": 1.1601648377855989, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 14495 + }, + { + "epoch": 0.14496, + "grad_norm": 1.009980094021526, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 14496 + }, + { + "epoch": 0.14497, + "grad_norm": 1.1757064908364916, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 14497 + }, + { + "epoch": 0.14498, + "grad_norm": 1.2557706682498537, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 14498 + }, + { + "epoch": 0.14499, + "grad_norm": 1.0063481278247886, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 14499 + }, + { + "epoch": 0.145, + "grad_norm": 1.2600965724146316, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 14500 + }, + { + "epoch": 0.14501, + "grad_norm": 1.058860986503806, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 14501 + }, + { + "epoch": 0.14502, + "grad_norm": 1.2407155637824785, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14502 + }, + { + "epoch": 0.14503, + "grad_norm": 1.309155248761219, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 14503 + }, + { + "epoch": 0.14504, + "grad_norm": 1.4214121220645077, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 14504 + }, + { + "epoch": 0.14505, + "grad_norm": 1.0404374530095637, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 14505 + }, + { + "epoch": 0.14506, + "grad_norm": 1.2873683959126516, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 14506 + }, + { + "epoch": 0.14507, + "grad_norm": 1.2001126154689117, + "learning_rate": 0.003, + "loss": 3.986, + "step": 14507 + }, + { + "epoch": 0.14508, + "grad_norm": 1.1862564918121277, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 14508 + }, + { + "epoch": 0.14509, + "grad_norm": 1.1429814837311603, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 14509 + }, + { + "epoch": 0.1451, + "grad_norm": 1.0366689188633833, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 14510 + }, + { + "epoch": 0.14511, + "grad_norm": 1.2689118182788326, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 14511 + }, + { + "epoch": 0.14512, + "grad_norm": 1.221645510450628, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 14512 + }, + { + "epoch": 0.14513, + "grad_norm": 1.0676268731187524, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 14513 + }, + { + "epoch": 0.14514, + "grad_norm": 1.3803401448631745, + "learning_rate": 0.003, + "loss": 4.037, + "step": 14514 + }, + { + "epoch": 0.14515, + "grad_norm": 1.0381879543839039, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 14515 + }, + { + "epoch": 0.14516, + "grad_norm": 1.3728094284848495, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 14516 + }, + { + "epoch": 0.14517, + "grad_norm": 1.2038839059792454, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 14517 + }, + { + "epoch": 0.14518, + "grad_norm": 1.5016318963257844, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 14518 + }, + { + "epoch": 0.14519, + "grad_norm": 1.1748809794876305, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 14519 + }, + { + "epoch": 0.1452, + "grad_norm": 1.4488835508199935, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 14520 + }, + { + "epoch": 0.14521, + "grad_norm": 0.9583289734041514, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 14521 + }, + { + "epoch": 0.14522, + "grad_norm": 1.2530762804116282, + "learning_rate": 0.003, + "loss": 3.986, + "step": 14522 + }, + { + "epoch": 0.14523, + "grad_norm": 1.3470868291082656, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 14523 + }, + { + "epoch": 0.14524, + "grad_norm": 1.516748058417483, + "learning_rate": 0.003, + "loss": 4.05, + "step": 14524 + }, + { + "epoch": 0.14525, + "grad_norm": 0.8688628064559125, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 14525 + }, + { + "epoch": 0.14526, + "grad_norm": 1.2290374421396448, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 14526 + }, + { + "epoch": 0.14527, + "grad_norm": 1.4106250986583697, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 14527 + }, + { + "epoch": 0.14528, + "grad_norm": 1.1960020034701262, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 14528 + }, + { + "epoch": 0.14529, + "grad_norm": 1.2461522806726895, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 14529 + }, + { + "epoch": 0.1453, + "grad_norm": 1.343435581151887, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 14530 + }, + { + "epoch": 0.14531, + "grad_norm": 1.1897863146723584, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 14531 + }, + { + "epoch": 0.14532, + "grad_norm": 1.2406366984067851, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 14532 + }, + { + "epoch": 0.14533, + "grad_norm": 1.2770937031087264, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 14533 + }, + { + "epoch": 0.14534, + "grad_norm": 1.1333862009610007, + "learning_rate": 0.003, + "loss": 4.027, + "step": 14534 + }, + { + "epoch": 0.14535, + "grad_norm": 1.1215687848618767, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 14535 + }, + { + "epoch": 0.14536, + "grad_norm": 1.2456154158145807, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 14536 + }, + { + "epoch": 0.14537, + "grad_norm": 0.9894400543840622, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 14537 + }, + { + "epoch": 0.14538, + "grad_norm": 1.4556690665591743, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 14538 + }, + { + "epoch": 0.14539, + "grad_norm": 1.1367777065612945, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 14539 + }, + { + "epoch": 0.1454, + "grad_norm": 1.301526864027351, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 14540 + }, + { + "epoch": 0.14541, + "grad_norm": 1.2157751688778053, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 14541 + }, + { + "epoch": 0.14542, + "grad_norm": 0.9290021911687516, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 14542 + }, + { + "epoch": 0.14543, + "grad_norm": 1.4507913936570669, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 14543 + }, + { + "epoch": 0.14544, + "grad_norm": 1.2253577184771525, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 14544 + }, + { + "epoch": 0.14545, + "grad_norm": 1.1595512555771437, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 14545 + }, + { + "epoch": 0.14546, + "grad_norm": 1.4697108397166665, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 14546 + }, + { + "epoch": 0.14547, + "grad_norm": 1.204273495324109, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 14547 + }, + { + "epoch": 0.14548, + "grad_norm": 1.2210909540672343, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 14548 + }, + { + "epoch": 0.14549, + "grad_norm": 1.1960358153078625, + "learning_rate": 0.003, + "loss": 4.031, + "step": 14549 + }, + { + "epoch": 0.1455, + "grad_norm": 1.168400476383576, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 14550 + }, + { + "epoch": 0.14551, + "grad_norm": 1.3522931121842838, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 14551 + }, + { + "epoch": 0.14552, + "grad_norm": 0.9815479797794744, + "learning_rate": 0.003, + "loss": 4.038, + "step": 14552 + }, + { + "epoch": 0.14553, + "grad_norm": 1.2883507256957447, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 14553 + }, + { + "epoch": 0.14554, + "grad_norm": 1.0742019647017573, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 14554 + }, + { + "epoch": 0.14555, + "grad_norm": 1.1017979923376091, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 14555 + }, + { + "epoch": 0.14556, + "grad_norm": 1.0805201358805017, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 14556 + }, + { + "epoch": 0.14557, + "grad_norm": 1.4082542123251285, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 14557 + }, + { + "epoch": 0.14558, + "grad_norm": 1.2638267673234083, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 14558 + }, + { + "epoch": 0.14559, + "grad_norm": 1.2636206096964009, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 14559 + }, + { + "epoch": 0.1456, + "grad_norm": 1.2256733740875092, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 14560 + }, + { + "epoch": 0.14561, + "grad_norm": 1.4019023401389583, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 14561 + }, + { + "epoch": 0.14562, + "grad_norm": 0.987825905196306, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 14562 + }, + { + "epoch": 0.14563, + "grad_norm": 1.3195123345596431, + "learning_rate": 0.003, + "loss": 4.038, + "step": 14563 + }, + { + "epoch": 0.14564, + "grad_norm": 1.1794370029884693, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 14564 + }, + { + "epoch": 0.14565, + "grad_norm": 1.1246915934472845, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 14565 + }, + { + "epoch": 0.14566, + "grad_norm": 1.1994594345382452, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 14566 + }, + { + "epoch": 0.14567, + "grad_norm": 1.2891949561316804, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 14567 + }, + { + "epoch": 0.14568, + "grad_norm": 1.234574364873565, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 14568 + }, + { + "epoch": 0.14569, + "grad_norm": 1.1069870175505885, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 14569 + }, + { + "epoch": 0.1457, + "grad_norm": 1.5472796516265446, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 14570 + }, + { + "epoch": 0.14571, + "grad_norm": 1.0606829134008178, + "learning_rate": 0.003, + "loss": 4.014, + "step": 14571 + }, + { + "epoch": 0.14572, + "grad_norm": 1.3441520210728717, + "learning_rate": 0.003, + "loss": 4.015, + "step": 14572 + }, + { + "epoch": 0.14573, + "grad_norm": 1.2160307278040676, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14573 + }, + { + "epoch": 0.14574, + "grad_norm": 1.3261360269716858, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 14574 + }, + { + "epoch": 0.14575, + "grad_norm": 1.0168868414381258, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 14575 + }, + { + "epoch": 0.14576, + "grad_norm": 1.398200028892131, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 14576 + }, + { + "epoch": 0.14577, + "grad_norm": 0.975521183755697, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 14577 + }, + { + "epoch": 0.14578, + "grad_norm": 1.347932974764977, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 14578 + }, + { + "epoch": 0.14579, + "grad_norm": 1.320107335942108, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 14579 + }, + { + "epoch": 0.1458, + "grad_norm": 1.1804506897454756, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 14580 + }, + { + "epoch": 0.14581, + "grad_norm": 1.511927370026859, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 14581 + }, + { + "epoch": 0.14582, + "grad_norm": 0.9657891579308665, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 14582 + }, + { + "epoch": 0.14583, + "grad_norm": 1.5982287604362448, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 14583 + }, + { + "epoch": 0.14584, + "grad_norm": 1.2063158861095558, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 14584 + }, + { + "epoch": 0.14585, + "grad_norm": 1.3204813971301048, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 14585 + }, + { + "epoch": 0.14586, + "grad_norm": 1.165718809426018, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 14586 + }, + { + "epoch": 0.14587, + "grad_norm": 1.4342957416193656, + "learning_rate": 0.003, + "loss": 4.029, + "step": 14587 + }, + { + "epoch": 0.14588, + "grad_norm": 1.3097579948327902, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14588 + }, + { + "epoch": 0.14589, + "grad_norm": 1.1819895455392044, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 14589 + }, + { + "epoch": 0.1459, + "grad_norm": 1.1519733997254258, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 14590 + }, + { + "epoch": 0.14591, + "grad_norm": 1.196795608704991, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 14591 + }, + { + "epoch": 0.14592, + "grad_norm": 1.212206219142562, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 14592 + }, + { + "epoch": 0.14593, + "grad_norm": 1.3748622890900157, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 14593 + }, + { + "epoch": 0.14594, + "grad_norm": 1.1321959038523033, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 14594 + }, + { + "epoch": 0.14595, + "grad_norm": 1.2479704493325805, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 14595 + }, + { + "epoch": 0.14596, + "grad_norm": 1.2157254751435294, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 14596 + }, + { + "epoch": 0.14597, + "grad_norm": 1.2113154403134787, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 14597 + }, + { + "epoch": 0.14598, + "grad_norm": 1.1652249828071028, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 14598 + }, + { + "epoch": 0.14599, + "grad_norm": 1.1249297186797358, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 14599 + }, + { + "epoch": 0.146, + "grad_norm": 1.1161516878373015, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 14600 + }, + { + "epoch": 0.14601, + "grad_norm": 1.2387498558393009, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14601 + }, + { + "epoch": 0.14602, + "grad_norm": 1.0064218516264403, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 14602 + }, + { + "epoch": 0.14603, + "grad_norm": 1.2928516477312995, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 14603 + }, + { + "epoch": 0.14604, + "grad_norm": 1.2079191749475882, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 14604 + }, + { + "epoch": 0.14605, + "grad_norm": 1.191874828203266, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 14605 + }, + { + "epoch": 0.14606, + "grad_norm": 1.2855761569668036, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 14606 + }, + { + "epoch": 0.14607, + "grad_norm": 1.3084030583655266, + "learning_rate": 0.003, + "loss": 4.025, + "step": 14607 + }, + { + "epoch": 0.14608, + "grad_norm": 1.0814048860065364, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 14608 + }, + { + "epoch": 0.14609, + "grad_norm": 1.318711360861515, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 14609 + }, + { + "epoch": 0.1461, + "grad_norm": 1.0843550800914612, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 14610 + }, + { + "epoch": 0.14611, + "grad_norm": 1.2146539882263698, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 14611 + }, + { + "epoch": 0.14612, + "grad_norm": 1.3111635235348604, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 14612 + }, + { + "epoch": 0.14613, + "grad_norm": 1.3431111310121864, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 14613 + }, + { + "epoch": 0.14614, + "grad_norm": 1.0556331991470425, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 14614 + }, + { + "epoch": 0.14615, + "grad_norm": 1.3210666923234133, + "learning_rate": 0.003, + "loss": 4.015, + "step": 14615 + }, + { + "epoch": 0.14616, + "grad_norm": 1.0424732096281708, + "learning_rate": 0.003, + "loss": 4.009, + "step": 14616 + }, + { + "epoch": 0.14617, + "grad_norm": 1.4222481728984127, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 14617 + }, + { + "epoch": 0.14618, + "grad_norm": 1.1671226113095667, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 14618 + }, + { + "epoch": 0.14619, + "grad_norm": 1.1566156305201265, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 14619 + }, + { + "epoch": 0.1462, + "grad_norm": 1.4435759830205142, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 14620 + }, + { + "epoch": 0.14621, + "grad_norm": 1.0040521628546581, + "learning_rate": 0.003, + "loss": 4.043, + "step": 14621 + }, + { + "epoch": 0.14622, + "grad_norm": 1.5013433657193513, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 14622 + }, + { + "epoch": 0.14623, + "grad_norm": 1.0041942566980275, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 14623 + }, + { + "epoch": 0.14624, + "grad_norm": 1.2788742166205531, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 14624 + }, + { + "epoch": 0.14625, + "grad_norm": 1.1526320122181346, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 14625 + }, + { + "epoch": 0.14626, + "grad_norm": 1.5960919046817466, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 14626 + }, + { + "epoch": 0.14627, + "grad_norm": 1.0154809815070478, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 14627 + }, + { + "epoch": 0.14628, + "grad_norm": 1.5591770055501386, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 14628 + }, + { + "epoch": 0.14629, + "grad_norm": 1.2433875029595995, + "learning_rate": 0.003, + "loss": 4.01, + "step": 14629 + }, + { + "epoch": 0.1463, + "grad_norm": 1.0446726464340053, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 14630 + }, + { + "epoch": 0.14631, + "grad_norm": 1.3196216495460884, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 14631 + }, + { + "epoch": 0.14632, + "grad_norm": 1.0795033261710623, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 14632 + }, + { + "epoch": 0.14633, + "grad_norm": 1.3754481147013444, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 14633 + }, + { + "epoch": 0.14634, + "grad_norm": 1.3076650469022453, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 14634 + }, + { + "epoch": 0.14635, + "grad_norm": 1.243730585112518, + "learning_rate": 0.003, + "loss": 4.037, + "step": 14635 + }, + { + "epoch": 0.14636, + "grad_norm": 1.2090911362187096, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14636 + }, + { + "epoch": 0.14637, + "grad_norm": 1.0881962337733553, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 14637 + }, + { + "epoch": 0.14638, + "grad_norm": 1.3315979686751431, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14638 + }, + { + "epoch": 0.14639, + "grad_norm": 1.0951279000513339, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 14639 + }, + { + "epoch": 0.1464, + "grad_norm": 1.2122185434801238, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 14640 + }, + { + "epoch": 0.14641, + "grad_norm": 1.1938786649463586, + "learning_rate": 0.003, + "loss": 4.03, + "step": 14641 + }, + { + "epoch": 0.14642, + "grad_norm": 1.1415369026720699, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 14642 + }, + { + "epoch": 0.14643, + "grad_norm": 0.9299398623994157, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 14643 + }, + { + "epoch": 0.14644, + "grad_norm": 0.8238908005739586, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 14644 + }, + { + "epoch": 0.14645, + "grad_norm": 1.0600623354228622, + "learning_rate": 0.003, + "loss": 4.028, + "step": 14645 + }, + { + "epoch": 0.14646, + "grad_norm": 1.4662081180329973, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 14646 + }, + { + "epoch": 0.14647, + "grad_norm": 1.0760830273484745, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 14647 + }, + { + "epoch": 0.14648, + "grad_norm": 1.5618154390340337, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 14648 + }, + { + "epoch": 0.14649, + "grad_norm": 1.029574143897717, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 14649 + }, + { + "epoch": 0.1465, + "grad_norm": 1.3437312355913529, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 14650 + }, + { + "epoch": 0.14651, + "grad_norm": 1.0808326315459529, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 14651 + }, + { + "epoch": 0.14652, + "grad_norm": 1.2761441979796349, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 14652 + }, + { + "epoch": 0.14653, + "grad_norm": 1.132692142274991, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 14653 + }, + { + "epoch": 0.14654, + "grad_norm": 1.3585879621138313, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14654 + }, + { + "epoch": 0.14655, + "grad_norm": 1.0463229754674632, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 14655 + }, + { + "epoch": 0.14656, + "grad_norm": 1.3918935935980463, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 14656 + }, + { + "epoch": 0.14657, + "grad_norm": 1.0312057243069037, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 14657 + }, + { + "epoch": 0.14658, + "grad_norm": 1.2350850647234748, + "learning_rate": 0.003, + "loss": 3.989, + "step": 14658 + }, + { + "epoch": 0.14659, + "grad_norm": 1.1383384238061172, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 14659 + }, + { + "epoch": 0.1466, + "grad_norm": 1.1225998777663533, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14660 + }, + { + "epoch": 0.14661, + "grad_norm": 1.1805257298545155, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 14661 + }, + { + "epoch": 0.14662, + "grad_norm": 0.9972619206686567, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 14662 + }, + { + "epoch": 0.14663, + "grad_norm": 1.3309615460438815, + "learning_rate": 0.003, + "loss": 4.043, + "step": 14663 + }, + { + "epoch": 0.14664, + "grad_norm": 1.2252668564509699, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 14664 + }, + { + "epoch": 0.14665, + "grad_norm": 1.2211356385365015, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 14665 + }, + { + "epoch": 0.14666, + "grad_norm": 1.248661926118289, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 14666 + }, + { + "epoch": 0.14667, + "grad_norm": 1.039743337030898, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 14667 + }, + { + "epoch": 0.14668, + "grad_norm": 1.3064724626176418, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 14668 + }, + { + "epoch": 0.14669, + "grad_norm": 1.0394125549913507, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 14669 + }, + { + "epoch": 0.1467, + "grad_norm": 1.523471981607469, + "learning_rate": 0.003, + "loss": 4.025, + "step": 14670 + }, + { + "epoch": 0.14671, + "grad_norm": 1.0238160913510082, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 14671 + }, + { + "epoch": 0.14672, + "grad_norm": 1.24781229708128, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 14672 + }, + { + "epoch": 0.14673, + "grad_norm": 1.1999723508752755, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14673 + }, + { + "epoch": 0.14674, + "grad_norm": 1.298555433229248, + "learning_rate": 0.003, + "loss": 3.987, + "step": 14674 + }, + { + "epoch": 0.14675, + "grad_norm": 1.1401112535442113, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 14675 + }, + { + "epoch": 0.14676, + "grad_norm": 1.2317303457429185, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 14676 + }, + { + "epoch": 0.14677, + "grad_norm": 1.1622651650359708, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 14677 + }, + { + "epoch": 0.14678, + "grad_norm": 1.456096924396937, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 14678 + }, + { + "epoch": 0.14679, + "grad_norm": 1.1842110405256305, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 14679 + }, + { + "epoch": 0.1468, + "grad_norm": 1.2673044205478405, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 14680 + }, + { + "epoch": 0.14681, + "grad_norm": 1.2099931664574737, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 14681 + }, + { + "epoch": 0.14682, + "grad_norm": 1.164083021610876, + "learning_rate": 0.003, + "loss": 4.008, + "step": 14682 + }, + { + "epoch": 0.14683, + "grad_norm": 1.2816239034400825, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 14683 + }, + { + "epoch": 0.14684, + "grad_norm": 1.1309571213745622, + "learning_rate": 0.003, + "loss": 4.009, + "step": 14684 + }, + { + "epoch": 0.14685, + "grad_norm": 1.598024848121282, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 14685 + }, + { + "epoch": 0.14686, + "grad_norm": 0.9280187236625966, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 14686 + }, + { + "epoch": 0.14687, + "grad_norm": 1.336377928291785, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 14687 + }, + { + "epoch": 0.14688, + "grad_norm": 1.2549502107529549, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 14688 + }, + { + "epoch": 0.14689, + "grad_norm": 1.3464356675156035, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 14689 + }, + { + "epoch": 0.1469, + "grad_norm": 0.9290986244236656, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 14690 + }, + { + "epoch": 0.14691, + "grad_norm": 1.13828097238937, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 14691 + }, + { + "epoch": 0.14692, + "grad_norm": 1.244377214165893, + "learning_rate": 0.003, + "loss": 4.013, + "step": 14692 + }, + { + "epoch": 0.14693, + "grad_norm": 1.061497537684576, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 14693 + }, + { + "epoch": 0.14694, + "grad_norm": 1.4475906454353726, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 14694 + }, + { + "epoch": 0.14695, + "grad_norm": 1.1298290215779774, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 14695 + }, + { + "epoch": 0.14696, + "grad_norm": 1.5073789409427754, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 14696 + }, + { + "epoch": 0.14697, + "grad_norm": 0.9979149420622156, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 14697 + }, + { + "epoch": 0.14698, + "grad_norm": 1.4211626262696988, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 14698 + }, + { + "epoch": 0.14699, + "grad_norm": 1.2299602276245312, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 14699 + }, + { + "epoch": 0.147, + "grad_norm": 1.103525583737697, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14700 + }, + { + "epoch": 0.14701, + "grad_norm": 1.5488878701382407, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 14701 + }, + { + "epoch": 0.14702, + "grad_norm": 1.1238259051215245, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 14702 + }, + { + "epoch": 0.14703, + "grad_norm": 1.182223801623779, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 14703 + }, + { + "epoch": 0.14704, + "grad_norm": 1.2729793388297799, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 14704 + }, + { + "epoch": 0.14705, + "grad_norm": 1.113098206652036, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 14705 + }, + { + "epoch": 0.14706, + "grad_norm": 1.0715062674554707, + "learning_rate": 0.003, + "loss": 4.02, + "step": 14706 + }, + { + "epoch": 0.14707, + "grad_norm": 1.351644327955927, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 14707 + }, + { + "epoch": 0.14708, + "grad_norm": 1.1156719712544556, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 14708 + }, + { + "epoch": 0.14709, + "grad_norm": 1.1937709438439554, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 14709 + }, + { + "epoch": 0.1471, + "grad_norm": 1.4057778479805412, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 14710 + }, + { + "epoch": 0.14711, + "grad_norm": 1.1272384668575615, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 14711 + }, + { + "epoch": 0.14712, + "grad_norm": 1.0489197743851302, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 14712 + }, + { + "epoch": 0.14713, + "grad_norm": 1.2933434063586542, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 14713 + }, + { + "epoch": 0.14714, + "grad_norm": 1.0627743043290785, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 14714 + }, + { + "epoch": 0.14715, + "grad_norm": 1.2665616873727155, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 14715 + }, + { + "epoch": 0.14716, + "grad_norm": 1.1546203680120108, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 14716 + }, + { + "epoch": 0.14717, + "grad_norm": 1.5435870011362298, + "learning_rate": 0.003, + "loss": 4.039, + "step": 14717 + }, + { + "epoch": 0.14718, + "grad_norm": 1.3015859295181127, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 14718 + }, + { + "epoch": 0.14719, + "grad_norm": 1.1719290752628724, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 14719 + }, + { + "epoch": 0.1472, + "grad_norm": 1.1957700973606493, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 14720 + }, + { + "epoch": 0.14721, + "grad_norm": 1.1863942728845447, + "learning_rate": 0.003, + "loss": 4.015, + "step": 14721 + }, + { + "epoch": 0.14722, + "grad_norm": 1.1910356355427554, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 14722 + }, + { + "epoch": 0.14723, + "grad_norm": 1.148093174968149, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 14723 + }, + { + "epoch": 0.14724, + "grad_norm": 1.1059163747624803, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 14724 + }, + { + "epoch": 0.14725, + "grad_norm": 1.2193482342947928, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 14725 + }, + { + "epoch": 0.14726, + "grad_norm": 1.2003680766415543, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 14726 + }, + { + "epoch": 0.14727, + "grad_norm": 1.1325797812355616, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 14727 + }, + { + "epoch": 0.14728, + "grad_norm": 1.1920454237282632, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 14728 + }, + { + "epoch": 0.14729, + "grad_norm": 1.1181121359816617, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 14729 + }, + { + "epoch": 0.1473, + "grad_norm": 1.3021322297897138, + "learning_rate": 0.003, + "loss": 4.009, + "step": 14730 + }, + { + "epoch": 0.14731, + "grad_norm": 1.3341848128938727, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 14731 + }, + { + "epoch": 0.14732, + "grad_norm": 1.1596557785781894, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 14732 + }, + { + "epoch": 0.14733, + "grad_norm": 1.182715193552188, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 14733 + }, + { + "epoch": 0.14734, + "grad_norm": 1.392854241245674, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 14734 + }, + { + "epoch": 0.14735, + "grad_norm": 1.0982923705804346, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 14735 + }, + { + "epoch": 0.14736, + "grad_norm": 1.3453311896029474, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14736 + }, + { + "epoch": 0.14737, + "grad_norm": 1.1451446830165797, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 14737 + }, + { + "epoch": 0.14738, + "grad_norm": 1.202671897629652, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 14738 + }, + { + "epoch": 0.14739, + "grad_norm": 1.3168548001667901, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 14739 + }, + { + "epoch": 0.1474, + "grad_norm": 1.1757506276149658, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14740 + }, + { + "epoch": 0.14741, + "grad_norm": 1.4512270939919778, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 14741 + }, + { + "epoch": 0.14742, + "grad_norm": 1.3569803064813937, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 14742 + }, + { + "epoch": 0.14743, + "grad_norm": 1.194524578480824, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 14743 + }, + { + "epoch": 0.14744, + "grad_norm": 1.2610976221389463, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 14744 + }, + { + "epoch": 0.14745, + "grad_norm": 0.9805340364303272, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 14745 + }, + { + "epoch": 0.14746, + "grad_norm": 1.239657032352779, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 14746 + }, + { + "epoch": 0.14747, + "grad_norm": 0.9568751883716193, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 14747 + }, + { + "epoch": 0.14748, + "grad_norm": 1.3422374066034728, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14748 + }, + { + "epoch": 0.14749, + "grad_norm": 1.2456508296319184, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 14749 + }, + { + "epoch": 0.1475, + "grad_norm": 1.2726613821317687, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 14750 + }, + { + "epoch": 0.14751, + "grad_norm": 1.1803598228688557, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 14751 + }, + { + "epoch": 0.14752, + "grad_norm": 1.3777719414878142, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 14752 + }, + { + "epoch": 0.14753, + "grad_norm": 1.4180466112113057, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 14753 + }, + { + "epoch": 0.14754, + "grad_norm": 0.8278486097470231, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 14754 + }, + { + "epoch": 0.14755, + "grad_norm": 0.9499984355175111, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 14755 + }, + { + "epoch": 0.14756, + "grad_norm": 1.3191873255352862, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 14756 + }, + { + "epoch": 0.14757, + "grad_norm": 1.1279864050152106, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 14757 + }, + { + "epoch": 0.14758, + "grad_norm": 1.112383834697207, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 14758 + }, + { + "epoch": 0.14759, + "grad_norm": 1.1386057827145057, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 14759 + }, + { + "epoch": 0.1476, + "grad_norm": 1.4423913059381863, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 14760 + }, + { + "epoch": 0.14761, + "grad_norm": 1.0955425381374513, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14761 + }, + { + "epoch": 0.14762, + "grad_norm": 1.4947533994825475, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 14762 + }, + { + "epoch": 0.14763, + "grad_norm": 1.0062595696736336, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 14763 + }, + { + "epoch": 0.14764, + "grad_norm": 1.3722162324539398, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 14764 + }, + { + "epoch": 0.14765, + "grad_norm": 1.201731727510067, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 14765 + }, + { + "epoch": 0.14766, + "grad_norm": 1.3068623774193429, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 14766 + }, + { + "epoch": 0.14767, + "grad_norm": 1.3451901636459578, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 14767 + }, + { + "epoch": 0.14768, + "grad_norm": 1.2390635797654017, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 14768 + }, + { + "epoch": 0.14769, + "grad_norm": 1.2865212311468084, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 14769 + }, + { + "epoch": 0.1477, + "grad_norm": 1.0752257596790689, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 14770 + }, + { + "epoch": 0.14771, + "grad_norm": 1.1960158251847295, + "learning_rate": 0.003, + "loss": 4.013, + "step": 14771 + }, + { + "epoch": 0.14772, + "grad_norm": 1.3336239932913452, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 14772 + }, + { + "epoch": 0.14773, + "grad_norm": 1.171080291569365, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 14773 + }, + { + "epoch": 0.14774, + "grad_norm": 1.2020832448468304, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 14774 + }, + { + "epoch": 0.14775, + "grad_norm": 1.0974795077534392, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 14775 + }, + { + "epoch": 0.14776, + "grad_norm": 1.319108091993031, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 14776 + }, + { + "epoch": 0.14777, + "grad_norm": 1.24349520064016, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 14777 + }, + { + "epoch": 0.14778, + "grad_norm": 1.3257656870019594, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 14778 + }, + { + "epoch": 0.14779, + "grad_norm": 1.0370219200672475, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 14779 + }, + { + "epoch": 0.1478, + "grad_norm": 1.7092061151710198, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 14780 + }, + { + "epoch": 0.14781, + "grad_norm": 1.1086415995649592, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 14781 + }, + { + "epoch": 0.14782, + "grad_norm": 1.1549386726359416, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 14782 + }, + { + "epoch": 0.14783, + "grad_norm": 1.0851773484666918, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 14783 + }, + { + "epoch": 0.14784, + "grad_norm": 1.3011776728231434, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 14784 + }, + { + "epoch": 0.14785, + "grad_norm": 1.3671877333509035, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 14785 + }, + { + "epoch": 0.14786, + "grad_norm": 1.0755374511587592, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 14786 + }, + { + "epoch": 0.14787, + "grad_norm": 1.2171324500253284, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 14787 + }, + { + "epoch": 0.14788, + "grad_norm": 1.0902739128743564, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 14788 + }, + { + "epoch": 0.14789, + "grad_norm": 1.186812643568498, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 14789 + }, + { + "epoch": 0.1479, + "grad_norm": 1.139087631217558, + "learning_rate": 0.003, + "loss": 4.008, + "step": 14790 + }, + { + "epoch": 0.14791, + "grad_norm": 1.1221619294261334, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 14791 + }, + { + "epoch": 0.14792, + "grad_norm": 1.3475228213640233, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 14792 + }, + { + "epoch": 0.14793, + "grad_norm": 0.9394196299688248, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14793 + }, + { + "epoch": 0.14794, + "grad_norm": 1.1896221994674032, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 14794 + }, + { + "epoch": 0.14795, + "grad_norm": 1.3234921820123722, + "learning_rate": 0.003, + "loss": 4.064, + "step": 14795 + }, + { + "epoch": 0.14796, + "grad_norm": 1.0543878759494105, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 14796 + }, + { + "epoch": 0.14797, + "grad_norm": 1.122857036841386, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14797 + }, + { + "epoch": 0.14798, + "grad_norm": 1.4259025619320995, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 14798 + }, + { + "epoch": 0.14799, + "grad_norm": 0.9954471138847973, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 14799 + }, + { + "epoch": 0.148, + "grad_norm": 1.3635514739065993, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 14800 + }, + { + "epoch": 0.14801, + "grad_norm": 1.355864793729671, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 14801 + }, + { + "epoch": 0.14802, + "grad_norm": 1.2951051954086201, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 14802 + }, + { + "epoch": 0.14803, + "grad_norm": 1.240531649391423, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 14803 + }, + { + "epoch": 0.14804, + "grad_norm": 1.1882329146439534, + "learning_rate": 0.003, + "loss": 4.003, + "step": 14804 + }, + { + "epoch": 0.14805, + "grad_norm": 1.0839561759223866, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 14805 + }, + { + "epoch": 0.14806, + "grad_norm": 1.1895202923837271, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 14806 + }, + { + "epoch": 0.14807, + "grad_norm": 1.0984721627105687, + "learning_rate": 0.003, + "loss": 3.979, + "step": 14807 + }, + { + "epoch": 0.14808, + "grad_norm": 1.5108702288765057, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 14808 + }, + { + "epoch": 0.14809, + "grad_norm": 1.0499726722241747, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 14809 + }, + { + "epoch": 0.1481, + "grad_norm": 1.2681284049610142, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 14810 + }, + { + "epoch": 0.14811, + "grad_norm": 1.1083564858942436, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 14811 + }, + { + "epoch": 0.14812, + "grad_norm": 1.2076947134347684, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 14812 + }, + { + "epoch": 0.14813, + "grad_norm": 1.307342413103559, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 14813 + }, + { + "epoch": 0.14814, + "grad_norm": 1.1750548839920394, + "learning_rate": 0.003, + "loss": 4.007, + "step": 14814 + }, + { + "epoch": 0.14815, + "grad_norm": 1.1184260700584692, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 14815 + }, + { + "epoch": 0.14816, + "grad_norm": 1.2485263669466287, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 14816 + }, + { + "epoch": 0.14817, + "grad_norm": 1.1876209717504822, + "learning_rate": 0.003, + "loss": 4.021, + "step": 14817 + }, + { + "epoch": 0.14818, + "grad_norm": 1.4994157297009993, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 14818 + }, + { + "epoch": 0.14819, + "grad_norm": 1.3463228432989796, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14819 + }, + { + "epoch": 0.1482, + "grad_norm": 1.2097675025712753, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 14820 + }, + { + "epoch": 0.14821, + "grad_norm": 1.2459321887205477, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 14821 + }, + { + "epoch": 0.14822, + "grad_norm": 1.145568372480429, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 14822 + }, + { + "epoch": 0.14823, + "grad_norm": 1.4011757991781326, + "learning_rate": 0.003, + "loss": 3.998, + "step": 14823 + }, + { + "epoch": 0.14824, + "grad_norm": 1.2026590669628292, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 14824 + }, + { + "epoch": 0.14825, + "grad_norm": 1.298554081521772, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 14825 + }, + { + "epoch": 0.14826, + "grad_norm": 1.1783814713327967, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 14826 + }, + { + "epoch": 0.14827, + "grad_norm": 1.4774961433233849, + "learning_rate": 0.003, + "loss": 4.036, + "step": 14827 + }, + { + "epoch": 0.14828, + "grad_norm": 0.9330859051005298, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 14828 + }, + { + "epoch": 0.14829, + "grad_norm": 1.6178608281109528, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 14829 + }, + { + "epoch": 0.1483, + "grad_norm": 0.9426351181235804, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 14830 + }, + { + "epoch": 0.14831, + "grad_norm": 1.1842062208065405, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 14831 + }, + { + "epoch": 0.14832, + "grad_norm": 1.168425062054937, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 14832 + }, + { + "epoch": 0.14833, + "grad_norm": 1.221455526865599, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 14833 + }, + { + "epoch": 0.14834, + "grad_norm": 1.0884537180758096, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 14834 + }, + { + "epoch": 0.14835, + "grad_norm": 1.342286263334974, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 14835 + }, + { + "epoch": 0.14836, + "grad_norm": 1.220178604711903, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 14836 + }, + { + "epoch": 0.14837, + "grad_norm": 1.2912726229956275, + "learning_rate": 0.003, + "loss": 4.025, + "step": 14837 + }, + { + "epoch": 0.14838, + "grad_norm": 1.0858951962638583, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 14838 + }, + { + "epoch": 0.14839, + "grad_norm": 1.3854215262077978, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 14839 + }, + { + "epoch": 0.1484, + "grad_norm": 1.3035567966579602, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 14840 + }, + { + "epoch": 0.14841, + "grad_norm": 1.026277131753848, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 14841 + }, + { + "epoch": 0.14842, + "grad_norm": 1.3953361705547809, + "learning_rate": 0.003, + "loss": 4.015, + "step": 14842 + }, + { + "epoch": 0.14843, + "grad_norm": 1.1548116580406775, + "learning_rate": 0.003, + "loss": 4.015, + "step": 14843 + }, + { + "epoch": 0.14844, + "grad_norm": 1.1248324434636736, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 14844 + }, + { + "epoch": 0.14845, + "grad_norm": 1.3108243504310484, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 14845 + }, + { + "epoch": 0.14846, + "grad_norm": 1.2222485880266103, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 14846 + }, + { + "epoch": 0.14847, + "grad_norm": 1.1495284914620727, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 14847 + }, + { + "epoch": 0.14848, + "grad_norm": 1.2515593067163204, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 14848 + }, + { + "epoch": 0.14849, + "grad_norm": 1.1804797578470736, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 14849 + }, + { + "epoch": 0.1485, + "grad_norm": 1.167497533650182, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 14850 + }, + { + "epoch": 0.14851, + "grad_norm": 1.052082466564659, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 14851 + }, + { + "epoch": 0.14852, + "grad_norm": 1.4757345167866347, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 14852 + }, + { + "epoch": 0.14853, + "grad_norm": 1.1738947991619102, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 14853 + }, + { + "epoch": 0.14854, + "grad_norm": 1.3960930732542705, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 14854 + }, + { + "epoch": 0.14855, + "grad_norm": 1.1977614959325054, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 14855 + }, + { + "epoch": 0.14856, + "grad_norm": 1.2826098837508362, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 14856 + }, + { + "epoch": 0.14857, + "grad_norm": 1.2771632194462115, + "learning_rate": 0.003, + "loss": 4.038, + "step": 14857 + }, + { + "epoch": 0.14858, + "grad_norm": 1.2649138401002105, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 14858 + }, + { + "epoch": 0.14859, + "grad_norm": 1.2621133328507186, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14859 + }, + { + "epoch": 0.1486, + "grad_norm": 1.165204313150529, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 14860 + }, + { + "epoch": 0.14861, + "grad_norm": 1.4595143374205735, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 14861 + }, + { + "epoch": 0.14862, + "grad_norm": 1.313118096897284, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 14862 + }, + { + "epoch": 0.14863, + "grad_norm": 0.9960033788648959, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 14863 + }, + { + "epoch": 0.14864, + "grad_norm": 1.3999675969687877, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 14864 + }, + { + "epoch": 0.14865, + "grad_norm": 0.971526221637175, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 14865 + }, + { + "epoch": 0.14866, + "grad_norm": 1.5885001965903711, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 14866 + }, + { + "epoch": 0.14867, + "grad_norm": 0.8794496752242735, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 14867 + }, + { + "epoch": 0.14868, + "grad_norm": 1.1457341472984988, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14868 + }, + { + "epoch": 0.14869, + "grad_norm": 1.2460462196880944, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 14869 + }, + { + "epoch": 0.1487, + "grad_norm": 1.3004944633604691, + "learning_rate": 0.003, + "loss": 4.014, + "step": 14870 + }, + { + "epoch": 0.14871, + "grad_norm": 1.0737833889267785, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 14871 + }, + { + "epoch": 0.14872, + "grad_norm": 1.1905628483988502, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 14872 + }, + { + "epoch": 0.14873, + "grad_norm": 1.2150252779319972, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 14873 + }, + { + "epoch": 0.14874, + "grad_norm": 1.413716137598804, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 14874 + }, + { + "epoch": 0.14875, + "grad_norm": 1.2319027017276702, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 14875 + }, + { + "epoch": 0.14876, + "grad_norm": 1.1590811670007173, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 14876 + }, + { + "epoch": 0.14877, + "grad_norm": 1.1502595888784863, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 14877 + }, + { + "epoch": 0.14878, + "grad_norm": 1.135405812618134, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14878 + }, + { + "epoch": 0.14879, + "grad_norm": 0.9817877807968799, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 14879 + }, + { + "epoch": 0.1488, + "grad_norm": 1.1457316336220291, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 14880 + }, + { + "epoch": 0.14881, + "grad_norm": 1.120078496043829, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 14881 + }, + { + "epoch": 0.14882, + "grad_norm": 1.3998521551348393, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 14882 + }, + { + "epoch": 0.14883, + "grad_norm": 1.2585695115298041, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 14883 + }, + { + "epoch": 0.14884, + "grad_norm": 1.1596322530573377, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14884 + }, + { + "epoch": 0.14885, + "grad_norm": 1.2900109792885086, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 14885 + }, + { + "epoch": 0.14886, + "grad_norm": 1.0941650068530546, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 14886 + }, + { + "epoch": 0.14887, + "grad_norm": 1.2701767655262914, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 14887 + }, + { + "epoch": 0.14888, + "grad_norm": 1.3785028945271498, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 14888 + }, + { + "epoch": 0.14889, + "grad_norm": 1.1640558071841207, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 14889 + }, + { + "epoch": 0.1489, + "grad_norm": 1.2350016360468847, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 14890 + }, + { + "epoch": 0.14891, + "grad_norm": 1.1909812740009393, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 14891 + }, + { + "epoch": 0.14892, + "grad_norm": 1.245236229330982, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 14892 + }, + { + "epoch": 0.14893, + "grad_norm": 1.0319952693329486, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 14893 + }, + { + "epoch": 0.14894, + "grad_norm": 1.404395060703985, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 14894 + }, + { + "epoch": 0.14895, + "grad_norm": 1.3312321228312276, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 14895 + }, + { + "epoch": 0.14896, + "grad_norm": 1.4031247593198042, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 14896 + }, + { + "epoch": 0.14897, + "grad_norm": 0.9251746593633249, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 14897 + }, + { + "epoch": 0.14898, + "grad_norm": 1.1089796838097807, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 14898 + }, + { + "epoch": 0.14899, + "grad_norm": 1.1160083080503167, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 14899 + }, + { + "epoch": 0.149, + "grad_norm": 1.3488201450168495, + "learning_rate": 0.003, + "loss": 4.02, + "step": 14900 + }, + { + "epoch": 0.14901, + "grad_norm": 1.4817746969596457, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 14901 + }, + { + "epoch": 0.14902, + "grad_norm": 0.9974601034053736, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 14902 + }, + { + "epoch": 0.14903, + "grad_norm": 1.5739684402092466, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 14903 + }, + { + "epoch": 0.14904, + "grad_norm": 1.2032902211761658, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 14904 + }, + { + "epoch": 0.14905, + "grad_norm": 1.1779051842351325, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 14905 + }, + { + "epoch": 0.14906, + "grad_norm": 1.304753964643637, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 14906 + }, + { + "epoch": 0.14907, + "grad_norm": 1.1017996705146436, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 14907 + }, + { + "epoch": 0.14908, + "grad_norm": 1.413055102243228, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 14908 + }, + { + "epoch": 0.14909, + "grad_norm": 1.1993242494244376, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14909 + }, + { + "epoch": 0.1491, + "grad_norm": 1.3960283841416354, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14910 + }, + { + "epoch": 0.14911, + "grad_norm": 0.9131043857587414, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 14911 + }, + { + "epoch": 0.14912, + "grad_norm": 1.251331191866348, + "learning_rate": 0.003, + "loss": 4.035, + "step": 14912 + }, + { + "epoch": 0.14913, + "grad_norm": 1.2244501109261372, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 14913 + }, + { + "epoch": 0.14914, + "grad_norm": 1.2586212474739633, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 14914 + }, + { + "epoch": 0.14915, + "grad_norm": 1.0999447784539422, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 14915 + }, + { + "epoch": 0.14916, + "grad_norm": 1.5530642038910838, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 14916 + }, + { + "epoch": 0.14917, + "grad_norm": 1.0477200805793776, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 14917 + }, + { + "epoch": 0.14918, + "grad_norm": 1.4866370403962696, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 14918 + }, + { + "epoch": 0.14919, + "grad_norm": 1.1401361636789986, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 14919 + }, + { + "epoch": 0.1492, + "grad_norm": 1.3919223969013914, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 14920 + }, + { + "epoch": 0.14921, + "grad_norm": 1.2378662180360431, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 14921 + }, + { + "epoch": 0.14922, + "grad_norm": 1.3964563554545333, + "learning_rate": 0.003, + "loss": 4.023, + "step": 14922 + }, + { + "epoch": 0.14923, + "grad_norm": 1.3434961179490488, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 14923 + }, + { + "epoch": 0.14924, + "grad_norm": 1.1174456729435764, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 14924 + }, + { + "epoch": 0.14925, + "grad_norm": 1.3257139886781153, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 14925 + }, + { + "epoch": 0.14926, + "grad_norm": 1.1451098945090743, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 14926 + }, + { + "epoch": 0.14927, + "grad_norm": 1.460942215878829, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 14927 + }, + { + "epoch": 0.14928, + "grad_norm": 1.0171135357041543, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 14928 + }, + { + "epoch": 0.14929, + "grad_norm": 1.2490868086048281, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 14929 + }, + { + "epoch": 0.1493, + "grad_norm": 0.9820383576630475, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 14930 + }, + { + "epoch": 0.14931, + "grad_norm": 1.4240111064662544, + "learning_rate": 0.003, + "loss": 4.016, + "step": 14931 + }, + { + "epoch": 0.14932, + "grad_norm": 1.0312709532553408, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14932 + }, + { + "epoch": 0.14933, + "grad_norm": 1.3033329876881157, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 14933 + }, + { + "epoch": 0.14934, + "grad_norm": 1.1022488175504013, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 14934 + }, + { + "epoch": 0.14935, + "grad_norm": 1.259008325842684, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 14935 + }, + { + "epoch": 0.14936, + "grad_norm": 1.2953376945757966, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 14936 + }, + { + "epoch": 0.14937, + "grad_norm": 0.8954211650418383, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 14937 + }, + { + "epoch": 0.14938, + "grad_norm": 1.108314327570641, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 14938 + }, + { + "epoch": 0.14939, + "grad_norm": 1.5431478241610768, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 14939 + }, + { + "epoch": 0.1494, + "grad_norm": 0.9261299070131537, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 14940 + }, + { + "epoch": 0.14941, + "grad_norm": 1.3487899537436916, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 14941 + }, + { + "epoch": 0.14942, + "grad_norm": 1.0720868304846356, + "learning_rate": 0.003, + "loss": 3.997, + "step": 14942 + }, + { + "epoch": 0.14943, + "grad_norm": 1.1356899271035579, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14943 + }, + { + "epoch": 0.14944, + "grad_norm": 1.159036689495631, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 14944 + }, + { + "epoch": 0.14945, + "grad_norm": 1.294910603968105, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 14945 + }, + { + "epoch": 0.14946, + "grad_norm": 1.2310211532261401, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 14946 + }, + { + "epoch": 0.14947, + "grad_norm": 1.147713769303728, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 14947 + }, + { + "epoch": 0.14948, + "grad_norm": 1.230167779841433, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 14948 + }, + { + "epoch": 0.14949, + "grad_norm": 1.2306050786117049, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 14949 + }, + { + "epoch": 0.1495, + "grad_norm": 1.3135866891956622, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 14950 + }, + { + "epoch": 0.14951, + "grad_norm": 1.3623435968877635, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 14951 + }, + { + "epoch": 0.14952, + "grad_norm": 1.152662938484965, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 14952 + }, + { + "epoch": 0.14953, + "grad_norm": 1.289088875479219, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 14953 + }, + { + "epoch": 0.14954, + "grad_norm": 1.1215923060881257, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 14954 + }, + { + "epoch": 0.14955, + "grad_norm": 1.3221633165675728, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 14955 + }, + { + "epoch": 0.14956, + "grad_norm": 0.9652269514638793, + "learning_rate": 0.003, + "loss": 3.997, + "step": 14956 + }, + { + "epoch": 0.14957, + "grad_norm": 1.2836725479916864, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 14957 + }, + { + "epoch": 0.14958, + "grad_norm": 1.222776077178544, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 14958 + }, + { + "epoch": 0.14959, + "grad_norm": 1.3037456196499888, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 14959 + }, + { + "epoch": 0.1496, + "grad_norm": 1.2338554444148002, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 14960 + }, + { + "epoch": 0.14961, + "grad_norm": 1.057582098078062, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 14961 + }, + { + "epoch": 0.14962, + "grad_norm": 1.2566371433084496, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 14962 + }, + { + "epoch": 0.14963, + "grad_norm": 1.2417135314919132, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 14963 + }, + { + "epoch": 0.14964, + "grad_norm": 1.240855453420091, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 14964 + }, + { + "epoch": 0.14965, + "grad_norm": 1.2042763426088086, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 14965 + }, + { + "epoch": 0.14966, + "grad_norm": 1.4578607611652201, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 14966 + }, + { + "epoch": 0.14967, + "grad_norm": 1.173609245954742, + "learning_rate": 0.003, + "loss": 4.039, + "step": 14967 + }, + { + "epoch": 0.14968, + "grad_norm": 1.0947873844270783, + "learning_rate": 0.003, + "loss": 4.014, + "step": 14968 + }, + { + "epoch": 0.14969, + "grad_norm": 1.3683915784044167, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 14969 + }, + { + "epoch": 0.1497, + "grad_norm": 1.2199880471450995, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 14970 + }, + { + "epoch": 0.14971, + "grad_norm": 1.309851887266059, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 14971 + }, + { + "epoch": 0.14972, + "grad_norm": 1.1758886914043591, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 14972 + }, + { + "epoch": 0.14973, + "grad_norm": 1.3609476281993964, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 14973 + }, + { + "epoch": 0.14974, + "grad_norm": 0.9816117678678208, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 14974 + }, + { + "epoch": 0.14975, + "grad_norm": 1.4053278904134376, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 14975 + }, + { + "epoch": 0.14976, + "grad_norm": 1.3177315280510327, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 14976 + }, + { + "epoch": 0.14977, + "grad_norm": 1.321177287660058, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 14977 + }, + { + "epoch": 0.14978, + "grad_norm": 1.1995771594106088, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 14978 + }, + { + "epoch": 0.14979, + "grad_norm": 1.1852421784022913, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 14979 + }, + { + "epoch": 0.1498, + "grad_norm": 1.0344875601882786, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 14980 + }, + { + "epoch": 0.14981, + "grad_norm": 1.218236211138091, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 14981 + }, + { + "epoch": 0.14982, + "grad_norm": 1.0531872899673196, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 14982 + }, + { + "epoch": 0.14983, + "grad_norm": 1.3701361125306657, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 14983 + }, + { + "epoch": 0.14984, + "grad_norm": 1.1561564936517743, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 14984 + }, + { + "epoch": 0.14985, + "grad_norm": 1.2203615274256243, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 14985 + }, + { + "epoch": 0.14986, + "grad_norm": 1.3589539058043363, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 14986 + }, + { + "epoch": 0.14987, + "grad_norm": 1.007262450095289, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14987 + }, + { + "epoch": 0.14988, + "grad_norm": 1.4869886265733574, + "learning_rate": 0.003, + "loss": 4.002, + "step": 14988 + }, + { + "epoch": 0.14989, + "grad_norm": 1.0009057794425513, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 14989 + }, + { + "epoch": 0.1499, + "grad_norm": 1.389189092010671, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14990 + }, + { + "epoch": 0.14991, + "grad_norm": 1.152524627524547, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 14991 + }, + { + "epoch": 0.14992, + "grad_norm": 1.1099450346476132, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 14992 + }, + { + "epoch": 0.14993, + "grad_norm": 1.3414137567732916, + "learning_rate": 0.003, + "loss": 4.008, + "step": 14993 + }, + { + "epoch": 0.14994, + "grad_norm": 1.1093458689260562, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 14994 + }, + { + "epoch": 0.14995, + "grad_norm": 1.3059990531641175, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 14995 + }, + { + "epoch": 0.14996, + "grad_norm": 1.2710449589157764, + "learning_rate": 0.003, + "loss": 4.014, + "step": 14996 + }, + { + "epoch": 0.14997, + "grad_norm": 1.324001036249548, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 14997 + }, + { + "epoch": 0.14998, + "grad_norm": 1.5132413834470855, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 14998 + }, + { + "epoch": 0.14999, + "grad_norm": 1.1022008116316366, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 14999 + }, + { + "epoch": 0.15, + "grad_norm": 1.2511547133707417, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 15000 + }, + { + "epoch": 0.15001, + "grad_norm": 1.0834107812897231, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 15001 + }, + { + "epoch": 0.15002, + "grad_norm": 1.4583003026965216, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 15002 + }, + { + "epoch": 0.15003, + "grad_norm": 1.0164623980748966, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 15003 + }, + { + "epoch": 0.15004, + "grad_norm": 1.343262825216341, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 15004 + }, + { + "epoch": 0.15005, + "grad_norm": 1.1370870689830004, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 15005 + }, + { + "epoch": 0.15006, + "grad_norm": 1.3223233899220281, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 15006 + }, + { + "epoch": 0.15007, + "grad_norm": 1.0370331143992404, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 15007 + }, + { + "epoch": 0.15008, + "grad_norm": 1.4445294117832153, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 15008 + }, + { + "epoch": 0.15009, + "grad_norm": 1.3518029276829822, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 15009 + }, + { + "epoch": 0.1501, + "grad_norm": 1.2309957046300406, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 15010 + }, + { + "epoch": 0.15011, + "grad_norm": 1.27589356901691, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 15011 + }, + { + "epoch": 0.15012, + "grad_norm": 1.2144644609278672, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 15012 + }, + { + "epoch": 0.15013, + "grad_norm": 1.1501449654053362, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 15013 + }, + { + "epoch": 0.15014, + "grad_norm": 1.6288277390725783, + "learning_rate": 0.003, + "loss": 4.004, + "step": 15014 + }, + { + "epoch": 0.15015, + "grad_norm": 0.9856979647827699, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 15015 + }, + { + "epoch": 0.15016, + "grad_norm": 1.4520037698240609, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 15016 + }, + { + "epoch": 0.15017, + "grad_norm": 1.040970103974602, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 15017 + }, + { + "epoch": 0.15018, + "grad_norm": 1.2634763102058675, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 15018 + }, + { + "epoch": 0.15019, + "grad_norm": 1.0772080110580602, + "learning_rate": 0.003, + "loss": 4.017, + "step": 15019 + }, + { + "epoch": 0.1502, + "grad_norm": 1.2687453816983203, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 15020 + }, + { + "epoch": 0.15021, + "grad_norm": 1.2614304451140905, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 15021 + }, + { + "epoch": 0.15022, + "grad_norm": 1.176662629932817, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 15022 + }, + { + "epoch": 0.15023, + "grad_norm": 1.3364690597724231, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 15023 + }, + { + "epoch": 0.15024, + "grad_norm": 0.9801971142283072, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 15024 + }, + { + "epoch": 0.15025, + "grad_norm": 1.290867464177766, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 15025 + }, + { + "epoch": 0.15026, + "grad_norm": 1.5256383410336505, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 15026 + }, + { + "epoch": 0.15027, + "grad_norm": 0.9554157287977645, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 15027 + }, + { + "epoch": 0.15028, + "grad_norm": 1.2344367686432, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 15028 + }, + { + "epoch": 0.15029, + "grad_norm": 1.058874736871547, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 15029 + }, + { + "epoch": 0.1503, + "grad_norm": 1.4628155784648897, + "learning_rate": 0.003, + "loss": 3.986, + "step": 15030 + }, + { + "epoch": 0.15031, + "grad_norm": 1.000440083099698, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 15031 + }, + { + "epoch": 0.15032, + "grad_norm": 1.6221985468900189, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 15032 + }, + { + "epoch": 0.15033, + "grad_norm": 1.0593767482313574, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 15033 + }, + { + "epoch": 0.15034, + "grad_norm": 1.3033259842482872, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 15034 + }, + { + "epoch": 0.15035, + "grad_norm": 1.271582243810947, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 15035 + }, + { + "epoch": 0.15036, + "grad_norm": 1.2504946374764305, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 15036 + }, + { + "epoch": 0.15037, + "grad_norm": 1.1392408868055681, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 15037 + }, + { + "epoch": 0.15038, + "grad_norm": 1.2827374034740893, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 15038 + }, + { + "epoch": 0.15039, + "grad_norm": 1.097446245620214, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 15039 + }, + { + "epoch": 0.1504, + "grad_norm": 1.3790289652472876, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 15040 + }, + { + "epoch": 0.15041, + "grad_norm": 1.2343541743317246, + "learning_rate": 0.003, + "loss": 4.061, + "step": 15041 + }, + { + "epoch": 0.15042, + "grad_norm": 1.3551423323541314, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 15042 + }, + { + "epoch": 0.15043, + "grad_norm": 1.0136000020055114, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 15043 + }, + { + "epoch": 0.15044, + "grad_norm": 1.386402889011217, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 15044 + }, + { + "epoch": 0.15045, + "grad_norm": 1.182671672958449, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 15045 + }, + { + "epoch": 0.15046, + "grad_norm": 1.1943280336924613, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 15046 + }, + { + "epoch": 0.15047, + "grad_norm": 1.067087083299192, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 15047 + }, + { + "epoch": 0.15048, + "grad_norm": 1.2101321504804678, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 15048 + }, + { + "epoch": 0.15049, + "grad_norm": 1.1918369767211583, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 15049 + }, + { + "epoch": 0.1505, + "grad_norm": 1.1865037174439503, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 15050 + }, + { + "epoch": 0.15051, + "grad_norm": 1.307294713323573, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 15051 + }, + { + "epoch": 0.15052, + "grad_norm": 1.0843302266527004, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 15052 + }, + { + "epoch": 0.15053, + "grad_norm": 1.368616511764391, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 15053 + }, + { + "epoch": 0.15054, + "grad_norm": 1.1503707138668346, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 15054 + }, + { + "epoch": 0.15055, + "grad_norm": 1.24232930332824, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 15055 + }, + { + "epoch": 0.15056, + "grad_norm": 1.2210548505793515, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 15056 + }, + { + "epoch": 0.15057, + "grad_norm": 1.2804783401997903, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 15057 + }, + { + "epoch": 0.15058, + "grad_norm": 1.1121947561924939, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 15058 + }, + { + "epoch": 0.15059, + "grad_norm": 1.2666290669327305, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 15059 + }, + { + "epoch": 0.1506, + "grad_norm": 1.1033006586389937, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 15060 + }, + { + "epoch": 0.15061, + "grad_norm": 1.3958301674075322, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 15061 + }, + { + "epoch": 0.15062, + "grad_norm": 1.1571985815771526, + "learning_rate": 0.003, + "loss": 4.023, + "step": 15062 + }, + { + "epoch": 0.15063, + "grad_norm": 1.1693679449700813, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 15063 + }, + { + "epoch": 0.15064, + "grad_norm": 1.479241432093687, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 15064 + }, + { + "epoch": 0.15065, + "grad_norm": 1.2611691539565617, + "learning_rate": 0.003, + "loss": 4.013, + "step": 15065 + }, + { + "epoch": 0.15066, + "grad_norm": 1.354519508848747, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 15066 + }, + { + "epoch": 0.15067, + "grad_norm": 1.1264546163898423, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 15067 + }, + { + "epoch": 0.15068, + "grad_norm": 1.0276693408258029, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 15068 + }, + { + "epoch": 0.15069, + "grad_norm": 1.2274172626033997, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 15069 + }, + { + "epoch": 0.1507, + "grad_norm": 1.104144935418708, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 15070 + }, + { + "epoch": 0.15071, + "grad_norm": 1.3419237823732055, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 15071 + }, + { + "epoch": 0.15072, + "grad_norm": 1.3650899614680156, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 15072 + }, + { + "epoch": 0.15073, + "grad_norm": 1.091256776156269, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 15073 + }, + { + "epoch": 0.15074, + "grad_norm": 1.1837415652261045, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 15074 + }, + { + "epoch": 0.15075, + "grad_norm": 1.070345352497782, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 15075 + }, + { + "epoch": 0.15076, + "grad_norm": 1.3670014269058248, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 15076 + }, + { + "epoch": 0.15077, + "grad_norm": 1.049884830296458, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 15077 + }, + { + "epoch": 0.15078, + "grad_norm": 1.3611626126884704, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 15078 + }, + { + "epoch": 0.15079, + "grad_norm": 1.049809942194643, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 15079 + }, + { + "epoch": 0.1508, + "grad_norm": 1.4038742234145964, + "learning_rate": 0.003, + "loss": 4.038, + "step": 15080 + }, + { + "epoch": 0.15081, + "grad_norm": 1.060595492700757, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 15081 + }, + { + "epoch": 0.15082, + "grad_norm": 1.3183300499048776, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 15082 + }, + { + "epoch": 0.15083, + "grad_norm": 1.076680798390479, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 15083 + }, + { + "epoch": 0.15084, + "grad_norm": 1.357153988614293, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 15084 + }, + { + "epoch": 0.15085, + "grad_norm": 1.0042635001063007, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 15085 + }, + { + "epoch": 0.15086, + "grad_norm": 1.3771181358622575, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 15086 + }, + { + "epoch": 0.15087, + "grad_norm": 1.147283702797723, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 15087 + }, + { + "epoch": 0.15088, + "grad_norm": 1.2341217391719035, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 15088 + }, + { + "epoch": 0.15089, + "grad_norm": 1.3516384833027888, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 15089 + }, + { + "epoch": 0.1509, + "grad_norm": 1.2292463982304918, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 15090 + }, + { + "epoch": 0.15091, + "grad_norm": 1.3564703702419483, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 15091 + }, + { + "epoch": 0.15092, + "grad_norm": 1.112792931335097, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 15092 + }, + { + "epoch": 0.15093, + "grad_norm": 1.3136370078155426, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 15093 + }, + { + "epoch": 0.15094, + "grad_norm": 1.221216042533791, + "learning_rate": 0.003, + "loss": 4.017, + "step": 15094 + }, + { + "epoch": 0.15095, + "grad_norm": 1.2145692979169989, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 15095 + }, + { + "epoch": 0.15096, + "grad_norm": 1.3465204947883336, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 15096 + }, + { + "epoch": 0.15097, + "grad_norm": 0.9430128887268873, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 15097 + }, + { + "epoch": 0.15098, + "grad_norm": 1.2901837174572874, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 15098 + }, + { + "epoch": 0.15099, + "grad_norm": 1.165445295507976, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 15099 + }, + { + "epoch": 0.151, + "grad_norm": 1.3232825503918688, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 15100 + }, + { + "epoch": 0.15101, + "grad_norm": 1.132837865211714, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 15101 + }, + { + "epoch": 0.15102, + "grad_norm": 1.241982532953572, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 15102 + }, + { + "epoch": 0.15103, + "grad_norm": 1.2383420097536164, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 15103 + }, + { + "epoch": 0.15104, + "grad_norm": 1.262949064967333, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 15104 + }, + { + "epoch": 0.15105, + "grad_norm": 1.513535550351292, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 15105 + }, + { + "epoch": 0.15106, + "grad_norm": 1.257750614192443, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 15106 + }, + { + "epoch": 0.15107, + "grad_norm": 1.116147918666188, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 15107 + }, + { + "epoch": 0.15108, + "grad_norm": 1.189241053979111, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 15108 + }, + { + "epoch": 0.15109, + "grad_norm": 1.1349573681696845, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 15109 + }, + { + "epoch": 0.1511, + "grad_norm": 1.272700322893393, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 15110 + }, + { + "epoch": 0.15111, + "grad_norm": 1.2802884196351139, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 15111 + }, + { + "epoch": 0.15112, + "grad_norm": 1.2370215688402926, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 15112 + }, + { + "epoch": 0.15113, + "grad_norm": 1.2016811206691678, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 15113 + }, + { + "epoch": 0.15114, + "grad_norm": 1.1700164562301176, + "learning_rate": 0.003, + "loss": 3.995, + "step": 15114 + }, + { + "epoch": 0.15115, + "grad_norm": 1.2068934294145686, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 15115 + }, + { + "epoch": 0.15116, + "grad_norm": 1.1332474111238924, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 15116 + }, + { + "epoch": 0.15117, + "grad_norm": 1.2410907164496658, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 15117 + }, + { + "epoch": 0.15118, + "grad_norm": 1.273940139764871, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 15118 + }, + { + "epoch": 0.15119, + "grad_norm": 1.367244553089254, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 15119 + }, + { + "epoch": 0.1512, + "grad_norm": 1.2379281616283513, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 15120 + }, + { + "epoch": 0.15121, + "grad_norm": 1.4274532742144836, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 15121 + }, + { + "epoch": 0.15122, + "grad_norm": 0.8765441414346433, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 15122 + }, + { + "epoch": 0.15123, + "grad_norm": 1.0869158342956753, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 15123 + }, + { + "epoch": 0.15124, + "grad_norm": 1.3050167137936268, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 15124 + }, + { + "epoch": 0.15125, + "grad_norm": 1.1404648288786916, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 15125 + }, + { + "epoch": 0.15126, + "grad_norm": 1.4148215055467332, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 15126 + }, + { + "epoch": 0.15127, + "grad_norm": 1.1147866554659929, + "learning_rate": 0.003, + "loss": 4.026, + "step": 15127 + }, + { + "epoch": 0.15128, + "grad_norm": 1.2610087486392272, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 15128 + }, + { + "epoch": 0.15129, + "grad_norm": 0.9286406884499967, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 15129 + }, + { + "epoch": 0.1513, + "grad_norm": 1.2832639151976415, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 15130 + }, + { + "epoch": 0.15131, + "grad_norm": 1.0621449933845621, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15131 + }, + { + "epoch": 0.15132, + "grad_norm": 1.4334878759851153, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 15132 + }, + { + "epoch": 0.15133, + "grad_norm": 1.3105529744462483, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 15133 + }, + { + "epoch": 0.15134, + "grad_norm": 1.0549616111045212, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 15134 + }, + { + "epoch": 0.15135, + "grad_norm": 1.286201041219918, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 15135 + }, + { + "epoch": 0.15136, + "grad_norm": 1.2613815297065765, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 15136 + }, + { + "epoch": 0.15137, + "grad_norm": 1.3049514329532583, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 15137 + }, + { + "epoch": 0.15138, + "grad_norm": 1.2523480185780507, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 15138 + }, + { + "epoch": 0.15139, + "grad_norm": 1.6897379703811235, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 15139 + }, + { + "epoch": 0.1514, + "grad_norm": 1.2114299946762168, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 15140 + }, + { + "epoch": 0.15141, + "grad_norm": 1.3552575020562225, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 15141 + }, + { + "epoch": 0.15142, + "grad_norm": 1.0988214288233396, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 15142 + }, + { + "epoch": 0.15143, + "grad_norm": 1.400017500835591, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 15143 + }, + { + "epoch": 0.15144, + "grad_norm": 1.294718444209177, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 15144 + }, + { + "epoch": 0.15145, + "grad_norm": 1.264073306861436, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 15145 + }, + { + "epoch": 0.15146, + "grad_norm": 1.147161227195975, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 15146 + }, + { + "epoch": 0.15147, + "grad_norm": 1.4421115859568234, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 15147 + }, + { + "epoch": 0.15148, + "grad_norm": 1.0578395740545363, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 15148 + }, + { + "epoch": 0.15149, + "grad_norm": 1.2865955220632557, + "learning_rate": 0.003, + "loss": 4.035, + "step": 15149 + }, + { + "epoch": 0.1515, + "grad_norm": 1.021024571309487, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 15150 + }, + { + "epoch": 0.15151, + "grad_norm": 1.5256734661903013, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 15151 + }, + { + "epoch": 0.15152, + "grad_norm": 1.074404522380423, + "learning_rate": 0.003, + "loss": 4.055, + "step": 15152 + }, + { + "epoch": 0.15153, + "grad_norm": 1.3833997517816128, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 15153 + }, + { + "epoch": 0.15154, + "grad_norm": 1.0423269119897456, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 15154 + }, + { + "epoch": 0.15155, + "grad_norm": 1.2210503757117748, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15155 + }, + { + "epoch": 0.15156, + "grad_norm": 1.108145226130091, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 15156 + }, + { + "epoch": 0.15157, + "grad_norm": 1.2422893744328773, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 15157 + }, + { + "epoch": 0.15158, + "grad_norm": 1.1585359463147444, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 15158 + }, + { + "epoch": 0.15159, + "grad_norm": 1.3052493913554555, + "learning_rate": 0.003, + "loss": 4.029, + "step": 15159 + }, + { + "epoch": 0.1516, + "grad_norm": 1.0391936207685226, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 15160 + }, + { + "epoch": 0.15161, + "grad_norm": 1.3719432821989905, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 15161 + }, + { + "epoch": 0.15162, + "grad_norm": 1.2060382045199585, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 15162 + }, + { + "epoch": 0.15163, + "grad_norm": 1.2938521061808144, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 15163 + }, + { + "epoch": 0.15164, + "grad_norm": 1.2728919704370556, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 15164 + }, + { + "epoch": 0.15165, + "grad_norm": 1.2976291980902035, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 15165 + }, + { + "epoch": 0.15166, + "grad_norm": 1.1318885985271303, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 15166 + }, + { + "epoch": 0.15167, + "grad_norm": 1.1218973473108103, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 15167 + }, + { + "epoch": 0.15168, + "grad_norm": 1.2558430257149893, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 15168 + }, + { + "epoch": 0.15169, + "grad_norm": 1.1632066368507694, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 15169 + }, + { + "epoch": 0.1517, + "grad_norm": 1.512747349596697, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 15170 + }, + { + "epoch": 0.15171, + "grad_norm": 1.1607688250338546, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 15171 + }, + { + "epoch": 0.15172, + "grad_norm": 1.5669471842154912, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 15172 + }, + { + "epoch": 0.15173, + "grad_norm": 0.9690746984375851, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 15173 + }, + { + "epoch": 0.15174, + "grad_norm": 1.424525009288961, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 15174 + }, + { + "epoch": 0.15175, + "grad_norm": 1.0884188106452057, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15175 + }, + { + "epoch": 0.15176, + "grad_norm": 1.3406032412403708, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 15176 + }, + { + "epoch": 0.15177, + "grad_norm": 1.2010227603841026, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 15177 + }, + { + "epoch": 0.15178, + "grad_norm": 1.1403832764384894, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 15178 + }, + { + "epoch": 0.15179, + "grad_norm": 1.3656623173642093, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 15179 + }, + { + "epoch": 0.1518, + "grad_norm": 0.9446342092378506, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 15180 + }, + { + "epoch": 0.15181, + "grad_norm": 1.1632519133270702, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 15181 + }, + { + "epoch": 0.15182, + "grad_norm": 1.2515943146416864, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 15182 + }, + { + "epoch": 0.15183, + "grad_norm": 1.3866824702413552, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 15183 + }, + { + "epoch": 0.15184, + "grad_norm": 1.1669861813371163, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 15184 + }, + { + "epoch": 0.15185, + "grad_norm": 1.1595091466120568, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 15185 + }, + { + "epoch": 0.15186, + "grad_norm": 1.3859410529014762, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 15186 + }, + { + "epoch": 0.15187, + "grad_norm": 1.3425423501975218, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 15187 + }, + { + "epoch": 0.15188, + "grad_norm": 1.0608433290080848, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 15188 + }, + { + "epoch": 0.15189, + "grad_norm": 1.0314082381097578, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 15189 + }, + { + "epoch": 0.1519, + "grad_norm": 1.2723553619786312, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 15190 + }, + { + "epoch": 0.15191, + "grad_norm": 1.142633684616142, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 15191 + }, + { + "epoch": 0.15192, + "grad_norm": 1.1497651283192816, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15192 + }, + { + "epoch": 0.15193, + "grad_norm": 1.141694588201836, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 15193 + }, + { + "epoch": 0.15194, + "grad_norm": 1.3196754641868844, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 15194 + }, + { + "epoch": 0.15195, + "grad_norm": 1.4382441258392895, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 15195 + }, + { + "epoch": 0.15196, + "grad_norm": 1.2905753364700843, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 15196 + }, + { + "epoch": 0.15197, + "grad_norm": 1.3322175812610237, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15197 + }, + { + "epoch": 0.15198, + "grad_norm": 1.2828552826700481, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 15198 + }, + { + "epoch": 0.15199, + "grad_norm": 1.1331332190523502, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 15199 + }, + { + "epoch": 0.152, + "grad_norm": 1.2949163051420256, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 15200 + }, + { + "epoch": 0.15201, + "grad_norm": 1.2667175452466484, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 15201 + }, + { + "epoch": 0.15202, + "grad_norm": 0.9320838745996926, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 15202 + }, + { + "epoch": 0.15203, + "grad_norm": 1.0453995384359727, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 15203 + }, + { + "epoch": 0.15204, + "grad_norm": 1.2597649012105097, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 15204 + }, + { + "epoch": 0.15205, + "grad_norm": 1.1961925801405553, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 15205 + }, + { + "epoch": 0.15206, + "grad_norm": 1.181160444706395, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 15206 + }, + { + "epoch": 0.15207, + "grad_norm": 1.4697980303009301, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 15207 + }, + { + "epoch": 0.15208, + "grad_norm": 1.0050373769088243, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 15208 + }, + { + "epoch": 0.15209, + "grad_norm": 1.4211911658134273, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 15209 + }, + { + "epoch": 0.1521, + "grad_norm": 1.0497449525528875, + "learning_rate": 0.003, + "loss": 4.018, + "step": 15210 + }, + { + "epoch": 0.15211, + "grad_norm": 1.3400095009725441, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 15211 + }, + { + "epoch": 0.15212, + "grad_norm": 1.035172309022198, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 15212 + }, + { + "epoch": 0.15213, + "grad_norm": 1.2720197272996543, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 15213 + }, + { + "epoch": 0.15214, + "grad_norm": 1.1914354855934521, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 15214 + }, + { + "epoch": 0.15215, + "grad_norm": 1.4567297823795853, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 15215 + }, + { + "epoch": 0.15216, + "grad_norm": 1.1699730057421303, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 15216 + }, + { + "epoch": 0.15217, + "grad_norm": 1.4255275150565567, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 15217 + }, + { + "epoch": 0.15218, + "grad_norm": 1.3457622723850626, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 15218 + }, + { + "epoch": 0.15219, + "grad_norm": 1.259914794142906, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 15219 + }, + { + "epoch": 0.1522, + "grad_norm": 1.3385804625484585, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 15220 + }, + { + "epoch": 0.15221, + "grad_norm": 1.1452926856393928, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 15221 + }, + { + "epoch": 0.15222, + "grad_norm": 1.2199673305085796, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 15222 + }, + { + "epoch": 0.15223, + "grad_norm": 1.1614406086533022, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 15223 + }, + { + "epoch": 0.15224, + "grad_norm": 1.2578319969789138, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 15224 + }, + { + "epoch": 0.15225, + "grad_norm": 1.0908275282131534, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 15225 + }, + { + "epoch": 0.15226, + "grad_norm": 1.436842674134083, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 15226 + }, + { + "epoch": 0.15227, + "grad_norm": 1.1763822273200253, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 15227 + }, + { + "epoch": 0.15228, + "grad_norm": 1.236367653955839, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 15228 + }, + { + "epoch": 0.15229, + "grad_norm": 1.1369189614223947, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 15229 + }, + { + "epoch": 0.1523, + "grad_norm": 1.3009550318043999, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 15230 + }, + { + "epoch": 0.15231, + "grad_norm": 1.0469260377035405, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 15231 + }, + { + "epoch": 0.15232, + "grad_norm": 1.316792046305704, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 15232 + }, + { + "epoch": 0.15233, + "grad_norm": 1.075327317499854, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 15233 + }, + { + "epoch": 0.15234, + "grad_norm": 1.5401713771580106, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 15234 + }, + { + "epoch": 0.15235, + "grad_norm": 1.3384975214423407, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 15235 + }, + { + "epoch": 0.15236, + "grad_norm": 1.2302142023888756, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 15236 + }, + { + "epoch": 0.15237, + "grad_norm": 1.1467050032159485, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 15237 + }, + { + "epoch": 0.15238, + "grad_norm": 1.2608892431989909, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 15238 + }, + { + "epoch": 0.15239, + "grad_norm": 1.1340177552695188, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 15239 + }, + { + "epoch": 0.1524, + "grad_norm": 1.23521863864813, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 15240 + }, + { + "epoch": 0.15241, + "grad_norm": 1.0968342906944502, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 15241 + }, + { + "epoch": 0.15242, + "grad_norm": 1.4950372693011262, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 15242 + }, + { + "epoch": 0.15243, + "grad_norm": 1.1002275896772342, + "learning_rate": 0.003, + "loss": 4.017, + "step": 15243 + }, + { + "epoch": 0.15244, + "grad_norm": 1.3100180086504831, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 15244 + }, + { + "epoch": 0.15245, + "grad_norm": 1.0125840553200571, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15245 + }, + { + "epoch": 0.15246, + "grad_norm": 1.6922019399605537, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 15246 + }, + { + "epoch": 0.15247, + "grad_norm": 1.0537686111932896, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 15247 + }, + { + "epoch": 0.15248, + "grad_norm": 1.2479280777607784, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 15248 + }, + { + "epoch": 0.15249, + "grad_norm": 1.2439706911893615, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 15249 + }, + { + "epoch": 0.1525, + "grad_norm": 1.2236146810031685, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 15250 + }, + { + "epoch": 0.15251, + "grad_norm": 1.3381460755089774, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 15251 + }, + { + "epoch": 0.15252, + "grad_norm": 1.2956721814013823, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 15252 + }, + { + "epoch": 0.15253, + "grad_norm": 1.1136046665572314, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 15253 + }, + { + "epoch": 0.15254, + "grad_norm": 1.4519019123615422, + "learning_rate": 0.003, + "loss": 4.017, + "step": 15254 + }, + { + "epoch": 0.15255, + "grad_norm": 1.218519913711217, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 15255 + }, + { + "epoch": 0.15256, + "grad_norm": 1.288323294027511, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 15256 + }, + { + "epoch": 0.15257, + "grad_norm": 1.0673560985444304, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 15257 + }, + { + "epoch": 0.15258, + "grad_norm": 1.4258252725808966, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 15258 + }, + { + "epoch": 0.15259, + "grad_norm": 1.1040006998672398, + "learning_rate": 0.003, + "loss": 4.009, + "step": 15259 + }, + { + "epoch": 0.1526, + "grad_norm": 1.2309129950569844, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 15260 + }, + { + "epoch": 0.15261, + "grad_norm": 1.1490315177687394, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 15261 + }, + { + "epoch": 0.15262, + "grad_norm": 1.140301200996627, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15262 + }, + { + "epoch": 0.15263, + "grad_norm": 1.288034047638439, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 15263 + }, + { + "epoch": 0.15264, + "grad_norm": 1.0811920833381112, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 15264 + }, + { + "epoch": 0.15265, + "grad_norm": 1.249639988258038, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 15265 + }, + { + "epoch": 0.15266, + "grad_norm": 1.1603153694948989, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 15266 + }, + { + "epoch": 0.15267, + "grad_norm": 1.3394422598380689, + "learning_rate": 0.003, + "loss": 4.009, + "step": 15267 + }, + { + "epoch": 0.15268, + "grad_norm": 1.0742268220328304, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 15268 + }, + { + "epoch": 0.15269, + "grad_norm": 1.3683478964654803, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 15269 + }, + { + "epoch": 0.1527, + "grad_norm": 1.2695653022628564, + "learning_rate": 0.003, + "loss": 4.011, + "step": 15270 + }, + { + "epoch": 0.15271, + "grad_norm": 1.1547539480855118, + "learning_rate": 0.003, + "loss": 4.027, + "step": 15271 + }, + { + "epoch": 0.15272, + "grad_norm": 1.1942241027035867, + "learning_rate": 0.003, + "loss": 3.995, + "step": 15272 + }, + { + "epoch": 0.15273, + "grad_norm": 1.0934806942778186, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 15273 + }, + { + "epoch": 0.15274, + "grad_norm": 1.2736523352348548, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 15274 + }, + { + "epoch": 0.15275, + "grad_norm": 1.2316668759346256, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 15275 + }, + { + "epoch": 0.15276, + "grad_norm": 0.8866068166018798, + "learning_rate": 0.003, + "loss": 4.023, + "step": 15276 + }, + { + "epoch": 0.15277, + "grad_norm": 1.0359599314182446, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 15277 + }, + { + "epoch": 0.15278, + "grad_norm": 1.4088391751793254, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 15278 + }, + { + "epoch": 0.15279, + "grad_norm": 1.1532474189545543, + "learning_rate": 0.003, + "loss": 4.012, + "step": 15279 + }, + { + "epoch": 0.1528, + "grad_norm": 1.5209095317461196, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 15280 + }, + { + "epoch": 0.15281, + "grad_norm": 0.9678818222371608, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 15281 + }, + { + "epoch": 0.15282, + "grad_norm": 1.3475866215440784, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 15282 + }, + { + "epoch": 0.15283, + "grad_norm": 1.0007327077437924, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 15283 + }, + { + "epoch": 0.15284, + "grad_norm": 1.5299823759901408, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 15284 + }, + { + "epoch": 0.15285, + "grad_norm": 1.115236900005875, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 15285 + }, + { + "epoch": 0.15286, + "grad_norm": 1.222239233298442, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 15286 + }, + { + "epoch": 0.15287, + "grad_norm": 1.105263151096791, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 15287 + }, + { + "epoch": 0.15288, + "grad_norm": 1.1716505993680435, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 15288 + }, + { + "epoch": 0.15289, + "grad_norm": 1.3563443429362065, + "learning_rate": 0.003, + "loss": 3.999, + "step": 15289 + }, + { + "epoch": 0.1529, + "grad_norm": 1.0872489410089983, + "learning_rate": 0.003, + "loss": 4.017, + "step": 15290 + }, + { + "epoch": 0.15291, + "grad_norm": 1.644791899750319, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 15291 + }, + { + "epoch": 0.15292, + "grad_norm": 1.006959413935244, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 15292 + }, + { + "epoch": 0.15293, + "grad_norm": 1.598524079754938, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 15293 + }, + { + "epoch": 0.15294, + "grad_norm": 1.0940183399275392, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 15294 + }, + { + "epoch": 0.15295, + "grad_norm": 1.460337046890657, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 15295 + }, + { + "epoch": 0.15296, + "grad_norm": 1.045725212899953, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 15296 + }, + { + "epoch": 0.15297, + "grad_norm": 1.5802895346212449, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 15297 + }, + { + "epoch": 0.15298, + "grad_norm": 1.3837242228352484, + "learning_rate": 0.003, + "loss": 4.049, + "step": 15298 + }, + { + "epoch": 0.15299, + "grad_norm": 1.2728385968327829, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 15299 + }, + { + "epoch": 0.153, + "grad_norm": 1.1823117384581505, + "learning_rate": 0.003, + "loss": 4.046, + "step": 15300 + }, + { + "epoch": 0.15301, + "grad_norm": 1.1320250057435113, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 15301 + }, + { + "epoch": 0.15302, + "grad_norm": 1.4255218577322437, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 15302 + }, + { + "epoch": 0.15303, + "grad_norm": 1.1908400209785568, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 15303 + }, + { + "epoch": 0.15304, + "grad_norm": 1.3528746438276917, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 15304 + }, + { + "epoch": 0.15305, + "grad_norm": 1.1938813749340769, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 15305 + }, + { + "epoch": 0.15306, + "grad_norm": 1.6421325596095928, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 15306 + }, + { + "epoch": 0.15307, + "grad_norm": 0.951216559987826, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 15307 + }, + { + "epoch": 0.15308, + "grad_norm": 1.2691738035815379, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 15308 + }, + { + "epoch": 0.15309, + "grad_norm": 0.9630179201647759, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 15309 + }, + { + "epoch": 0.1531, + "grad_norm": 1.3626385785070245, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 15310 + }, + { + "epoch": 0.15311, + "grad_norm": 1.0798825353592207, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 15311 + }, + { + "epoch": 0.15312, + "grad_norm": 1.2748408005075207, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 15312 + }, + { + "epoch": 0.15313, + "grad_norm": 1.3160143379370914, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 15313 + }, + { + "epoch": 0.15314, + "grad_norm": 1.3381560563598656, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 15314 + }, + { + "epoch": 0.15315, + "grad_norm": 1.1003058742942267, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 15315 + }, + { + "epoch": 0.15316, + "grad_norm": 1.28462379604353, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 15316 + }, + { + "epoch": 0.15317, + "grad_norm": 1.0558852666618743, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 15317 + }, + { + "epoch": 0.15318, + "grad_norm": 1.1925633714135881, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15318 + }, + { + "epoch": 0.15319, + "grad_norm": 1.1516827476554599, + "learning_rate": 0.003, + "loss": 3.993, + "step": 15319 + }, + { + "epoch": 0.1532, + "grad_norm": 1.1443399520062345, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 15320 + }, + { + "epoch": 0.15321, + "grad_norm": 1.2249706571549668, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 15321 + }, + { + "epoch": 0.15322, + "grad_norm": 1.22913366138089, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 15322 + }, + { + "epoch": 0.15323, + "grad_norm": 1.3641024561312933, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 15323 + }, + { + "epoch": 0.15324, + "grad_norm": 1.1013868436436511, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15324 + }, + { + "epoch": 0.15325, + "grad_norm": 1.3058725139288032, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 15325 + }, + { + "epoch": 0.15326, + "grad_norm": 1.2659469483828032, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 15326 + }, + { + "epoch": 0.15327, + "grad_norm": 1.261903749150775, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 15327 + }, + { + "epoch": 0.15328, + "grad_norm": 1.0976890628915055, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 15328 + }, + { + "epoch": 0.15329, + "grad_norm": 1.3073593212321624, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 15329 + }, + { + "epoch": 0.1533, + "grad_norm": 1.2307470804948377, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 15330 + }, + { + "epoch": 0.15331, + "grad_norm": 1.1548746627152964, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 15331 + }, + { + "epoch": 0.15332, + "grad_norm": 1.1745992236943885, + "learning_rate": 0.003, + "loss": 4.0, + "step": 15332 + }, + { + "epoch": 0.15333, + "grad_norm": 1.29812710663051, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 15333 + }, + { + "epoch": 0.15334, + "grad_norm": 0.9390035244787395, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 15334 + }, + { + "epoch": 0.15335, + "grad_norm": 1.122648019952921, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 15335 + }, + { + "epoch": 0.15336, + "grad_norm": 1.5149705395129167, + "learning_rate": 0.003, + "loss": 4.021, + "step": 15336 + }, + { + "epoch": 0.15337, + "grad_norm": 1.0013137038070798, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 15337 + }, + { + "epoch": 0.15338, + "grad_norm": 1.299481327762594, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 15338 + }, + { + "epoch": 0.15339, + "grad_norm": 1.4471932995791132, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 15339 + }, + { + "epoch": 0.1534, + "grad_norm": 0.9648166180616551, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 15340 + }, + { + "epoch": 0.15341, + "grad_norm": 1.4509801117246968, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 15341 + }, + { + "epoch": 0.15342, + "grad_norm": 1.1254236266710087, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 15342 + }, + { + "epoch": 0.15343, + "grad_norm": 1.3692642680880276, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 15343 + }, + { + "epoch": 0.15344, + "grad_norm": 1.2619386001461093, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 15344 + }, + { + "epoch": 0.15345, + "grad_norm": 1.2290685149803946, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 15345 + }, + { + "epoch": 0.15346, + "grad_norm": 1.1694162094841682, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 15346 + }, + { + "epoch": 0.15347, + "grad_norm": 1.224932391008914, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 15347 + }, + { + "epoch": 0.15348, + "grad_norm": 1.2410636760246108, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 15348 + }, + { + "epoch": 0.15349, + "grad_norm": 1.118028257618968, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 15349 + }, + { + "epoch": 0.1535, + "grad_norm": 1.2307429212771872, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 15350 + }, + { + "epoch": 0.15351, + "grad_norm": 1.220610134666719, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 15351 + }, + { + "epoch": 0.15352, + "grad_norm": 1.489724020189981, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15352 + }, + { + "epoch": 0.15353, + "grad_norm": 1.1392495632135529, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 15353 + }, + { + "epoch": 0.15354, + "grad_norm": 1.3664724518429507, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 15354 + }, + { + "epoch": 0.15355, + "grad_norm": 1.0740409634784727, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 15355 + }, + { + "epoch": 0.15356, + "grad_norm": 1.2778573411778194, + "learning_rate": 0.003, + "loss": 4.007, + "step": 15356 + }, + { + "epoch": 0.15357, + "grad_norm": 1.506825805600986, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 15357 + }, + { + "epoch": 0.15358, + "grad_norm": 1.0304333936951433, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 15358 + }, + { + "epoch": 0.15359, + "grad_norm": 1.4557553962981578, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 15359 + }, + { + "epoch": 0.1536, + "grad_norm": 0.9998670788123999, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 15360 + }, + { + "epoch": 0.15361, + "grad_norm": 1.3285068607989634, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 15361 + }, + { + "epoch": 0.15362, + "grad_norm": 1.1524236129558298, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 15362 + }, + { + "epoch": 0.15363, + "grad_norm": 1.3631793589844639, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 15363 + }, + { + "epoch": 0.15364, + "grad_norm": 1.3841052064121926, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 15364 + }, + { + "epoch": 0.15365, + "grad_norm": 1.148427490969218, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 15365 + }, + { + "epoch": 0.15366, + "grad_norm": 1.3036553111173732, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 15366 + }, + { + "epoch": 0.15367, + "grad_norm": 1.0642389435975332, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 15367 + }, + { + "epoch": 0.15368, + "grad_norm": 1.4958489290963009, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 15368 + }, + { + "epoch": 0.15369, + "grad_norm": 0.9916113308781349, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 15369 + }, + { + "epoch": 0.1537, + "grad_norm": 1.521643615228874, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 15370 + }, + { + "epoch": 0.15371, + "grad_norm": 1.3960310353379668, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 15371 + }, + { + "epoch": 0.15372, + "grad_norm": 1.1601899119026602, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 15372 + }, + { + "epoch": 0.15373, + "grad_norm": 1.1807936206065663, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 15373 + }, + { + "epoch": 0.15374, + "grad_norm": 1.127100500721332, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15374 + }, + { + "epoch": 0.15375, + "grad_norm": 1.031832545244089, + "learning_rate": 0.003, + "loss": 4.009, + "step": 15375 + }, + { + "epoch": 0.15376, + "grad_norm": 1.4215603656354039, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 15376 + }, + { + "epoch": 0.15377, + "grad_norm": 1.0606115114134587, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 15377 + }, + { + "epoch": 0.15378, + "grad_norm": 1.3744579165162751, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 15378 + }, + { + "epoch": 0.15379, + "grad_norm": 1.2139090258538139, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 15379 + }, + { + "epoch": 0.1538, + "grad_norm": 1.2386901117854758, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 15380 + }, + { + "epoch": 0.15381, + "grad_norm": 1.147397648133237, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15381 + }, + { + "epoch": 0.15382, + "grad_norm": 1.332374579095073, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 15382 + }, + { + "epoch": 0.15383, + "grad_norm": 1.0059314793160214, + "learning_rate": 0.003, + "loss": 4.04, + "step": 15383 + }, + { + "epoch": 0.15384, + "grad_norm": 1.2971956854525335, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 15384 + }, + { + "epoch": 0.15385, + "grad_norm": 1.1863692569687379, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 15385 + }, + { + "epoch": 0.15386, + "grad_norm": 1.147517167221806, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 15386 + }, + { + "epoch": 0.15387, + "grad_norm": 1.0845869428548975, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 15387 + }, + { + "epoch": 0.15388, + "grad_norm": 1.4415714589859954, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 15388 + }, + { + "epoch": 0.15389, + "grad_norm": 1.01200996419366, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 15389 + }, + { + "epoch": 0.1539, + "grad_norm": 1.1968289569868962, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 15390 + }, + { + "epoch": 0.15391, + "grad_norm": 1.4124652006603855, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 15391 + }, + { + "epoch": 0.15392, + "grad_norm": 1.1429694971029492, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 15392 + }, + { + "epoch": 0.15393, + "grad_norm": 1.3989460206903803, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 15393 + }, + { + "epoch": 0.15394, + "grad_norm": 1.1275817712338951, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 15394 + }, + { + "epoch": 0.15395, + "grad_norm": 1.6527313295290693, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 15395 + }, + { + "epoch": 0.15396, + "grad_norm": 1.1606484889693214, + "learning_rate": 0.003, + "loss": 3.996, + "step": 15396 + }, + { + "epoch": 0.15397, + "grad_norm": 1.3063989661917172, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 15397 + }, + { + "epoch": 0.15398, + "grad_norm": 1.1175556092640544, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 15398 + }, + { + "epoch": 0.15399, + "grad_norm": 1.2611769514814661, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 15399 + }, + { + "epoch": 0.154, + "grad_norm": 1.0227787885978328, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 15400 + }, + { + "epoch": 0.15401, + "grad_norm": 1.175010936201811, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 15401 + }, + { + "epoch": 0.15402, + "grad_norm": 1.301898398708719, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15402 + }, + { + "epoch": 0.15403, + "grad_norm": 1.0108000392959706, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 15403 + }, + { + "epoch": 0.15404, + "grad_norm": 1.3416860945801337, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 15404 + }, + { + "epoch": 0.15405, + "grad_norm": 1.1570945452891142, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 15405 + }, + { + "epoch": 0.15406, + "grad_norm": 1.3655121369212266, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 15406 + }, + { + "epoch": 0.15407, + "grad_norm": 1.3960011028879589, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 15407 + }, + { + "epoch": 0.15408, + "grad_norm": 1.141425990161296, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 15408 + }, + { + "epoch": 0.15409, + "grad_norm": 1.3061620835789476, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 15409 + }, + { + "epoch": 0.1541, + "grad_norm": 0.9288136755410322, + "learning_rate": 0.003, + "loss": 4.025, + "step": 15410 + }, + { + "epoch": 0.15411, + "grad_norm": 1.2780898944597026, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 15411 + }, + { + "epoch": 0.15412, + "grad_norm": 1.1165309242772365, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15412 + }, + { + "epoch": 0.15413, + "grad_norm": 1.2588491423612371, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 15413 + }, + { + "epoch": 0.15414, + "grad_norm": 1.153293339632718, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 15414 + }, + { + "epoch": 0.15415, + "grad_norm": 1.2763101314205272, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 15415 + }, + { + "epoch": 0.15416, + "grad_norm": 1.445718992177873, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 15416 + }, + { + "epoch": 0.15417, + "grad_norm": 1.1010078858794556, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 15417 + }, + { + "epoch": 0.15418, + "grad_norm": 1.3389978908349613, + "learning_rate": 0.003, + "loss": 3.992, + "step": 15418 + }, + { + "epoch": 0.15419, + "grad_norm": 1.1161796140807403, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 15419 + }, + { + "epoch": 0.1542, + "grad_norm": 1.2183616590008166, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15420 + }, + { + "epoch": 0.15421, + "grad_norm": 1.0704559979666672, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 15421 + }, + { + "epoch": 0.15422, + "grad_norm": 1.0742924028397265, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 15422 + }, + { + "epoch": 0.15423, + "grad_norm": 1.2111405828356576, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 15423 + }, + { + "epoch": 0.15424, + "grad_norm": 1.0235465294837345, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 15424 + }, + { + "epoch": 0.15425, + "grad_norm": 1.416013184766158, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 15425 + }, + { + "epoch": 0.15426, + "grad_norm": 1.0329446806219278, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 15426 + }, + { + "epoch": 0.15427, + "grad_norm": 1.3111226902340347, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 15427 + }, + { + "epoch": 0.15428, + "grad_norm": 1.330497005134314, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 15428 + }, + { + "epoch": 0.15429, + "grad_norm": 1.403787119578501, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 15429 + }, + { + "epoch": 0.1543, + "grad_norm": 1.0628919762318099, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 15430 + }, + { + "epoch": 0.15431, + "grad_norm": 1.4918675346184322, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 15431 + }, + { + "epoch": 0.15432, + "grad_norm": 1.0839100804515, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 15432 + }, + { + "epoch": 0.15433, + "grad_norm": 1.3741229327105866, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 15433 + }, + { + "epoch": 0.15434, + "grad_norm": 1.014335655092048, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 15434 + }, + { + "epoch": 0.15435, + "grad_norm": 1.5198132343959803, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 15435 + }, + { + "epoch": 0.15436, + "grad_norm": 1.2232189794341666, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 15436 + }, + { + "epoch": 0.15437, + "grad_norm": 1.1922898524941339, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 15437 + }, + { + "epoch": 0.15438, + "grad_norm": 1.4312060880257944, + "learning_rate": 0.003, + "loss": 4.049, + "step": 15438 + }, + { + "epoch": 0.15439, + "grad_norm": 1.3761522420073637, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 15439 + }, + { + "epoch": 0.1544, + "grad_norm": 1.1878663856256735, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 15440 + }, + { + "epoch": 0.15441, + "grad_norm": 1.409308011730964, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 15441 + }, + { + "epoch": 0.15442, + "grad_norm": 0.8818847587507577, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 15442 + }, + { + "epoch": 0.15443, + "grad_norm": 1.181172095811864, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 15443 + }, + { + "epoch": 0.15444, + "grad_norm": 1.1697211130303764, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 15444 + }, + { + "epoch": 0.15445, + "grad_norm": 1.0309657250127224, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 15445 + }, + { + "epoch": 0.15446, + "grad_norm": 1.3297375349823675, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 15446 + }, + { + "epoch": 0.15447, + "grad_norm": 1.1139099407835233, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 15447 + }, + { + "epoch": 0.15448, + "grad_norm": 1.237655786195009, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 15448 + }, + { + "epoch": 0.15449, + "grad_norm": 1.2789726132717136, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 15449 + }, + { + "epoch": 0.1545, + "grad_norm": 1.2968964530353215, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 15450 + }, + { + "epoch": 0.15451, + "grad_norm": 1.1922491983748065, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 15451 + }, + { + "epoch": 0.15452, + "grad_norm": 1.399508785372692, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 15452 + }, + { + "epoch": 0.15453, + "grad_norm": 0.9931795257761599, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 15453 + }, + { + "epoch": 0.15454, + "grad_norm": 1.343499007254131, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 15454 + }, + { + "epoch": 0.15455, + "grad_norm": 1.0748854282639977, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 15455 + }, + { + "epoch": 0.15456, + "grad_norm": 1.4505517385476958, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 15456 + }, + { + "epoch": 0.15457, + "grad_norm": 1.1089755331302413, + "learning_rate": 0.003, + "loss": 4.02, + "step": 15457 + }, + { + "epoch": 0.15458, + "grad_norm": 1.4647167086584743, + "learning_rate": 0.003, + "loss": 4.023, + "step": 15458 + }, + { + "epoch": 0.15459, + "grad_norm": 1.0699807413746714, + "learning_rate": 0.003, + "loss": 4.007, + "step": 15459 + }, + { + "epoch": 0.1546, + "grad_norm": 1.425807860008332, + "learning_rate": 0.003, + "loss": 3.997, + "step": 15460 + }, + { + "epoch": 0.15461, + "grad_norm": 1.0401575014221385, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 15461 + }, + { + "epoch": 0.15462, + "grad_norm": 1.3587624825650655, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 15462 + }, + { + "epoch": 0.15463, + "grad_norm": 1.1371008023815934, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 15463 + }, + { + "epoch": 0.15464, + "grad_norm": 1.2999090728439768, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 15464 + }, + { + "epoch": 0.15465, + "grad_norm": 1.2658049390348993, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 15465 + }, + { + "epoch": 0.15466, + "grad_norm": 1.1640506691228396, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 15466 + }, + { + "epoch": 0.15467, + "grad_norm": 1.2667810282519953, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 15467 + }, + { + "epoch": 0.15468, + "grad_norm": 1.0106880311117121, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 15468 + }, + { + "epoch": 0.15469, + "grad_norm": 1.1254762579400168, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 15469 + }, + { + "epoch": 0.1547, + "grad_norm": 1.3597427881980186, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 15470 + }, + { + "epoch": 0.15471, + "grad_norm": 1.0451997736704248, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 15471 + }, + { + "epoch": 0.15472, + "grad_norm": 1.4638421033676579, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 15472 + }, + { + "epoch": 0.15473, + "grad_norm": 0.9818904122801336, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 15473 + }, + { + "epoch": 0.15474, + "grad_norm": 1.3113123003195055, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 15474 + }, + { + "epoch": 0.15475, + "grad_norm": 1.0219220699211995, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 15475 + }, + { + "epoch": 0.15476, + "grad_norm": 1.3721315900566593, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 15476 + }, + { + "epoch": 0.15477, + "grad_norm": 1.0382470290868924, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 15477 + }, + { + "epoch": 0.15478, + "grad_norm": 1.391397070429852, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15478 + }, + { + "epoch": 0.15479, + "grad_norm": 1.4210267801203937, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15479 + }, + { + "epoch": 0.1548, + "grad_norm": 1.1612045895435392, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 15480 + }, + { + "epoch": 0.15481, + "grad_norm": 1.0937338117526525, + "learning_rate": 0.003, + "loss": 4.025, + "step": 15481 + }, + { + "epoch": 0.15482, + "grad_norm": 1.3604106267692817, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15482 + }, + { + "epoch": 0.15483, + "grad_norm": 1.1859626833742631, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 15483 + }, + { + "epoch": 0.15484, + "grad_norm": 1.201582766515893, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 15484 + }, + { + "epoch": 0.15485, + "grad_norm": 1.3486346050306894, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 15485 + }, + { + "epoch": 0.15486, + "grad_norm": 1.0764828931379589, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 15486 + }, + { + "epoch": 0.15487, + "grad_norm": 1.3408549143673125, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15487 + }, + { + "epoch": 0.15488, + "grad_norm": 1.1300827202543724, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 15488 + }, + { + "epoch": 0.15489, + "grad_norm": 1.1709510441950886, + "learning_rate": 0.003, + "loss": 4.01, + "step": 15489 + }, + { + "epoch": 0.1549, + "grad_norm": 1.249822152716845, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 15490 + }, + { + "epoch": 0.15491, + "grad_norm": 1.291335979959531, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 15491 + }, + { + "epoch": 0.15492, + "grad_norm": 1.3396259417898158, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 15492 + }, + { + "epoch": 0.15493, + "grad_norm": 1.2814611940312006, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 15493 + }, + { + "epoch": 0.15494, + "grad_norm": 1.1777452240371598, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 15494 + }, + { + "epoch": 0.15495, + "grad_norm": 1.1592037969328344, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 15495 + }, + { + "epoch": 0.15496, + "grad_norm": 1.1044762203900087, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 15496 + }, + { + "epoch": 0.15497, + "grad_norm": 1.2668704762351173, + "learning_rate": 0.003, + "loss": 4.011, + "step": 15497 + }, + { + "epoch": 0.15498, + "grad_norm": 1.2622274894185679, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 15498 + }, + { + "epoch": 0.15499, + "grad_norm": 1.1551704191426173, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 15499 + }, + { + "epoch": 0.155, + "grad_norm": 1.2922799353230623, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 15500 + }, + { + "epoch": 0.15501, + "grad_norm": 1.2177967791089919, + "learning_rate": 0.003, + "loss": 4.002, + "step": 15501 + }, + { + "epoch": 0.15502, + "grad_norm": 1.1380041505072185, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 15502 + }, + { + "epoch": 0.15503, + "grad_norm": 1.2326803571856018, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15503 + }, + { + "epoch": 0.15504, + "grad_norm": 1.0677489422471353, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 15504 + }, + { + "epoch": 0.15505, + "grad_norm": 1.4375907997925543, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 15505 + }, + { + "epoch": 0.15506, + "grad_norm": 0.9940022896813783, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 15506 + }, + { + "epoch": 0.15507, + "grad_norm": 1.8163683369923858, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 15507 + }, + { + "epoch": 0.15508, + "grad_norm": 0.9821455908225213, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 15508 + }, + { + "epoch": 0.15509, + "grad_norm": 1.1321623600415103, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 15509 + }, + { + "epoch": 0.1551, + "grad_norm": 1.4757627649908576, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 15510 + }, + { + "epoch": 0.15511, + "grad_norm": 1.1836268113812825, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 15511 + }, + { + "epoch": 0.15512, + "grad_norm": 1.4550460571051078, + "learning_rate": 0.003, + "loss": 4.052, + "step": 15512 + }, + { + "epoch": 0.15513, + "grad_norm": 1.0424281490788763, + "learning_rate": 0.003, + "loss": 4.011, + "step": 15513 + }, + { + "epoch": 0.15514, + "grad_norm": 1.3060251931644995, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 15514 + }, + { + "epoch": 0.15515, + "grad_norm": 1.1840927008821804, + "learning_rate": 0.003, + "loss": 4.014, + "step": 15515 + }, + { + "epoch": 0.15516, + "grad_norm": 1.3355984599902573, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 15516 + }, + { + "epoch": 0.15517, + "grad_norm": 1.3854475674976467, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 15517 + }, + { + "epoch": 0.15518, + "grad_norm": 1.1903115013555348, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 15518 + }, + { + "epoch": 0.15519, + "grad_norm": 1.3645758618524382, + "learning_rate": 0.003, + "loss": 4.003, + "step": 15519 + }, + { + "epoch": 0.1552, + "grad_norm": 1.1420658547143592, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 15520 + }, + { + "epoch": 0.15521, + "grad_norm": 1.0788356898283442, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 15521 + }, + { + "epoch": 0.15522, + "grad_norm": 1.2409217539764137, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 15522 + }, + { + "epoch": 0.15523, + "grad_norm": 1.1395473943660315, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 15523 + }, + { + "epoch": 0.15524, + "grad_norm": 1.1662676153801086, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 15524 + }, + { + "epoch": 0.15525, + "grad_norm": 1.2325072549210612, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 15525 + }, + { + "epoch": 0.15526, + "grad_norm": 1.1737878425195596, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 15526 + }, + { + "epoch": 0.15527, + "grad_norm": 1.3796725225222926, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 15527 + }, + { + "epoch": 0.15528, + "grad_norm": 1.2538521278635026, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 15528 + }, + { + "epoch": 0.15529, + "grad_norm": 1.1800656715900641, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 15529 + }, + { + "epoch": 0.1553, + "grad_norm": 1.205066071322585, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 15530 + }, + { + "epoch": 0.15531, + "grad_norm": 1.019614767721003, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 15531 + }, + { + "epoch": 0.15532, + "grad_norm": 1.1759045019842371, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 15532 + }, + { + "epoch": 0.15533, + "grad_norm": 1.1854990433424775, + "learning_rate": 0.003, + "loss": 4.025, + "step": 15533 + }, + { + "epoch": 0.15534, + "grad_norm": 1.196400638478638, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 15534 + }, + { + "epoch": 0.15535, + "grad_norm": 1.5511982731401823, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 15535 + }, + { + "epoch": 0.15536, + "grad_norm": 0.8202768827545235, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 15536 + }, + { + "epoch": 0.15537, + "grad_norm": 1.0404641833258768, + "learning_rate": 0.003, + "loss": 4.027, + "step": 15537 + }, + { + "epoch": 0.15538, + "grad_norm": 1.3069588417679192, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 15538 + }, + { + "epoch": 0.15539, + "grad_norm": 1.272868134888428, + "learning_rate": 0.003, + "loss": 4.042, + "step": 15539 + }, + { + "epoch": 0.1554, + "grad_norm": 1.427051787014361, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 15540 + }, + { + "epoch": 0.15541, + "grad_norm": 1.1839923373571692, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 15541 + }, + { + "epoch": 0.15542, + "grad_norm": 1.3716600078861727, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 15542 + }, + { + "epoch": 0.15543, + "grad_norm": 1.0728442089606187, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 15543 + }, + { + "epoch": 0.15544, + "grad_norm": 1.3486558614426765, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 15544 + }, + { + "epoch": 0.15545, + "grad_norm": 1.1889424087910203, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 15545 + }, + { + "epoch": 0.15546, + "grad_norm": 1.4268357117744535, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 15546 + }, + { + "epoch": 0.15547, + "grad_norm": 1.1236587367880921, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15547 + }, + { + "epoch": 0.15548, + "grad_norm": 1.4448499863917188, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 15548 + }, + { + "epoch": 0.15549, + "grad_norm": 0.9604154047060119, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 15549 + }, + { + "epoch": 0.1555, + "grad_norm": 1.257451362138832, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 15550 + }, + { + "epoch": 0.15551, + "grad_norm": 1.1220385706073506, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 15551 + }, + { + "epoch": 0.15552, + "grad_norm": 1.3838135909506402, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 15552 + }, + { + "epoch": 0.15553, + "grad_norm": 0.9600748836538427, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 15553 + }, + { + "epoch": 0.15554, + "grad_norm": 1.1852568034872384, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 15554 + }, + { + "epoch": 0.15555, + "grad_norm": 1.174714423688401, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 15555 + }, + { + "epoch": 0.15556, + "grad_norm": 1.2123765713461248, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 15556 + }, + { + "epoch": 0.15557, + "grad_norm": 1.1638765547588363, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 15557 + }, + { + "epoch": 0.15558, + "grad_norm": 1.0850237294958363, + "learning_rate": 0.003, + "loss": 4.012, + "step": 15558 + }, + { + "epoch": 0.15559, + "grad_norm": 1.5111875014671927, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 15559 + }, + { + "epoch": 0.1556, + "grad_norm": 1.4196270330515373, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 15560 + }, + { + "epoch": 0.15561, + "grad_norm": 1.2845578894800977, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 15561 + }, + { + "epoch": 0.15562, + "grad_norm": 1.107438276853264, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 15562 + }, + { + "epoch": 0.15563, + "grad_norm": 0.9621144226586266, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 15563 + }, + { + "epoch": 0.15564, + "grad_norm": 1.2080448978519494, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 15564 + }, + { + "epoch": 0.15565, + "grad_norm": 1.177355515944768, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 15565 + }, + { + "epoch": 0.15566, + "grad_norm": 1.3013818446067609, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 15566 + }, + { + "epoch": 0.15567, + "grad_norm": 1.4693577918031457, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 15567 + }, + { + "epoch": 0.15568, + "grad_norm": 1.1977300101325452, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 15568 + }, + { + "epoch": 0.15569, + "grad_norm": 1.30836741460951, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 15569 + }, + { + "epoch": 0.1557, + "grad_norm": 1.1907578962424745, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 15570 + }, + { + "epoch": 0.15571, + "grad_norm": 1.4503119718230573, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 15571 + }, + { + "epoch": 0.15572, + "grad_norm": 1.278895214753018, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 15572 + }, + { + "epoch": 0.15573, + "grad_norm": 1.0395796991744881, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 15573 + }, + { + "epoch": 0.15574, + "grad_norm": 1.6556038524578698, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 15574 + }, + { + "epoch": 0.15575, + "grad_norm": 1.0835814542160225, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 15575 + }, + { + "epoch": 0.15576, + "grad_norm": 1.5193905404710812, + "learning_rate": 0.003, + "loss": 4.029, + "step": 15576 + }, + { + "epoch": 0.15577, + "grad_norm": 1.1346197793117552, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 15577 + }, + { + "epoch": 0.15578, + "grad_norm": 1.1313532216410387, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 15578 + }, + { + "epoch": 0.15579, + "grad_norm": 1.4615977378034375, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 15579 + }, + { + "epoch": 0.1558, + "grad_norm": 1.1023308878737557, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 15580 + }, + { + "epoch": 0.15581, + "grad_norm": 1.3832846645080352, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 15581 + }, + { + "epoch": 0.15582, + "grad_norm": 1.1717556196242371, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 15582 + }, + { + "epoch": 0.15583, + "grad_norm": 1.4744072510739992, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 15583 + }, + { + "epoch": 0.15584, + "grad_norm": 1.052391300109064, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 15584 + }, + { + "epoch": 0.15585, + "grad_norm": 1.3762594933567556, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 15585 + }, + { + "epoch": 0.15586, + "grad_norm": 1.2243391222677933, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 15586 + }, + { + "epoch": 0.15587, + "grad_norm": 1.267717794998198, + "learning_rate": 0.003, + "loss": 3.996, + "step": 15587 + }, + { + "epoch": 0.15588, + "grad_norm": 1.3117570916822947, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 15588 + }, + { + "epoch": 0.15589, + "grad_norm": 1.1515375084571582, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 15589 + }, + { + "epoch": 0.1559, + "grad_norm": 1.2237350512132354, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15590 + }, + { + "epoch": 0.15591, + "grad_norm": 1.2770771620758707, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 15591 + }, + { + "epoch": 0.15592, + "grad_norm": 1.2947035832215696, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 15592 + }, + { + "epoch": 0.15593, + "grad_norm": 1.131252977144435, + "learning_rate": 0.003, + "loss": 3.98, + "step": 15593 + }, + { + "epoch": 0.15594, + "grad_norm": 1.060910237472633, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 15594 + }, + { + "epoch": 0.15595, + "grad_norm": 1.1386813912532787, + "learning_rate": 0.003, + "loss": 3.985, + "step": 15595 + }, + { + "epoch": 0.15596, + "grad_norm": 1.038484224163973, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 15596 + }, + { + "epoch": 0.15597, + "grad_norm": 1.3268715049524036, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 15597 + }, + { + "epoch": 0.15598, + "grad_norm": 1.1932150584347168, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 15598 + }, + { + "epoch": 0.15599, + "grad_norm": 1.631553122367338, + "learning_rate": 0.003, + "loss": 4.001, + "step": 15599 + }, + { + "epoch": 0.156, + "grad_norm": 1.031280774003672, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 15600 + }, + { + "epoch": 0.15601, + "grad_norm": 1.3259177715354908, + "learning_rate": 0.003, + "loss": 4.012, + "step": 15601 + }, + { + "epoch": 0.15602, + "grad_norm": 1.3552168329577372, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 15602 + }, + { + "epoch": 0.15603, + "grad_norm": 1.05154774472057, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 15603 + }, + { + "epoch": 0.15604, + "grad_norm": 1.4783019636415102, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 15604 + }, + { + "epoch": 0.15605, + "grad_norm": 1.1577328894270635, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 15605 + }, + { + "epoch": 0.15606, + "grad_norm": 1.378054590499981, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 15606 + }, + { + "epoch": 0.15607, + "grad_norm": 1.196110950152625, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 15607 + }, + { + "epoch": 0.15608, + "grad_norm": 1.257660491301664, + "learning_rate": 0.003, + "loss": 4.02, + "step": 15608 + }, + { + "epoch": 0.15609, + "grad_norm": 1.3290291137316042, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 15609 + }, + { + "epoch": 0.1561, + "grad_norm": 1.0273929162359139, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 15610 + }, + { + "epoch": 0.15611, + "grad_norm": 1.1489898023789653, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 15611 + }, + { + "epoch": 0.15612, + "grad_norm": 1.1284407520548465, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 15612 + }, + { + "epoch": 0.15613, + "grad_norm": 1.362815278658877, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 15613 + }, + { + "epoch": 0.15614, + "grad_norm": 1.0728144070534393, + "learning_rate": 0.003, + "loss": 3.997, + "step": 15614 + }, + { + "epoch": 0.15615, + "grad_norm": 1.6027777923758262, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 15615 + }, + { + "epoch": 0.15616, + "grad_norm": 1.0754283703159526, + "learning_rate": 0.003, + "loss": 4.011, + "step": 15616 + }, + { + "epoch": 0.15617, + "grad_norm": 1.526599269818748, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 15617 + }, + { + "epoch": 0.15618, + "grad_norm": 1.0939402000174385, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 15618 + }, + { + "epoch": 0.15619, + "grad_norm": 1.2494915990304731, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 15619 + }, + { + "epoch": 0.1562, + "grad_norm": 1.226717648159094, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 15620 + }, + { + "epoch": 0.15621, + "grad_norm": 1.3602513260753404, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 15621 + }, + { + "epoch": 0.15622, + "grad_norm": 1.1010272012241384, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 15622 + }, + { + "epoch": 0.15623, + "grad_norm": 1.4585470776423317, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 15623 + }, + { + "epoch": 0.15624, + "grad_norm": 1.0050101911556795, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 15624 + }, + { + "epoch": 0.15625, + "grad_norm": 1.2981447508412527, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 15625 + }, + { + "epoch": 0.15626, + "grad_norm": 1.170417358387042, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15626 + }, + { + "epoch": 0.15627, + "grad_norm": 1.2314174237285826, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 15627 + }, + { + "epoch": 0.15628, + "grad_norm": 1.1524994522301018, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 15628 + }, + { + "epoch": 0.15629, + "grad_norm": 1.318418554553263, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 15629 + }, + { + "epoch": 0.1563, + "grad_norm": 1.3949926386579614, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 15630 + }, + { + "epoch": 0.15631, + "grad_norm": 1.0556249059216092, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15631 + }, + { + "epoch": 0.15632, + "grad_norm": 1.2179348218164991, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 15632 + }, + { + "epoch": 0.15633, + "grad_norm": 1.3034648865720937, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 15633 + }, + { + "epoch": 0.15634, + "grad_norm": 1.1642828471343658, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 15634 + }, + { + "epoch": 0.15635, + "grad_norm": 1.3870366758727755, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 15635 + }, + { + "epoch": 0.15636, + "grad_norm": 1.162394387820439, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 15636 + }, + { + "epoch": 0.15637, + "grad_norm": 1.1264573153720103, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 15637 + }, + { + "epoch": 0.15638, + "grad_norm": 1.2210700342418637, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 15638 + }, + { + "epoch": 0.15639, + "grad_norm": 1.3816056938231887, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 15639 + }, + { + "epoch": 0.1564, + "grad_norm": 1.1240051895362912, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 15640 + }, + { + "epoch": 0.15641, + "grad_norm": 1.2286781372547828, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 15641 + }, + { + "epoch": 0.15642, + "grad_norm": 1.0770756354352453, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 15642 + }, + { + "epoch": 0.15643, + "grad_norm": 1.1868289717266771, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 15643 + }, + { + "epoch": 0.15644, + "grad_norm": 1.2934548543139381, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 15644 + }, + { + "epoch": 0.15645, + "grad_norm": 1.0885130534618273, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 15645 + }, + { + "epoch": 0.15646, + "grad_norm": 1.4594782536308275, + "learning_rate": 0.003, + "loss": 3.986, + "step": 15646 + }, + { + "epoch": 0.15647, + "grad_norm": 1.171122790155785, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 15647 + }, + { + "epoch": 0.15648, + "grad_norm": 1.3247598811351102, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 15648 + }, + { + "epoch": 0.15649, + "grad_norm": 1.0884078643809036, + "learning_rate": 0.003, + "loss": 4.014, + "step": 15649 + }, + { + "epoch": 0.1565, + "grad_norm": 1.2625242738306834, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 15650 + }, + { + "epoch": 0.15651, + "grad_norm": 1.2590093124234436, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 15651 + }, + { + "epoch": 0.15652, + "grad_norm": 1.0386595781942116, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 15652 + }, + { + "epoch": 0.15653, + "grad_norm": 1.4220855535299173, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 15653 + }, + { + "epoch": 0.15654, + "grad_norm": 1.2390563975423328, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 15654 + }, + { + "epoch": 0.15655, + "grad_norm": 1.5156655582788026, + "learning_rate": 0.003, + "loss": 4.059, + "step": 15655 + }, + { + "epoch": 0.15656, + "grad_norm": 1.0683641273964668, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 15656 + }, + { + "epoch": 0.15657, + "grad_norm": 1.3594544693303718, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 15657 + }, + { + "epoch": 0.15658, + "grad_norm": 1.2731269504661384, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 15658 + }, + { + "epoch": 0.15659, + "grad_norm": 1.2439084811529881, + "learning_rate": 0.003, + "loss": 4.032, + "step": 15659 + }, + { + "epoch": 0.1566, + "grad_norm": 1.3989181078805446, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 15660 + }, + { + "epoch": 0.15661, + "grad_norm": 1.1266789991409576, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 15661 + }, + { + "epoch": 0.15662, + "grad_norm": 1.25549306388821, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 15662 + }, + { + "epoch": 0.15663, + "grad_norm": 1.2024359146994443, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15663 + }, + { + "epoch": 0.15664, + "grad_norm": 1.4445860300332145, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 15664 + }, + { + "epoch": 0.15665, + "grad_norm": 1.0638694520822147, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15665 + }, + { + "epoch": 0.15666, + "grad_norm": 1.35890651645828, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 15666 + }, + { + "epoch": 0.15667, + "grad_norm": 1.1594542945395403, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 15667 + }, + { + "epoch": 0.15668, + "grad_norm": 1.3836444966834545, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 15668 + }, + { + "epoch": 0.15669, + "grad_norm": 1.2809819371574236, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 15669 + }, + { + "epoch": 0.1567, + "grad_norm": 1.1595670146662453, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 15670 + }, + { + "epoch": 0.15671, + "grad_norm": 1.4693245437056486, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 15671 + }, + { + "epoch": 0.15672, + "grad_norm": 0.9537301411473873, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 15672 + }, + { + "epoch": 0.15673, + "grad_norm": 1.2160253322792343, + "learning_rate": 0.003, + "loss": 4.036, + "step": 15673 + }, + { + "epoch": 0.15674, + "grad_norm": 1.1795293676711307, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 15674 + }, + { + "epoch": 0.15675, + "grad_norm": 1.311684259378266, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 15675 + }, + { + "epoch": 0.15676, + "grad_norm": 1.2582466749306869, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 15676 + }, + { + "epoch": 0.15677, + "grad_norm": 1.3351466316642864, + "learning_rate": 0.003, + "loss": 3.991, + "step": 15677 + }, + { + "epoch": 0.15678, + "grad_norm": 1.0633834566426348, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 15678 + }, + { + "epoch": 0.15679, + "grad_norm": 1.37332463707254, + "learning_rate": 0.003, + "loss": 4.028, + "step": 15679 + }, + { + "epoch": 0.1568, + "grad_norm": 1.1712721225087668, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 15680 + }, + { + "epoch": 0.15681, + "grad_norm": 1.3630506657556807, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 15681 + }, + { + "epoch": 0.15682, + "grad_norm": 1.2004379728704972, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 15682 + }, + { + "epoch": 0.15683, + "grad_norm": 1.2342667491771402, + "learning_rate": 0.003, + "loss": 4.009, + "step": 15683 + }, + { + "epoch": 0.15684, + "grad_norm": 1.1773798120343846, + "learning_rate": 0.003, + "loss": 4.025, + "step": 15684 + }, + { + "epoch": 0.15685, + "grad_norm": 1.0595683897554626, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 15685 + }, + { + "epoch": 0.15686, + "grad_norm": 1.3224476326633003, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 15686 + }, + { + "epoch": 0.15687, + "grad_norm": 1.191516704767411, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 15687 + }, + { + "epoch": 0.15688, + "grad_norm": 1.3277753161978578, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 15688 + }, + { + "epoch": 0.15689, + "grad_norm": 1.05958924159554, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 15689 + }, + { + "epoch": 0.1569, + "grad_norm": 1.4217155196174134, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 15690 + }, + { + "epoch": 0.15691, + "grad_norm": 1.075912120136982, + "learning_rate": 0.003, + "loss": 4.009, + "step": 15691 + }, + { + "epoch": 0.15692, + "grad_norm": 1.3995474722503682, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 15692 + }, + { + "epoch": 0.15693, + "grad_norm": 1.0650293846026575, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 15693 + }, + { + "epoch": 0.15694, + "grad_norm": 1.111432765077753, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 15694 + }, + { + "epoch": 0.15695, + "grad_norm": 1.2400496626879633, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 15695 + }, + { + "epoch": 0.15696, + "grad_norm": 1.2820035726545784, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 15696 + }, + { + "epoch": 0.15697, + "grad_norm": 1.144222892582624, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 15697 + }, + { + "epoch": 0.15698, + "grad_norm": 1.2560877918014903, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 15698 + }, + { + "epoch": 0.15699, + "grad_norm": 1.1004322491024552, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 15699 + }, + { + "epoch": 0.157, + "grad_norm": 1.1219445606257716, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 15700 + }, + { + "epoch": 0.15701, + "grad_norm": 1.2534425420426756, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 15701 + }, + { + "epoch": 0.15702, + "grad_norm": 1.0829170645941426, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 15702 + }, + { + "epoch": 0.15703, + "grad_norm": 1.361611697850092, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 15703 + }, + { + "epoch": 0.15704, + "grad_norm": 1.1696394963040575, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 15704 + }, + { + "epoch": 0.15705, + "grad_norm": 1.1887616996703583, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 15705 + }, + { + "epoch": 0.15706, + "grad_norm": 1.2745663354960037, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 15706 + }, + { + "epoch": 0.15707, + "grad_norm": 1.2381550064409779, + "learning_rate": 0.003, + "loss": 4.01, + "step": 15707 + }, + { + "epoch": 0.15708, + "grad_norm": 1.2739542548057263, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 15708 + }, + { + "epoch": 0.15709, + "grad_norm": 1.2398633286941458, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 15709 + }, + { + "epoch": 0.1571, + "grad_norm": 1.3097628322922201, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 15710 + }, + { + "epoch": 0.15711, + "grad_norm": 1.2036284641889141, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 15711 + }, + { + "epoch": 0.15712, + "grad_norm": 1.1727130198403828, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15712 + }, + { + "epoch": 0.15713, + "grad_norm": 1.3216356222953638, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 15713 + }, + { + "epoch": 0.15714, + "grad_norm": 1.230233431018454, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 15714 + }, + { + "epoch": 0.15715, + "grad_norm": 1.0479851957765023, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 15715 + }, + { + "epoch": 0.15716, + "grad_norm": 1.510663539638075, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 15716 + }, + { + "epoch": 0.15717, + "grad_norm": 1.003824020129898, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 15717 + }, + { + "epoch": 0.15718, + "grad_norm": 1.6092334246727755, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 15718 + }, + { + "epoch": 0.15719, + "grad_norm": 0.9666745117091241, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 15719 + }, + { + "epoch": 0.1572, + "grad_norm": 1.540739592405409, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 15720 + }, + { + "epoch": 0.15721, + "grad_norm": 1.0197508957957866, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 15721 + }, + { + "epoch": 0.15722, + "grad_norm": 1.4017671085031251, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 15722 + }, + { + "epoch": 0.15723, + "grad_norm": 1.410379383302518, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 15723 + }, + { + "epoch": 0.15724, + "grad_norm": 1.0326787094813932, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15724 + }, + { + "epoch": 0.15725, + "grad_norm": 1.4714017123876677, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 15725 + }, + { + "epoch": 0.15726, + "grad_norm": 1.1761253766406978, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 15726 + }, + { + "epoch": 0.15727, + "grad_norm": 1.1971053477586984, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 15727 + }, + { + "epoch": 0.15728, + "grad_norm": 1.202301682287537, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 15728 + }, + { + "epoch": 0.15729, + "grad_norm": 1.394061714539972, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 15729 + }, + { + "epoch": 0.1573, + "grad_norm": 1.0839884842598562, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 15730 + }, + { + "epoch": 0.15731, + "grad_norm": 1.3454603881216958, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 15731 + }, + { + "epoch": 0.15732, + "grad_norm": 1.1226471976758234, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 15732 + }, + { + "epoch": 0.15733, + "grad_norm": 1.341650590026742, + "learning_rate": 0.003, + "loss": 4.005, + "step": 15733 + }, + { + "epoch": 0.15734, + "grad_norm": 1.333745219797826, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 15734 + }, + { + "epoch": 0.15735, + "grad_norm": 1.3820529306632514, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 15735 + }, + { + "epoch": 0.15736, + "grad_norm": 1.075898810263639, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 15736 + }, + { + "epoch": 0.15737, + "grad_norm": 1.1927507040975849, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 15737 + }, + { + "epoch": 0.15738, + "grad_norm": 1.1560093494963364, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 15738 + }, + { + "epoch": 0.15739, + "grad_norm": 1.2765791949268857, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 15739 + }, + { + "epoch": 0.1574, + "grad_norm": 1.2261031464072705, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 15740 + }, + { + "epoch": 0.15741, + "grad_norm": 1.3856210456393485, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 15741 + }, + { + "epoch": 0.15742, + "grad_norm": 1.2308173545958936, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 15742 + }, + { + "epoch": 0.15743, + "grad_norm": 1.305667574962459, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 15743 + }, + { + "epoch": 0.15744, + "grad_norm": 1.122296256777786, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 15744 + }, + { + "epoch": 0.15745, + "grad_norm": 1.4580327525916033, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 15745 + }, + { + "epoch": 0.15746, + "grad_norm": 1.014207682133263, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 15746 + }, + { + "epoch": 0.15747, + "grad_norm": 1.2128222596163458, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 15747 + }, + { + "epoch": 0.15748, + "grad_norm": 1.618226744308094, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 15748 + }, + { + "epoch": 0.15749, + "grad_norm": 1.0926705269460224, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15749 + }, + { + "epoch": 0.1575, + "grad_norm": 1.3084121480238136, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 15750 + }, + { + "epoch": 0.15751, + "grad_norm": 1.0881441144059336, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 15751 + }, + { + "epoch": 0.15752, + "grad_norm": 1.3828878959169328, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 15752 + }, + { + "epoch": 0.15753, + "grad_norm": 1.0582213292468492, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 15753 + }, + { + "epoch": 0.15754, + "grad_norm": 1.2243364731940443, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 15754 + }, + { + "epoch": 0.15755, + "grad_norm": 0.9489447170472749, + "learning_rate": 0.003, + "loss": 3.999, + "step": 15755 + }, + { + "epoch": 0.15756, + "grad_norm": 1.3767375660460266, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 15756 + }, + { + "epoch": 0.15757, + "grad_norm": 1.1513102637885906, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 15757 + }, + { + "epoch": 0.15758, + "grad_norm": 1.2342191344271383, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 15758 + }, + { + "epoch": 0.15759, + "grad_norm": 1.3732495854092959, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 15759 + }, + { + "epoch": 0.1576, + "grad_norm": 1.1450681370573774, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 15760 + }, + { + "epoch": 0.15761, + "grad_norm": 1.1808603286300208, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 15761 + }, + { + "epoch": 0.15762, + "grad_norm": 1.3347176288852123, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 15762 + }, + { + "epoch": 0.15763, + "grad_norm": 1.1791546013231482, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 15763 + }, + { + "epoch": 0.15764, + "grad_norm": 1.4660199504829574, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 15764 + }, + { + "epoch": 0.15765, + "grad_norm": 1.1500449240770518, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 15765 + }, + { + "epoch": 0.15766, + "grad_norm": 1.3741568751174897, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 15766 + }, + { + "epoch": 0.15767, + "grad_norm": 1.1140070005830738, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 15767 + }, + { + "epoch": 0.15768, + "grad_norm": 1.257002863470519, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 15768 + }, + { + "epoch": 0.15769, + "grad_norm": 1.2053429049376951, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 15769 + }, + { + "epoch": 0.1577, + "grad_norm": 1.240049321747172, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 15770 + }, + { + "epoch": 0.15771, + "grad_norm": 1.1955814457159923, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15771 + }, + { + "epoch": 0.15772, + "grad_norm": 1.1340920104271954, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 15772 + }, + { + "epoch": 0.15773, + "grad_norm": 1.430938579515546, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 15773 + }, + { + "epoch": 0.15774, + "grad_norm": 1.1187858894078215, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 15774 + }, + { + "epoch": 0.15775, + "grad_norm": 1.6429846620387065, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 15775 + }, + { + "epoch": 0.15776, + "grad_norm": 0.9314203897015222, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 15776 + }, + { + "epoch": 0.15777, + "grad_norm": 1.3991092016747086, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 15777 + }, + { + "epoch": 0.15778, + "grad_norm": 1.2251718959934537, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 15778 + }, + { + "epoch": 0.15779, + "grad_norm": 1.3636857510201814, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 15779 + }, + { + "epoch": 0.1578, + "grad_norm": 1.185420826916895, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 15780 + }, + { + "epoch": 0.15781, + "grad_norm": 1.1484718610891285, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 15781 + }, + { + "epoch": 0.15782, + "grad_norm": 1.41598546417389, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 15782 + }, + { + "epoch": 0.15783, + "grad_norm": 1.138196369639764, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 15783 + }, + { + "epoch": 0.15784, + "grad_norm": 1.2565530652564345, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 15784 + }, + { + "epoch": 0.15785, + "grad_norm": 1.1563218182105854, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 15785 + }, + { + "epoch": 0.15786, + "grad_norm": 1.513960434660906, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 15786 + }, + { + "epoch": 0.15787, + "grad_norm": 1.1167315182601023, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15787 + }, + { + "epoch": 0.15788, + "grad_norm": 1.3727606902010836, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 15788 + }, + { + "epoch": 0.15789, + "grad_norm": 1.205202750661887, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 15789 + }, + { + "epoch": 0.1579, + "grad_norm": 1.2550885048186762, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 15790 + }, + { + "epoch": 0.15791, + "grad_norm": 1.0700136621537168, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 15791 + }, + { + "epoch": 0.15792, + "grad_norm": 1.3657975058002139, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 15792 + }, + { + "epoch": 0.15793, + "grad_norm": 1.3055020328555231, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 15793 + }, + { + "epoch": 0.15794, + "grad_norm": 1.1440251825531926, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 15794 + }, + { + "epoch": 0.15795, + "grad_norm": 1.612311812811049, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 15795 + }, + { + "epoch": 0.15796, + "grad_norm": 0.9678674348930872, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 15796 + }, + { + "epoch": 0.15797, + "grad_norm": 1.3041001041416083, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 15797 + }, + { + "epoch": 0.15798, + "grad_norm": 1.2287015904010035, + "learning_rate": 0.003, + "loss": 4.035, + "step": 15798 + }, + { + "epoch": 0.15799, + "grad_norm": 1.3510588695396477, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 15799 + }, + { + "epoch": 0.158, + "grad_norm": 1.2788150573835075, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 15800 + }, + { + "epoch": 0.15801, + "grad_norm": 1.0967642729913403, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 15801 + }, + { + "epoch": 0.15802, + "grad_norm": 1.4235868887732133, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 15802 + }, + { + "epoch": 0.15803, + "grad_norm": 1.0357255025069987, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 15803 + }, + { + "epoch": 0.15804, + "grad_norm": 1.3479991991782512, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 15804 + }, + { + "epoch": 0.15805, + "grad_norm": 1.2020102114084525, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 15805 + }, + { + "epoch": 0.15806, + "grad_norm": 1.2822871872220123, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 15806 + }, + { + "epoch": 0.15807, + "grad_norm": 1.1520001903352206, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 15807 + }, + { + "epoch": 0.15808, + "grad_norm": 1.0943066946968263, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 15808 + }, + { + "epoch": 0.15809, + "grad_norm": 1.301562984316627, + "learning_rate": 0.003, + "loss": 4.012, + "step": 15809 + }, + { + "epoch": 0.1581, + "grad_norm": 1.2776885758676504, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 15810 + }, + { + "epoch": 0.15811, + "grad_norm": 1.2606813015228042, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 15811 + }, + { + "epoch": 0.15812, + "grad_norm": 1.1857629716473423, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 15812 + }, + { + "epoch": 0.15813, + "grad_norm": 1.1636799706327965, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 15813 + }, + { + "epoch": 0.15814, + "grad_norm": 1.0337612537657892, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 15814 + }, + { + "epoch": 0.15815, + "grad_norm": 1.381683836345086, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 15815 + }, + { + "epoch": 0.15816, + "grad_norm": 1.1577125621268765, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 15816 + }, + { + "epoch": 0.15817, + "grad_norm": 1.5654365347116856, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 15817 + }, + { + "epoch": 0.15818, + "grad_norm": 0.9782645089506308, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 15818 + }, + { + "epoch": 0.15819, + "grad_norm": 1.2096831177291703, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 15819 + }, + { + "epoch": 0.1582, + "grad_norm": 1.1564805628697856, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 15820 + }, + { + "epoch": 0.15821, + "grad_norm": 1.263751306335342, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 15821 + }, + { + "epoch": 0.15822, + "grad_norm": 1.2472942820956714, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 15822 + }, + { + "epoch": 0.15823, + "grad_norm": 1.228267008633194, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 15823 + }, + { + "epoch": 0.15824, + "grad_norm": 1.4449510191156791, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 15824 + }, + { + "epoch": 0.15825, + "grad_norm": 1.1035898511648348, + "learning_rate": 0.003, + "loss": 4.026, + "step": 15825 + }, + { + "epoch": 0.15826, + "grad_norm": 1.5642574174563366, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 15826 + }, + { + "epoch": 0.15827, + "grad_norm": 0.9767703214885841, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 15827 + }, + { + "epoch": 0.15828, + "grad_norm": 1.4108549284549041, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 15828 + }, + { + "epoch": 0.15829, + "grad_norm": 1.377700822184997, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 15829 + }, + { + "epoch": 0.1583, + "grad_norm": 1.4692157047059422, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 15830 + }, + { + "epoch": 0.15831, + "grad_norm": 1.0706932402820692, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 15831 + }, + { + "epoch": 0.15832, + "grad_norm": 1.0949706715343381, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 15832 + }, + { + "epoch": 0.15833, + "grad_norm": 1.3196349499006816, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 15833 + }, + { + "epoch": 0.15834, + "grad_norm": 1.0917848610990246, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 15834 + }, + { + "epoch": 0.15835, + "grad_norm": 1.2928050134155717, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 15835 + }, + { + "epoch": 0.15836, + "grad_norm": 1.1200830203602705, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 15836 + }, + { + "epoch": 0.15837, + "grad_norm": 1.1344965900511697, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 15837 + }, + { + "epoch": 0.15838, + "grad_norm": 1.3132272233496367, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 15838 + }, + { + "epoch": 0.15839, + "grad_norm": 1.0778549547619092, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 15839 + }, + { + "epoch": 0.1584, + "grad_norm": 1.6529929221658228, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 15840 + }, + { + "epoch": 0.15841, + "grad_norm": 1.0816388554578351, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 15841 + }, + { + "epoch": 0.15842, + "grad_norm": 1.4517008573751615, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 15842 + }, + { + "epoch": 0.15843, + "grad_norm": 1.171938801187063, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15843 + }, + { + "epoch": 0.15844, + "grad_norm": 1.3534559256239043, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 15844 + }, + { + "epoch": 0.15845, + "grad_norm": 1.1834282268056617, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 15845 + }, + { + "epoch": 0.15846, + "grad_norm": 1.305954416737127, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 15846 + }, + { + "epoch": 0.15847, + "grad_norm": 1.0950390047460483, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 15847 + }, + { + "epoch": 0.15848, + "grad_norm": 1.3653974193756213, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 15848 + }, + { + "epoch": 0.15849, + "grad_norm": 1.1928352001024034, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 15849 + }, + { + "epoch": 0.1585, + "grad_norm": 1.3276197716502633, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 15850 + }, + { + "epoch": 0.15851, + "grad_norm": 1.2364669875438297, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 15851 + }, + { + "epoch": 0.15852, + "grad_norm": 1.257690636110746, + "learning_rate": 0.003, + "loss": 3.991, + "step": 15852 + }, + { + "epoch": 0.15853, + "grad_norm": 1.214967178374091, + "learning_rate": 0.003, + "loss": 4.002, + "step": 15853 + }, + { + "epoch": 0.15854, + "grad_norm": 1.4023811054543367, + "learning_rate": 0.003, + "loss": 3.996, + "step": 15854 + }, + { + "epoch": 0.15855, + "grad_norm": 1.1323216534833997, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 15855 + }, + { + "epoch": 0.15856, + "grad_norm": 1.2568759760812227, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 15856 + }, + { + "epoch": 0.15857, + "grad_norm": 1.0837214166550875, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 15857 + }, + { + "epoch": 0.15858, + "grad_norm": 1.136985798690449, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 15858 + }, + { + "epoch": 0.15859, + "grad_norm": 1.1450669993849945, + "learning_rate": 0.003, + "loss": 4.033, + "step": 15859 + }, + { + "epoch": 0.1586, + "grad_norm": 1.347547922002838, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 15860 + }, + { + "epoch": 0.15861, + "grad_norm": 1.0419008819958953, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 15861 + }, + { + "epoch": 0.15862, + "grad_norm": 1.413079565207084, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 15862 + }, + { + "epoch": 0.15863, + "grad_norm": 1.2743728315364327, + "learning_rate": 0.003, + "loss": 3.987, + "step": 15863 + }, + { + "epoch": 0.15864, + "grad_norm": 1.3169789757405919, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 15864 + }, + { + "epoch": 0.15865, + "grad_norm": 1.1232260543531478, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 15865 + }, + { + "epoch": 0.15866, + "grad_norm": 1.4253620354108139, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15866 + }, + { + "epoch": 0.15867, + "grad_norm": 1.3233768560673893, + "learning_rate": 0.003, + "loss": 4.017, + "step": 15867 + }, + { + "epoch": 0.15868, + "grad_norm": 1.3844891114669013, + "learning_rate": 0.003, + "loss": 3.987, + "step": 15868 + }, + { + "epoch": 0.15869, + "grad_norm": 1.063792568617661, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 15869 + }, + { + "epoch": 0.1587, + "grad_norm": 1.3356734637445744, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 15870 + }, + { + "epoch": 0.15871, + "grad_norm": 1.269230312806111, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 15871 + }, + { + "epoch": 0.15872, + "grad_norm": 1.5384143518783961, + "learning_rate": 0.003, + "loss": 4.032, + "step": 15872 + }, + { + "epoch": 0.15873, + "grad_norm": 1.107342231744455, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 15873 + }, + { + "epoch": 0.15874, + "grad_norm": 1.1973264901705576, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 15874 + }, + { + "epoch": 0.15875, + "grad_norm": 1.2196899595513655, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 15875 + }, + { + "epoch": 0.15876, + "grad_norm": 1.3007578819437007, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 15876 + }, + { + "epoch": 0.15877, + "grad_norm": 1.119209826200725, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 15877 + }, + { + "epoch": 0.15878, + "grad_norm": 1.4334426482819758, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 15878 + }, + { + "epoch": 0.15879, + "grad_norm": 1.067441178690247, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 15879 + }, + { + "epoch": 0.1588, + "grad_norm": 1.3339181131879854, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 15880 + }, + { + "epoch": 0.15881, + "grad_norm": 1.073006918871515, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 15881 + }, + { + "epoch": 0.15882, + "grad_norm": 1.1485741696188854, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 15882 + }, + { + "epoch": 0.15883, + "grad_norm": 1.2806079384299174, + "learning_rate": 0.003, + "loss": 3.997, + "step": 15883 + }, + { + "epoch": 0.15884, + "grad_norm": 1.230772825912422, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15884 + }, + { + "epoch": 0.15885, + "grad_norm": 1.3410068056507027, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 15885 + }, + { + "epoch": 0.15886, + "grad_norm": 1.1937651506835003, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 15886 + }, + { + "epoch": 0.15887, + "grad_norm": 1.3281232831561387, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 15887 + }, + { + "epoch": 0.15888, + "grad_norm": 1.3357080580184066, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 15888 + }, + { + "epoch": 0.15889, + "grad_norm": 1.450221699192516, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 15889 + }, + { + "epoch": 0.1589, + "grad_norm": 1.2063283535601301, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 15890 + }, + { + "epoch": 0.15891, + "grad_norm": 1.422104119354722, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 15891 + }, + { + "epoch": 0.15892, + "grad_norm": 1.1148680025652393, + "learning_rate": 0.003, + "loss": 4.001, + "step": 15892 + }, + { + "epoch": 0.15893, + "grad_norm": 1.2580714382446376, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 15893 + }, + { + "epoch": 0.15894, + "grad_norm": 1.1380887118267102, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 15894 + }, + { + "epoch": 0.15895, + "grad_norm": 1.4597095195310927, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15895 + }, + { + "epoch": 0.15896, + "grad_norm": 1.126224181908572, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 15896 + }, + { + "epoch": 0.15897, + "grad_norm": 1.2223812256140263, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 15897 + }, + { + "epoch": 0.15898, + "grad_norm": 1.270821791198787, + "learning_rate": 0.003, + "loss": 4.013, + "step": 15898 + }, + { + "epoch": 0.15899, + "grad_norm": 1.5603404791555529, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 15899 + }, + { + "epoch": 0.159, + "grad_norm": 1.3936087745617098, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 15900 + }, + { + "epoch": 0.15901, + "grad_norm": 1.0934864565818274, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 15901 + }, + { + "epoch": 0.15902, + "grad_norm": 1.1888037236970748, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 15902 + }, + { + "epoch": 0.15903, + "grad_norm": 1.2600477409961077, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15903 + }, + { + "epoch": 0.15904, + "grad_norm": 1.0810865586160259, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 15904 + }, + { + "epoch": 0.15905, + "grad_norm": 1.3802911177096677, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 15905 + }, + { + "epoch": 0.15906, + "grad_norm": 1.1233192397248655, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 15906 + }, + { + "epoch": 0.15907, + "grad_norm": 1.3272323602114968, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 15907 + }, + { + "epoch": 0.15908, + "grad_norm": 1.0506256131069907, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 15908 + }, + { + "epoch": 0.15909, + "grad_norm": 1.2115197839476954, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 15909 + }, + { + "epoch": 0.1591, + "grad_norm": 1.2868681085818645, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 15910 + }, + { + "epoch": 0.15911, + "grad_norm": 1.0897951917112798, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 15911 + }, + { + "epoch": 0.15912, + "grad_norm": 1.3936606217052407, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 15912 + }, + { + "epoch": 0.15913, + "grad_norm": 1.1805285692118155, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 15913 + }, + { + "epoch": 0.15914, + "grad_norm": 1.5115259210744625, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 15914 + }, + { + "epoch": 0.15915, + "grad_norm": 0.976486704028311, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 15915 + }, + { + "epoch": 0.15916, + "grad_norm": 1.5681460695365494, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 15916 + }, + { + "epoch": 0.15917, + "grad_norm": 1.1470059509439832, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 15917 + }, + { + "epoch": 0.15918, + "grad_norm": 1.3069453710337169, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15918 + }, + { + "epoch": 0.15919, + "grad_norm": 1.0735072095383267, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 15919 + }, + { + "epoch": 0.1592, + "grad_norm": 1.2391610510497388, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 15920 + }, + { + "epoch": 0.15921, + "grad_norm": 1.0875601848350542, + "learning_rate": 0.003, + "loss": 4.049, + "step": 15921 + }, + { + "epoch": 0.15922, + "grad_norm": 1.4961613661364455, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 15922 + }, + { + "epoch": 0.15923, + "grad_norm": 1.198264842703091, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 15923 + }, + { + "epoch": 0.15924, + "grad_norm": 1.251445635848106, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 15924 + }, + { + "epoch": 0.15925, + "grad_norm": 1.0844267077040648, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 15925 + }, + { + "epoch": 0.15926, + "grad_norm": 1.5594659925155585, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 15926 + }, + { + "epoch": 0.15927, + "grad_norm": 1.0809068769756605, + "learning_rate": 0.003, + "loss": 3.98, + "step": 15927 + }, + { + "epoch": 0.15928, + "grad_norm": 1.4555096692438068, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 15928 + }, + { + "epoch": 0.15929, + "grad_norm": 1.1285997467928173, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 15929 + }, + { + "epoch": 0.1593, + "grad_norm": 1.3324282147057867, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 15930 + }, + { + "epoch": 0.15931, + "grad_norm": 1.2380611666099195, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 15931 + }, + { + "epoch": 0.15932, + "grad_norm": 1.1506683674394433, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 15932 + }, + { + "epoch": 0.15933, + "grad_norm": 1.2834021241100766, + "learning_rate": 0.003, + "loss": 4.0, + "step": 15933 + }, + { + "epoch": 0.15934, + "grad_norm": 1.2501225263853923, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 15934 + }, + { + "epoch": 0.15935, + "grad_norm": 1.2661937164721981, + "learning_rate": 0.003, + "loss": 4.019, + "step": 15935 + }, + { + "epoch": 0.15936, + "grad_norm": 1.2736002273731495, + "learning_rate": 0.003, + "loss": 4.008, + "step": 15936 + }, + { + "epoch": 0.15937, + "grad_norm": 1.064288489544132, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 15937 + }, + { + "epoch": 0.15938, + "grad_norm": 1.2084424600212056, + "learning_rate": 0.003, + "loss": 4.001, + "step": 15938 + }, + { + "epoch": 0.15939, + "grad_norm": 1.3309803366665889, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15939 + }, + { + "epoch": 0.1594, + "grad_norm": 1.032362097892044, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 15940 + }, + { + "epoch": 0.15941, + "grad_norm": 1.2671311295721919, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 15941 + }, + { + "epoch": 0.15942, + "grad_norm": 1.0980979707465504, + "learning_rate": 0.003, + "loss": 3.974, + "step": 15942 + }, + { + "epoch": 0.15943, + "grad_norm": 1.2788912070867908, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 15943 + }, + { + "epoch": 0.15944, + "grad_norm": 1.3409571028024285, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 15944 + }, + { + "epoch": 0.15945, + "grad_norm": 1.250160402820269, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 15945 + }, + { + "epoch": 0.15946, + "grad_norm": 1.084368530449594, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 15946 + }, + { + "epoch": 0.15947, + "grad_norm": 1.3489970350127387, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 15947 + }, + { + "epoch": 0.15948, + "grad_norm": 1.2994259205767582, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 15948 + }, + { + "epoch": 0.15949, + "grad_norm": 1.2973830305893017, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 15949 + }, + { + "epoch": 0.1595, + "grad_norm": 1.2073403007883003, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 15950 + }, + { + "epoch": 0.15951, + "grad_norm": 1.2077527594878215, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 15951 + }, + { + "epoch": 0.15952, + "grad_norm": 1.1281748449482183, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 15952 + }, + { + "epoch": 0.15953, + "grad_norm": 1.2386195576290036, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 15953 + }, + { + "epoch": 0.15954, + "grad_norm": 1.3117440639263585, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 15954 + }, + { + "epoch": 0.15955, + "grad_norm": 1.1144295795884482, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 15955 + }, + { + "epoch": 0.15956, + "grad_norm": 1.3513662104221202, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 15956 + }, + { + "epoch": 0.15957, + "grad_norm": 1.1644768303251782, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 15957 + }, + { + "epoch": 0.15958, + "grad_norm": 1.5178654834472558, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 15958 + }, + { + "epoch": 0.15959, + "grad_norm": 1.009003208684371, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 15959 + }, + { + "epoch": 0.1596, + "grad_norm": 1.747584605838126, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 15960 + }, + { + "epoch": 0.15961, + "grad_norm": 1.2904165960056244, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 15961 + }, + { + "epoch": 0.15962, + "grad_norm": 1.2051900969929517, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 15962 + }, + { + "epoch": 0.15963, + "grad_norm": 1.3437324584870094, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 15963 + }, + { + "epoch": 0.15964, + "grad_norm": 1.0610913888424898, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 15964 + }, + { + "epoch": 0.15965, + "grad_norm": 1.1370997544613122, + "learning_rate": 0.003, + "loss": 4.024, + "step": 15965 + }, + { + "epoch": 0.15966, + "grad_norm": 1.311349509904151, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 15966 + }, + { + "epoch": 0.15967, + "grad_norm": 1.2046874971861148, + "learning_rate": 0.003, + "loss": 4.012, + "step": 15967 + }, + { + "epoch": 0.15968, + "grad_norm": 1.32732408381076, + "learning_rate": 0.003, + "loss": 4.02, + "step": 15968 + }, + { + "epoch": 0.15969, + "grad_norm": 1.0236651951170457, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 15969 + }, + { + "epoch": 0.1597, + "grad_norm": 1.2998798579086002, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 15970 + }, + { + "epoch": 0.15971, + "grad_norm": 1.059402665772283, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 15971 + }, + { + "epoch": 0.15972, + "grad_norm": 1.2886639535877384, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 15972 + }, + { + "epoch": 0.15973, + "grad_norm": 1.4074638650887707, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 15973 + }, + { + "epoch": 0.15974, + "grad_norm": 1.034321743162951, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 15974 + }, + { + "epoch": 0.15975, + "grad_norm": 1.5559879296046673, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 15975 + }, + { + "epoch": 0.15976, + "grad_norm": 0.9732904162293462, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 15976 + }, + { + "epoch": 0.15977, + "grad_norm": 1.2254248219761825, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 15977 + }, + { + "epoch": 0.15978, + "grad_norm": 1.1405023127368015, + "learning_rate": 0.003, + "loss": 4.017, + "step": 15978 + }, + { + "epoch": 0.15979, + "grad_norm": 1.3818040951816377, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 15979 + }, + { + "epoch": 0.1598, + "grad_norm": 1.1749784241455272, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 15980 + }, + { + "epoch": 0.15981, + "grad_norm": 1.4577035816327795, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15981 + }, + { + "epoch": 0.15982, + "grad_norm": 1.23743809212499, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 15982 + }, + { + "epoch": 0.15983, + "grad_norm": 1.3583355235631573, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 15983 + }, + { + "epoch": 0.15984, + "grad_norm": 1.1240289417059142, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 15984 + }, + { + "epoch": 0.15985, + "grad_norm": 1.4041849938566662, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 15985 + }, + { + "epoch": 0.15986, + "grad_norm": 1.3684400747095355, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 15986 + }, + { + "epoch": 0.15987, + "grad_norm": 1.210295400952199, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 15987 + }, + { + "epoch": 0.15988, + "grad_norm": 1.557409213131596, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 15988 + }, + { + "epoch": 0.15989, + "grad_norm": 1.2080860727995004, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 15989 + }, + { + "epoch": 0.1599, + "grad_norm": 1.12769210010326, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 15990 + }, + { + "epoch": 0.15991, + "grad_norm": 1.3193033992992307, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 15991 + }, + { + "epoch": 0.15992, + "grad_norm": 1.1530373535937724, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 15992 + }, + { + "epoch": 0.15993, + "grad_norm": 1.3786125097767274, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15993 + }, + { + "epoch": 0.15994, + "grad_norm": 1.01651306819352, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 15994 + }, + { + "epoch": 0.15995, + "grad_norm": 1.4312493173148968, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 15995 + }, + { + "epoch": 0.15996, + "grad_norm": 0.9999668997559916, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 15996 + }, + { + "epoch": 0.15997, + "grad_norm": 1.3035104354159652, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 15997 + }, + { + "epoch": 0.15998, + "grad_norm": 1.100398140602815, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 15998 + }, + { + "epoch": 0.15999, + "grad_norm": 1.3803552999940325, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 15999 + }, + { + "epoch": 0.16, + "grad_norm": 1.1605678043519492, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 16000 + }, + { + "epoch": 0.16001, + "grad_norm": 1.2404721333235633, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16001 + }, + { + "epoch": 0.16002, + "grad_norm": 1.2338390822554808, + "learning_rate": 0.003, + "loss": 4.007, + "step": 16002 + }, + { + "epoch": 0.16003, + "grad_norm": 1.1938666541910892, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 16003 + }, + { + "epoch": 0.16004, + "grad_norm": 1.2229940449205243, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 16004 + }, + { + "epoch": 0.16005, + "grad_norm": 1.5471699933630838, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 16005 + }, + { + "epoch": 0.16006, + "grad_norm": 1.1594371365697964, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 16006 + }, + { + "epoch": 0.16007, + "grad_norm": 1.5214616998953365, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 16007 + }, + { + "epoch": 0.16008, + "grad_norm": 1.134113177523023, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 16008 + }, + { + "epoch": 0.16009, + "grad_norm": 1.2947310732024733, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 16009 + }, + { + "epoch": 0.1601, + "grad_norm": 1.3787130295348902, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 16010 + }, + { + "epoch": 0.16011, + "grad_norm": 1.2233450040254008, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 16011 + }, + { + "epoch": 0.16012, + "grad_norm": 1.188757317575267, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 16012 + }, + { + "epoch": 0.16013, + "grad_norm": 1.2802531708167952, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 16013 + }, + { + "epoch": 0.16014, + "grad_norm": 1.3036325770857242, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 16014 + }, + { + "epoch": 0.16015, + "grad_norm": 1.3075081535296704, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 16015 + }, + { + "epoch": 0.16016, + "grad_norm": 1.2868315758360842, + "learning_rate": 0.003, + "loss": 3.994, + "step": 16016 + }, + { + "epoch": 0.16017, + "grad_norm": 1.2192453725004495, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 16017 + }, + { + "epoch": 0.16018, + "grad_norm": 1.4179947259720587, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 16018 + }, + { + "epoch": 0.16019, + "grad_norm": 0.9933268676060425, + "learning_rate": 0.003, + "loss": 3.972, + "step": 16019 + }, + { + "epoch": 0.1602, + "grad_norm": 1.2677574055498495, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 16020 + }, + { + "epoch": 0.16021, + "grad_norm": 1.04914501921573, + "learning_rate": 0.003, + "loss": 4.001, + "step": 16021 + }, + { + "epoch": 0.16022, + "grad_norm": 1.5658787036199897, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 16022 + }, + { + "epoch": 0.16023, + "grad_norm": 0.9380166563877902, + "learning_rate": 0.003, + "loss": 4.007, + "step": 16023 + }, + { + "epoch": 0.16024, + "grad_norm": 1.3563993540979358, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 16024 + }, + { + "epoch": 0.16025, + "grad_norm": 1.2223720877097546, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 16025 + }, + { + "epoch": 0.16026, + "grad_norm": 1.3958468360337928, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 16026 + }, + { + "epoch": 0.16027, + "grad_norm": 1.1496040425392546, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 16027 + }, + { + "epoch": 0.16028, + "grad_norm": 1.265978581286196, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 16028 + }, + { + "epoch": 0.16029, + "grad_norm": 1.1970656146860956, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 16029 + }, + { + "epoch": 0.1603, + "grad_norm": 1.3719632920241611, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 16030 + }, + { + "epoch": 0.16031, + "grad_norm": 1.4888979743666881, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 16031 + }, + { + "epoch": 0.16032, + "grad_norm": 0.9405753892538025, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 16032 + }, + { + "epoch": 0.16033, + "grad_norm": 1.3972562621926694, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 16033 + }, + { + "epoch": 0.16034, + "grad_norm": 1.2076091171066374, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 16034 + }, + { + "epoch": 0.16035, + "grad_norm": 1.2152831213427373, + "learning_rate": 0.003, + "loss": 4.026, + "step": 16035 + }, + { + "epoch": 0.16036, + "grad_norm": 1.323796096636007, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 16036 + }, + { + "epoch": 0.16037, + "grad_norm": 1.0753704479667336, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 16037 + }, + { + "epoch": 0.16038, + "grad_norm": 1.1599005027463496, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 16038 + }, + { + "epoch": 0.16039, + "grad_norm": 1.2709321988871976, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 16039 + }, + { + "epoch": 0.1604, + "grad_norm": 1.3670081025524143, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 16040 + }, + { + "epoch": 0.16041, + "grad_norm": 0.9935916137257558, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 16041 + }, + { + "epoch": 0.16042, + "grad_norm": 1.1991884036840277, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 16042 + }, + { + "epoch": 0.16043, + "grad_norm": 1.2122572540792345, + "learning_rate": 0.003, + "loss": 4.011, + "step": 16043 + }, + { + "epoch": 0.16044, + "grad_norm": 1.051149596266568, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 16044 + }, + { + "epoch": 0.16045, + "grad_norm": 1.2696961833905007, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16045 + }, + { + "epoch": 0.16046, + "grad_norm": 1.136441039790602, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 16046 + }, + { + "epoch": 0.16047, + "grad_norm": 1.2746520841443807, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 16047 + }, + { + "epoch": 0.16048, + "grad_norm": 1.3310326798956609, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 16048 + }, + { + "epoch": 0.16049, + "grad_norm": 1.1906862885352336, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 16049 + }, + { + "epoch": 0.1605, + "grad_norm": 1.4284071734558204, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 16050 + }, + { + "epoch": 0.16051, + "grad_norm": 1.1554474651545776, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 16051 + }, + { + "epoch": 0.16052, + "grad_norm": 1.6483568951641105, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16052 + }, + { + "epoch": 0.16053, + "grad_norm": 1.050790920334589, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 16053 + }, + { + "epoch": 0.16054, + "grad_norm": 1.1913184336153997, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 16054 + }, + { + "epoch": 0.16055, + "grad_norm": 1.299799165461864, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 16055 + }, + { + "epoch": 0.16056, + "grad_norm": 1.2197434747744236, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 16056 + }, + { + "epoch": 0.16057, + "grad_norm": 1.3485521233853122, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 16057 + }, + { + "epoch": 0.16058, + "grad_norm": 1.2529999632379034, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 16058 + }, + { + "epoch": 0.16059, + "grad_norm": 1.2757609779826649, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 16059 + }, + { + "epoch": 0.1606, + "grad_norm": 1.2624204085329382, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 16060 + }, + { + "epoch": 0.16061, + "grad_norm": 1.01916293730141, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 16061 + }, + { + "epoch": 0.16062, + "grad_norm": 1.3918647677121883, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 16062 + }, + { + "epoch": 0.16063, + "grad_norm": 1.1299985681926117, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16063 + }, + { + "epoch": 0.16064, + "grad_norm": 1.3989755290151158, + "learning_rate": 0.003, + "loss": 4.041, + "step": 16064 + }, + { + "epoch": 0.16065, + "grad_norm": 1.1444329874813979, + "learning_rate": 0.003, + "loss": 4.01, + "step": 16065 + }, + { + "epoch": 0.16066, + "grad_norm": 1.2769218025740188, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 16066 + }, + { + "epoch": 0.16067, + "grad_norm": 1.2657726384704442, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 16067 + }, + { + "epoch": 0.16068, + "grad_norm": 1.3584528947680767, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 16068 + }, + { + "epoch": 0.16069, + "grad_norm": 1.3849230623983817, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16069 + }, + { + "epoch": 0.1607, + "grad_norm": 1.1612567846070265, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 16070 + }, + { + "epoch": 0.16071, + "grad_norm": 0.9850898731815557, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 16071 + }, + { + "epoch": 0.16072, + "grad_norm": 1.3979078076442122, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 16072 + }, + { + "epoch": 0.16073, + "grad_norm": 1.1829207947805287, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16073 + }, + { + "epoch": 0.16074, + "grad_norm": 1.3956568785676735, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 16074 + }, + { + "epoch": 0.16075, + "grad_norm": 1.2601557349528247, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 16075 + }, + { + "epoch": 0.16076, + "grad_norm": 1.175950105019326, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 16076 + }, + { + "epoch": 0.16077, + "grad_norm": 1.2399943688853072, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 16077 + }, + { + "epoch": 0.16078, + "grad_norm": 1.38283249365418, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 16078 + }, + { + "epoch": 0.16079, + "grad_norm": 1.0971987983802745, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 16079 + }, + { + "epoch": 0.1608, + "grad_norm": 1.300851282510104, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 16080 + }, + { + "epoch": 0.16081, + "grad_norm": 1.095857488943779, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 16081 + }, + { + "epoch": 0.16082, + "grad_norm": 1.299630758811905, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 16082 + }, + { + "epoch": 0.16083, + "grad_norm": 1.0260005769947862, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 16083 + }, + { + "epoch": 0.16084, + "grad_norm": 1.3001667629050337, + "learning_rate": 0.003, + "loss": 4.029, + "step": 16084 + }, + { + "epoch": 0.16085, + "grad_norm": 1.4688412215739906, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 16085 + }, + { + "epoch": 0.16086, + "grad_norm": 1.1117184144622974, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 16086 + }, + { + "epoch": 0.16087, + "grad_norm": 1.370685907343411, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 16087 + }, + { + "epoch": 0.16088, + "grad_norm": 1.202322765052148, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16088 + }, + { + "epoch": 0.16089, + "grad_norm": 1.2356402782688343, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 16089 + }, + { + "epoch": 0.1609, + "grad_norm": 1.1182959162689072, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 16090 + }, + { + "epoch": 0.16091, + "grad_norm": 1.2642362698117995, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 16091 + }, + { + "epoch": 0.16092, + "grad_norm": 1.2352144019840015, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 16092 + }, + { + "epoch": 0.16093, + "grad_norm": 1.076256655348046, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 16093 + }, + { + "epoch": 0.16094, + "grad_norm": 1.3548711640288833, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16094 + }, + { + "epoch": 0.16095, + "grad_norm": 1.5575501903390845, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 16095 + }, + { + "epoch": 0.16096, + "grad_norm": 0.9637468052786332, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 16096 + }, + { + "epoch": 0.16097, + "grad_norm": 1.4523002755247842, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 16097 + }, + { + "epoch": 0.16098, + "grad_norm": 1.0730202043900385, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 16098 + }, + { + "epoch": 0.16099, + "grad_norm": 1.3524062057099446, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 16099 + }, + { + "epoch": 0.161, + "grad_norm": 1.1546396738849403, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 16100 + }, + { + "epoch": 0.16101, + "grad_norm": 1.5156718082592469, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 16101 + }, + { + "epoch": 0.16102, + "grad_norm": 1.1864275339496917, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 16102 + }, + { + "epoch": 0.16103, + "grad_norm": 1.15265366212102, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 16103 + }, + { + "epoch": 0.16104, + "grad_norm": 1.3678819136882205, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 16104 + }, + { + "epoch": 0.16105, + "grad_norm": 1.1852836254515984, + "learning_rate": 0.003, + "loss": 4.004, + "step": 16105 + }, + { + "epoch": 0.16106, + "grad_norm": 1.2642708883562752, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16106 + }, + { + "epoch": 0.16107, + "grad_norm": 1.4118138366225041, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 16107 + }, + { + "epoch": 0.16108, + "grad_norm": 1.2442807137062035, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 16108 + }, + { + "epoch": 0.16109, + "grad_norm": 1.3907189421395418, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 16109 + }, + { + "epoch": 0.1611, + "grad_norm": 1.3454169707594963, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 16110 + }, + { + "epoch": 0.16111, + "grad_norm": 1.1334365616559023, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16111 + }, + { + "epoch": 0.16112, + "grad_norm": 1.223835544837327, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 16112 + }, + { + "epoch": 0.16113, + "grad_norm": 1.1003578422714257, + "learning_rate": 0.003, + "loss": 4.024, + "step": 16113 + }, + { + "epoch": 0.16114, + "grad_norm": 1.422400823040409, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 16114 + }, + { + "epoch": 0.16115, + "grad_norm": 1.0493058061084093, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16115 + }, + { + "epoch": 0.16116, + "grad_norm": 1.3919970554520462, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 16116 + }, + { + "epoch": 0.16117, + "grad_norm": 1.0525169085581814, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 16117 + }, + { + "epoch": 0.16118, + "grad_norm": 1.5199701224654893, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 16118 + }, + { + "epoch": 0.16119, + "grad_norm": 1.1893543622240004, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 16119 + }, + { + "epoch": 0.1612, + "grad_norm": 1.481317893128953, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 16120 + }, + { + "epoch": 0.16121, + "grad_norm": 1.3689173482711632, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 16121 + }, + { + "epoch": 0.16122, + "grad_norm": 1.3670484911413183, + "learning_rate": 0.003, + "loss": 4.037, + "step": 16122 + }, + { + "epoch": 0.16123, + "grad_norm": 1.2552896372472673, + "learning_rate": 0.003, + "loss": 4.046, + "step": 16123 + }, + { + "epoch": 0.16124, + "grad_norm": 1.3335344033573147, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 16124 + }, + { + "epoch": 0.16125, + "grad_norm": 1.1700116034864976, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 16125 + }, + { + "epoch": 0.16126, + "grad_norm": 1.1471011013024555, + "learning_rate": 0.003, + "loss": 4.005, + "step": 16126 + }, + { + "epoch": 0.16127, + "grad_norm": 1.1549395887326956, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 16127 + }, + { + "epoch": 0.16128, + "grad_norm": 1.189703932631664, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 16128 + }, + { + "epoch": 0.16129, + "grad_norm": 1.3082653225319467, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 16129 + }, + { + "epoch": 0.1613, + "grad_norm": 1.2416398957273194, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 16130 + }, + { + "epoch": 0.16131, + "grad_norm": 1.4663947896272986, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 16131 + }, + { + "epoch": 0.16132, + "grad_norm": 1.2135324744775915, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 16132 + }, + { + "epoch": 0.16133, + "grad_norm": 1.1562967314921637, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 16133 + }, + { + "epoch": 0.16134, + "grad_norm": 1.1763928022748928, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16134 + }, + { + "epoch": 0.16135, + "grad_norm": 1.2976195079820243, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 16135 + }, + { + "epoch": 0.16136, + "grad_norm": 1.1471018842279348, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 16136 + }, + { + "epoch": 0.16137, + "grad_norm": 1.2413117497347381, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16137 + }, + { + "epoch": 0.16138, + "grad_norm": 0.9923693305364467, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 16138 + }, + { + "epoch": 0.16139, + "grad_norm": 1.4381837952111263, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 16139 + }, + { + "epoch": 0.1614, + "grad_norm": 0.9472733316173573, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 16140 + }, + { + "epoch": 0.16141, + "grad_norm": 1.33400120176943, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 16141 + }, + { + "epoch": 0.16142, + "grad_norm": 1.2127378779303366, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 16142 + }, + { + "epoch": 0.16143, + "grad_norm": 1.4219686336200243, + "learning_rate": 0.003, + "loss": 4.01, + "step": 16143 + }, + { + "epoch": 0.16144, + "grad_norm": 1.39719751895515, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 16144 + }, + { + "epoch": 0.16145, + "grad_norm": 1.1694230089176398, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 16145 + }, + { + "epoch": 0.16146, + "grad_norm": 1.3097712187358037, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 16146 + }, + { + "epoch": 0.16147, + "grad_norm": 1.1227197415916592, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 16147 + }, + { + "epoch": 0.16148, + "grad_norm": 1.1213560161179337, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 16148 + }, + { + "epoch": 0.16149, + "grad_norm": 1.3743437454563598, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 16149 + }, + { + "epoch": 0.1615, + "grad_norm": 1.2906143546730646, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 16150 + }, + { + "epoch": 0.16151, + "grad_norm": 1.3214154585018147, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 16151 + }, + { + "epoch": 0.16152, + "grad_norm": 1.0230395562856147, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 16152 + }, + { + "epoch": 0.16153, + "grad_norm": 1.3229402182409677, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 16153 + }, + { + "epoch": 0.16154, + "grad_norm": 1.2864381791192716, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 16154 + }, + { + "epoch": 0.16155, + "grad_norm": 1.0760908943022593, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 16155 + }, + { + "epoch": 0.16156, + "grad_norm": 1.2856867476042058, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 16156 + }, + { + "epoch": 0.16157, + "grad_norm": 1.105141525818919, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 16157 + }, + { + "epoch": 0.16158, + "grad_norm": 1.3770087818300023, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 16158 + }, + { + "epoch": 0.16159, + "grad_norm": 1.0357188749854245, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 16159 + }, + { + "epoch": 0.1616, + "grad_norm": 1.4336389701380712, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16160 + }, + { + "epoch": 0.16161, + "grad_norm": 1.2736841522738496, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 16161 + }, + { + "epoch": 0.16162, + "grad_norm": 1.2190076877210108, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 16162 + }, + { + "epoch": 0.16163, + "grad_norm": 1.1169781715691094, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 16163 + }, + { + "epoch": 0.16164, + "grad_norm": 1.2873936761081572, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 16164 + }, + { + "epoch": 0.16165, + "grad_norm": 1.0721605161117134, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 16165 + }, + { + "epoch": 0.16166, + "grad_norm": 1.5210271055832423, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 16166 + }, + { + "epoch": 0.16167, + "grad_norm": 1.2603117087981106, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 16167 + }, + { + "epoch": 0.16168, + "grad_norm": 1.574483052352801, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 16168 + }, + { + "epoch": 0.16169, + "grad_norm": 0.8664294582071033, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 16169 + }, + { + "epoch": 0.1617, + "grad_norm": 0.9914409134469694, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 16170 + }, + { + "epoch": 0.16171, + "grad_norm": 1.4567984425664773, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 16171 + }, + { + "epoch": 0.16172, + "grad_norm": 1.2118153987214155, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 16172 + }, + { + "epoch": 0.16173, + "grad_norm": 1.1642355951845538, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 16173 + }, + { + "epoch": 0.16174, + "grad_norm": 1.277516708904266, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 16174 + }, + { + "epoch": 0.16175, + "grad_norm": 1.1793587223376114, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 16175 + }, + { + "epoch": 0.16176, + "grad_norm": 1.3705621332644309, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16176 + }, + { + "epoch": 0.16177, + "grad_norm": 0.9276949023792542, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 16177 + }, + { + "epoch": 0.16178, + "grad_norm": 1.41636566857606, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 16178 + }, + { + "epoch": 0.16179, + "grad_norm": 1.0679985032966415, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 16179 + }, + { + "epoch": 0.1618, + "grad_norm": 1.5060257983082257, + "learning_rate": 0.003, + "loss": 4.021, + "step": 16180 + }, + { + "epoch": 0.16181, + "grad_norm": 1.0021990238480751, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 16181 + }, + { + "epoch": 0.16182, + "grad_norm": 1.3496364921628858, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 16182 + }, + { + "epoch": 0.16183, + "grad_norm": 1.3215118120120168, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 16183 + }, + { + "epoch": 0.16184, + "grad_norm": 0.9261056839001097, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 16184 + }, + { + "epoch": 0.16185, + "grad_norm": 1.2410333950635242, + "learning_rate": 0.003, + "loss": 4.044, + "step": 16185 + }, + { + "epoch": 0.16186, + "grad_norm": 1.2122607787387762, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 16186 + }, + { + "epoch": 0.16187, + "grad_norm": 1.074138998508649, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 16187 + }, + { + "epoch": 0.16188, + "grad_norm": 1.3439154528218888, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 16188 + }, + { + "epoch": 0.16189, + "grad_norm": 1.2327094920539634, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 16189 + }, + { + "epoch": 0.1619, + "grad_norm": 1.2945842411187498, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 16190 + }, + { + "epoch": 0.16191, + "grad_norm": 1.1839140649632616, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 16191 + }, + { + "epoch": 0.16192, + "grad_norm": 1.3758659634642418, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 16192 + }, + { + "epoch": 0.16193, + "grad_norm": 1.5754222985713842, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 16193 + }, + { + "epoch": 0.16194, + "grad_norm": 1.168003610994215, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 16194 + }, + { + "epoch": 0.16195, + "grad_norm": 1.2598666586763887, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16195 + }, + { + "epoch": 0.16196, + "grad_norm": 1.2484636707932206, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 16196 + }, + { + "epoch": 0.16197, + "grad_norm": 1.120079978131868, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 16197 + }, + { + "epoch": 0.16198, + "grad_norm": 1.4012787009699093, + "learning_rate": 0.003, + "loss": 4.01, + "step": 16198 + }, + { + "epoch": 0.16199, + "grad_norm": 1.240228254694417, + "learning_rate": 0.003, + "loss": 4.013, + "step": 16199 + }, + { + "epoch": 0.162, + "grad_norm": 1.1772413535073507, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 16200 + }, + { + "epoch": 0.16201, + "grad_norm": 1.2900263087529662, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 16201 + }, + { + "epoch": 0.16202, + "grad_norm": 1.257610807599335, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16202 + }, + { + "epoch": 0.16203, + "grad_norm": 1.175552574679173, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 16203 + }, + { + "epoch": 0.16204, + "grad_norm": 1.4500182220732118, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 16204 + }, + { + "epoch": 0.16205, + "grad_norm": 1.1768076652887332, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 16205 + }, + { + "epoch": 0.16206, + "grad_norm": 1.3175082599762675, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 16206 + }, + { + "epoch": 0.16207, + "grad_norm": 1.083670859157658, + "learning_rate": 0.003, + "loss": 4.021, + "step": 16207 + }, + { + "epoch": 0.16208, + "grad_norm": 1.3829291622680255, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 16208 + }, + { + "epoch": 0.16209, + "grad_norm": 1.1075036833669838, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 16209 + }, + { + "epoch": 0.1621, + "grad_norm": 1.4023288738470112, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 16210 + }, + { + "epoch": 0.16211, + "grad_norm": 1.0568643946962788, + "learning_rate": 0.003, + "loss": 4.003, + "step": 16211 + }, + { + "epoch": 0.16212, + "grad_norm": 1.2822784907103308, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 16212 + }, + { + "epoch": 0.16213, + "grad_norm": 1.0865919739046765, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 16213 + }, + { + "epoch": 0.16214, + "grad_norm": 1.1462534560610254, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 16214 + }, + { + "epoch": 0.16215, + "grad_norm": 1.4350052523536598, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 16215 + }, + { + "epoch": 0.16216, + "grad_norm": 1.2013093101179828, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 16216 + }, + { + "epoch": 0.16217, + "grad_norm": 1.215098202740069, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 16217 + }, + { + "epoch": 0.16218, + "grad_norm": 1.5608825590690358, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 16218 + }, + { + "epoch": 0.16219, + "grad_norm": 0.979091370861085, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 16219 + }, + { + "epoch": 0.1622, + "grad_norm": 1.4947609668871131, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 16220 + }, + { + "epoch": 0.16221, + "grad_norm": 1.0998543006615162, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 16221 + }, + { + "epoch": 0.16222, + "grad_norm": 1.2523996094659415, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 16222 + }, + { + "epoch": 0.16223, + "grad_norm": 1.2217073500770703, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 16223 + }, + { + "epoch": 0.16224, + "grad_norm": 1.5546837574704633, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 16224 + }, + { + "epoch": 0.16225, + "grad_norm": 1.2099323804351048, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 16225 + }, + { + "epoch": 0.16226, + "grad_norm": 1.1681462870412578, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 16226 + }, + { + "epoch": 0.16227, + "grad_norm": 1.2745384192534517, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16227 + }, + { + "epoch": 0.16228, + "grad_norm": 1.2910726751301218, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 16228 + }, + { + "epoch": 0.16229, + "grad_norm": 1.232591354216802, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 16229 + }, + { + "epoch": 0.1623, + "grad_norm": 1.1829997005740684, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 16230 + }, + { + "epoch": 0.16231, + "grad_norm": 1.3163228024010354, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16231 + }, + { + "epoch": 0.16232, + "grad_norm": 1.233936463295609, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 16232 + }, + { + "epoch": 0.16233, + "grad_norm": 1.2132889495425785, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 16233 + }, + { + "epoch": 0.16234, + "grad_norm": 1.2822756762471483, + "learning_rate": 0.003, + "loss": 4.024, + "step": 16234 + }, + { + "epoch": 0.16235, + "grad_norm": 1.0788167153224901, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 16235 + }, + { + "epoch": 0.16236, + "grad_norm": 1.3068920944816198, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16236 + }, + { + "epoch": 0.16237, + "grad_norm": 1.1490896028469753, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 16237 + }, + { + "epoch": 0.16238, + "grad_norm": 1.4329996245008865, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 16238 + }, + { + "epoch": 0.16239, + "grad_norm": 1.0738850110421332, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 16239 + }, + { + "epoch": 0.1624, + "grad_norm": 1.4199152589907686, + "learning_rate": 0.003, + "loss": 4.011, + "step": 16240 + }, + { + "epoch": 0.16241, + "grad_norm": 1.1707238634143304, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 16241 + }, + { + "epoch": 0.16242, + "grad_norm": 1.1452780655232648, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16242 + }, + { + "epoch": 0.16243, + "grad_norm": 1.2250158805487876, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 16243 + }, + { + "epoch": 0.16244, + "grad_norm": 1.3050529176307324, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 16244 + }, + { + "epoch": 0.16245, + "grad_norm": 1.2853826796033716, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 16245 + }, + { + "epoch": 0.16246, + "grad_norm": 1.425835266955398, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 16246 + }, + { + "epoch": 0.16247, + "grad_norm": 1.009245169525709, + "learning_rate": 0.003, + "loss": 4.009, + "step": 16247 + }, + { + "epoch": 0.16248, + "grad_norm": 1.6861590499758379, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 16248 + }, + { + "epoch": 0.16249, + "grad_norm": 1.0409671667743918, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 16249 + }, + { + "epoch": 0.1625, + "grad_norm": 1.4350539872300103, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 16250 + }, + { + "epoch": 0.16251, + "grad_norm": 1.1367477765992338, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16251 + }, + { + "epoch": 0.16252, + "grad_norm": 1.2107808951722476, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 16252 + }, + { + "epoch": 0.16253, + "grad_norm": 1.1892489577790892, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 16253 + }, + { + "epoch": 0.16254, + "grad_norm": 1.2124577797786444, + "learning_rate": 0.003, + "loss": 3.99, + "step": 16254 + }, + { + "epoch": 0.16255, + "grad_norm": 1.208255102284434, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 16255 + }, + { + "epoch": 0.16256, + "grad_norm": 1.299753068132065, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16256 + }, + { + "epoch": 0.16257, + "grad_norm": 1.17324629592306, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 16257 + }, + { + "epoch": 0.16258, + "grad_norm": 1.6489190676746355, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 16258 + }, + { + "epoch": 0.16259, + "grad_norm": 0.9070137256189884, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 16259 + }, + { + "epoch": 0.1626, + "grad_norm": 1.323713007881002, + "learning_rate": 0.003, + "loss": 4.014, + "step": 16260 + }, + { + "epoch": 0.16261, + "grad_norm": 1.2576953485729692, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 16261 + }, + { + "epoch": 0.16262, + "grad_norm": 1.0306207731281276, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 16262 + }, + { + "epoch": 0.16263, + "grad_norm": 1.3418643829455337, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 16263 + }, + { + "epoch": 0.16264, + "grad_norm": 1.0795537290135617, + "learning_rate": 0.003, + "loss": 4.023, + "step": 16264 + }, + { + "epoch": 0.16265, + "grad_norm": 1.3765411902940559, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 16265 + }, + { + "epoch": 0.16266, + "grad_norm": 1.1264219486711267, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 16266 + }, + { + "epoch": 0.16267, + "grad_norm": 1.5943807470861864, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 16267 + }, + { + "epoch": 0.16268, + "grad_norm": 1.1331519759746556, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 16268 + }, + { + "epoch": 0.16269, + "grad_norm": 1.4076888603028517, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 16269 + }, + { + "epoch": 0.1627, + "grad_norm": 1.1338980685085958, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 16270 + }, + { + "epoch": 0.16271, + "grad_norm": 1.2263541104881175, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 16271 + }, + { + "epoch": 0.16272, + "grad_norm": 1.2477132012700038, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 16272 + }, + { + "epoch": 0.16273, + "grad_norm": 1.4033482768081984, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 16273 + }, + { + "epoch": 0.16274, + "grad_norm": 1.2407092930558066, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 16274 + }, + { + "epoch": 0.16275, + "grad_norm": 1.1690149723541716, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 16275 + }, + { + "epoch": 0.16276, + "grad_norm": 1.4404826682453318, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 16276 + }, + { + "epoch": 0.16277, + "grad_norm": 1.0793605987424126, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16277 + }, + { + "epoch": 0.16278, + "grad_norm": 1.2410935408130972, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 16278 + }, + { + "epoch": 0.16279, + "grad_norm": 1.1018110785157487, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 16279 + }, + { + "epoch": 0.1628, + "grad_norm": 1.4350516445394483, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 16280 + }, + { + "epoch": 0.16281, + "grad_norm": 1.040440143488395, + "learning_rate": 0.003, + "loss": 3.999, + "step": 16281 + }, + { + "epoch": 0.16282, + "grad_norm": 1.5725596187641915, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 16282 + }, + { + "epoch": 0.16283, + "grad_norm": 1.174488443518014, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 16283 + }, + { + "epoch": 0.16284, + "grad_norm": 1.1418980978133457, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 16284 + }, + { + "epoch": 0.16285, + "grad_norm": 1.3176145082440558, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 16285 + }, + { + "epoch": 0.16286, + "grad_norm": 1.2729396117612566, + "learning_rate": 0.003, + "loss": 4.031, + "step": 16286 + }, + { + "epoch": 0.16287, + "grad_norm": 1.0582851913805245, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 16287 + }, + { + "epoch": 0.16288, + "grad_norm": 1.448962771931629, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 16288 + }, + { + "epoch": 0.16289, + "grad_norm": 1.0817148956757496, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16289 + }, + { + "epoch": 0.1629, + "grad_norm": 1.2740098023804838, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16290 + }, + { + "epoch": 0.16291, + "grad_norm": 1.1614133630841075, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16291 + }, + { + "epoch": 0.16292, + "grad_norm": 1.236868613041996, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 16292 + }, + { + "epoch": 0.16293, + "grad_norm": 1.2869133677979343, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 16293 + }, + { + "epoch": 0.16294, + "grad_norm": 1.2907804379730377, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 16294 + }, + { + "epoch": 0.16295, + "grad_norm": 1.099536496169472, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 16295 + }, + { + "epoch": 0.16296, + "grad_norm": 1.3335936028645292, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 16296 + }, + { + "epoch": 0.16297, + "grad_norm": 1.1645557198097691, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16297 + }, + { + "epoch": 0.16298, + "grad_norm": 1.3446573090864091, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 16298 + }, + { + "epoch": 0.16299, + "grad_norm": 1.277875127298129, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 16299 + }, + { + "epoch": 0.163, + "grad_norm": 1.360414138348981, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 16300 + }, + { + "epoch": 0.16301, + "grad_norm": 1.056558589312762, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 16301 + }, + { + "epoch": 0.16302, + "grad_norm": 1.2740972430036706, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 16302 + }, + { + "epoch": 0.16303, + "grad_norm": 1.0433904292266207, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16303 + }, + { + "epoch": 0.16304, + "grad_norm": 1.4690931174665078, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 16304 + }, + { + "epoch": 0.16305, + "grad_norm": 1.3585764808502565, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 16305 + }, + { + "epoch": 0.16306, + "grad_norm": 0.9738797029139665, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 16306 + }, + { + "epoch": 0.16307, + "grad_norm": 1.1886139683680965, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 16307 + }, + { + "epoch": 0.16308, + "grad_norm": 1.4949177477980182, + "learning_rate": 0.003, + "loss": 4.001, + "step": 16308 + }, + { + "epoch": 0.16309, + "grad_norm": 1.2034762337178353, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 16309 + }, + { + "epoch": 0.1631, + "grad_norm": 1.2990119340504764, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 16310 + }, + { + "epoch": 0.16311, + "grad_norm": 1.2898335711830429, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16311 + }, + { + "epoch": 0.16312, + "grad_norm": 1.1909578646778416, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 16312 + }, + { + "epoch": 0.16313, + "grad_norm": 1.410969870509698, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 16313 + }, + { + "epoch": 0.16314, + "grad_norm": 1.1728459534717985, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 16314 + }, + { + "epoch": 0.16315, + "grad_norm": 1.4326695495326103, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 16315 + }, + { + "epoch": 0.16316, + "grad_norm": 1.200058845173753, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 16316 + }, + { + "epoch": 0.16317, + "grad_norm": 1.1029542964356904, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 16317 + }, + { + "epoch": 0.16318, + "grad_norm": 1.4904739266686984, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 16318 + }, + { + "epoch": 0.16319, + "grad_norm": 1.008823446422317, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 16319 + }, + { + "epoch": 0.1632, + "grad_norm": 1.5958122881545798, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 16320 + }, + { + "epoch": 0.16321, + "grad_norm": 1.0776684869138313, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 16321 + }, + { + "epoch": 0.16322, + "grad_norm": 1.3250262438870306, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 16322 + }, + { + "epoch": 0.16323, + "grad_norm": 1.2879418917363281, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 16323 + }, + { + "epoch": 0.16324, + "grad_norm": 1.277337742300279, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 16324 + }, + { + "epoch": 0.16325, + "grad_norm": 1.2046109700156538, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 16325 + }, + { + "epoch": 0.16326, + "grad_norm": 1.1594972035327338, + "learning_rate": 0.003, + "loss": 4.02, + "step": 16326 + }, + { + "epoch": 0.16327, + "grad_norm": 1.2993757871545841, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 16327 + }, + { + "epoch": 0.16328, + "grad_norm": 0.9459441462719139, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 16328 + }, + { + "epoch": 0.16329, + "grad_norm": 1.089640361789377, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 16329 + }, + { + "epoch": 0.1633, + "grad_norm": 1.357347324378313, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 16330 + }, + { + "epoch": 0.16331, + "grad_norm": 1.3832348141460884, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 16331 + }, + { + "epoch": 0.16332, + "grad_norm": 1.1657345267078214, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16332 + }, + { + "epoch": 0.16333, + "grad_norm": 1.2352142220832043, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16333 + }, + { + "epoch": 0.16334, + "grad_norm": 1.3011743066663761, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 16334 + }, + { + "epoch": 0.16335, + "grad_norm": 1.301849855697704, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 16335 + }, + { + "epoch": 0.16336, + "grad_norm": 1.197578674404236, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 16336 + }, + { + "epoch": 0.16337, + "grad_norm": 1.2289120276716436, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 16337 + }, + { + "epoch": 0.16338, + "grad_norm": 1.0665177380441002, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 16338 + }, + { + "epoch": 0.16339, + "grad_norm": 1.4593108939078878, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16339 + }, + { + "epoch": 0.1634, + "grad_norm": 0.9955358818303156, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 16340 + }, + { + "epoch": 0.16341, + "grad_norm": 1.1762072836442636, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 16341 + }, + { + "epoch": 0.16342, + "grad_norm": 1.2875094008417542, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 16342 + }, + { + "epoch": 0.16343, + "grad_norm": 1.1438620403444335, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 16343 + }, + { + "epoch": 0.16344, + "grad_norm": 1.2411591665010742, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 16344 + }, + { + "epoch": 0.16345, + "grad_norm": 1.344220362883522, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 16345 + }, + { + "epoch": 0.16346, + "grad_norm": 1.1197276916624015, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16346 + }, + { + "epoch": 0.16347, + "grad_norm": 1.2267385558994073, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 16347 + }, + { + "epoch": 0.16348, + "grad_norm": 1.4237698868352837, + "learning_rate": 0.003, + "loss": 4.036, + "step": 16348 + }, + { + "epoch": 0.16349, + "grad_norm": 1.2159721118278595, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 16349 + }, + { + "epoch": 0.1635, + "grad_norm": 1.2293083591735972, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 16350 + }, + { + "epoch": 0.16351, + "grad_norm": 1.2889322429236827, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16351 + }, + { + "epoch": 0.16352, + "grad_norm": 1.3685760771446185, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 16352 + }, + { + "epoch": 0.16353, + "grad_norm": 1.222184218179719, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 16353 + }, + { + "epoch": 0.16354, + "grad_norm": 1.4062042052693153, + "learning_rate": 0.003, + "loss": 4.045, + "step": 16354 + }, + { + "epoch": 0.16355, + "grad_norm": 1.1982614837664545, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 16355 + }, + { + "epoch": 0.16356, + "grad_norm": 1.0474864519503508, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 16356 + }, + { + "epoch": 0.16357, + "grad_norm": 1.210625814816236, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 16357 + }, + { + "epoch": 0.16358, + "grad_norm": 1.1218492901639943, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 16358 + }, + { + "epoch": 0.16359, + "grad_norm": 1.3571042584631752, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 16359 + }, + { + "epoch": 0.1636, + "grad_norm": 1.097890712606949, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 16360 + }, + { + "epoch": 0.16361, + "grad_norm": 1.322708037059437, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 16361 + }, + { + "epoch": 0.16362, + "grad_norm": 1.465739171631752, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 16362 + }, + { + "epoch": 0.16363, + "grad_norm": 1.2289048539225673, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 16363 + }, + { + "epoch": 0.16364, + "grad_norm": 1.3255134052006838, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 16364 + }, + { + "epoch": 0.16365, + "grad_norm": 1.631364182785758, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 16365 + }, + { + "epoch": 0.16366, + "grad_norm": 1.0307348934901521, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 16366 + }, + { + "epoch": 0.16367, + "grad_norm": 1.5072865545072853, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 16367 + }, + { + "epoch": 0.16368, + "grad_norm": 1.0483240140762637, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 16368 + }, + { + "epoch": 0.16369, + "grad_norm": 1.5864667762571492, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 16369 + }, + { + "epoch": 0.1637, + "grad_norm": 0.9633118780727623, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 16370 + }, + { + "epoch": 0.16371, + "grad_norm": 1.3838750786501524, + "learning_rate": 0.003, + "loss": 4.041, + "step": 16371 + }, + { + "epoch": 0.16372, + "grad_norm": 1.0879613892858553, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 16372 + }, + { + "epoch": 0.16373, + "grad_norm": 1.4076474578927347, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 16373 + }, + { + "epoch": 0.16374, + "grad_norm": 1.2549560835076428, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 16374 + }, + { + "epoch": 0.16375, + "grad_norm": 1.2218421563642234, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 16375 + }, + { + "epoch": 0.16376, + "grad_norm": 1.1068188648679878, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16376 + }, + { + "epoch": 0.16377, + "grad_norm": 1.356022831472547, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 16377 + }, + { + "epoch": 0.16378, + "grad_norm": 1.2627628500161356, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 16378 + }, + { + "epoch": 0.16379, + "grad_norm": 1.1543103781748223, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 16379 + }, + { + "epoch": 0.1638, + "grad_norm": 1.3650629405227102, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16380 + }, + { + "epoch": 0.16381, + "grad_norm": 1.0863105705930893, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 16381 + }, + { + "epoch": 0.16382, + "grad_norm": 1.442162970226098, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 16382 + }, + { + "epoch": 0.16383, + "grad_norm": 1.2105949431390675, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 16383 + }, + { + "epoch": 0.16384, + "grad_norm": 1.2445309684091068, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 16384 + }, + { + "epoch": 0.16385, + "grad_norm": 1.2627835888797698, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 16385 + }, + { + "epoch": 0.16386, + "grad_norm": 1.2739873244060556, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 16386 + }, + { + "epoch": 0.16387, + "grad_norm": 1.136346771273126, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 16387 + }, + { + "epoch": 0.16388, + "grad_norm": 1.2313390681202687, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 16388 + }, + { + "epoch": 0.16389, + "grad_norm": 1.3446402546969176, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 16389 + }, + { + "epoch": 0.1639, + "grad_norm": 1.1748177586116015, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 16390 + }, + { + "epoch": 0.16391, + "grad_norm": 1.4496101096216325, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 16391 + }, + { + "epoch": 0.16392, + "grad_norm": 1.1392287821141556, + "learning_rate": 0.003, + "loss": 3.989, + "step": 16392 + }, + { + "epoch": 0.16393, + "grad_norm": 1.218511368003899, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 16393 + }, + { + "epoch": 0.16394, + "grad_norm": 1.146180596348099, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 16394 + }, + { + "epoch": 0.16395, + "grad_norm": 1.207602007972604, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 16395 + }, + { + "epoch": 0.16396, + "grad_norm": 1.249250791280474, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 16396 + }, + { + "epoch": 0.16397, + "grad_norm": 1.2974792736158025, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 16397 + }, + { + "epoch": 0.16398, + "grad_norm": 1.385400203784108, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 16398 + }, + { + "epoch": 0.16399, + "grad_norm": 0.9583698373756244, + "learning_rate": 0.003, + "loss": 4.004, + "step": 16399 + }, + { + "epoch": 0.164, + "grad_norm": 1.401565056887489, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 16400 + }, + { + "epoch": 0.16401, + "grad_norm": 1.1642142279096959, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 16401 + }, + { + "epoch": 0.16402, + "grad_norm": 1.2183166639913017, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 16402 + }, + { + "epoch": 0.16403, + "grad_norm": 1.2507916104231493, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 16403 + }, + { + "epoch": 0.16404, + "grad_norm": 1.14410261257001, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 16404 + }, + { + "epoch": 0.16405, + "grad_norm": 1.3882639592151966, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16405 + }, + { + "epoch": 0.16406, + "grad_norm": 1.2800368116147272, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16406 + }, + { + "epoch": 0.16407, + "grad_norm": 1.249683573714354, + "learning_rate": 0.003, + "loss": 4.038, + "step": 16407 + }, + { + "epoch": 0.16408, + "grad_norm": 1.1167375177906984, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 16408 + }, + { + "epoch": 0.16409, + "grad_norm": 1.263879094909315, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 16409 + }, + { + "epoch": 0.1641, + "grad_norm": 1.1124381509251915, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16410 + }, + { + "epoch": 0.16411, + "grad_norm": 1.3936589885324306, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 16411 + }, + { + "epoch": 0.16412, + "grad_norm": 1.12451852583607, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 16412 + }, + { + "epoch": 0.16413, + "grad_norm": 1.460067412372667, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 16413 + }, + { + "epoch": 0.16414, + "grad_norm": 0.9747630232628195, + "learning_rate": 0.003, + "loss": 3.987, + "step": 16414 + }, + { + "epoch": 0.16415, + "grad_norm": 1.410573158295569, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16415 + }, + { + "epoch": 0.16416, + "grad_norm": 1.1376270299664892, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 16416 + }, + { + "epoch": 0.16417, + "grad_norm": 1.2495604579613826, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 16417 + }, + { + "epoch": 0.16418, + "grad_norm": 1.3823509772498197, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 16418 + }, + { + "epoch": 0.16419, + "grad_norm": 1.2138146849977327, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 16419 + }, + { + "epoch": 0.1642, + "grad_norm": 1.1861034753425292, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 16420 + }, + { + "epoch": 0.16421, + "grad_norm": 1.5003316384806735, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 16421 + }, + { + "epoch": 0.16422, + "grad_norm": 1.1063407146960826, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 16422 + }, + { + "epoch": 0.16423, + "grad_norm": 1.7022747407394883, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 16423 + }, + { + "epoch": 0.16424, + "grad_norm": 0.9854746270716073, + "learning_rate": 0.003, + "loss": 4.02, + "step": 16424 + }, + { + "epoch": 0.16425, + "grad_norm": 1.3461991942792693, + "learning_rate": 0.003, + "loss": 4.013, + "step": 16425 + }, + { + "epoch": 0.16426, + "grad_norm": 1.2611638827136702, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 16426 + }, + { + "epoch": 0.16427, + "grad_norm": 1.2392649592127025, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16427 + }, + { + "epoch": 0.16428, + "grad_norm": 1.1600141364720165, + "learning_rate": 0.003, + "loss": 4.022, + "step": 16428 + }, + { + "epoch": 0.16429, + "grad_norm": 1.365743175043391, + "learning_rate": 0.003, + "loss": 4.021, + "step": 16429 + }, + { + "epoch": 0.1643, + "grad_norm": 1.2341725722983907, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 16430 + }, + { + "epoch": 0.16431, + "grad_norm": 1.2664336922254367, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 16431 + }, + { + "epoch": 0.16432, + "grad_norm": 1.162134504903983, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 16432 + }, + { + "epoch": 0.16433, + "grad_norm": 1.1890031067087414, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 16433 + }, + { + "epoch": 0.16434, + "grad_norm": 1.268343217918924, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 16434 + }, + { + "epoch": 0.16435, + "grad_norm": 0.9735238227671814, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 16435 + }, + { + "epoch": 0.16436, + "grad_norm": 1.2501414862931537, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 16436 + }, + { + "epoch": 0.16437, + "grad_norm": 1.1033374229738238, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 16437 + }, + { + "epoch": 0.16438, + "grad_norm": 1.2530212734335278, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16438 + }, + { + "epoch": 0.16439, + "grad_norm": 1.1415596092894056, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16439 + }, + { + "epoch": 0.1644, + "grad_norm": 1.258427707531266, + "learning_rate": 0.003, + "loss": 4.002, + "step": 16440 + }, + { + "epoch": 0.16441, + "grad_norm": 1.3504692944034622, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 16441 + }, + { + "epoch": 0.16442, + "grad_norm": 1.0471200830711485, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 16442 + }, + { + "epoch": 0.16443, + "grad_norm": 1.3586376355519827, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 16443 + }, + { + "epoch": 0.16444, + "grad_norm": 1.2291363527030847, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 16444 + }, + { + "epoch": 0.16445, + "grad_norm": 1.2462129480970447, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 16445 + }, + { + "epoch": 0.16446, + "grad_norm": 1.081352439140012, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 16446 + }, + { + "epoch": 0.16447, + "grad_norm": 1.671540047008696, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 16447 + }, + { + "epoch": 0.16448, + "grad_norm": 1.3643640246569684, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 16448 + }, + { + "epoch": 0.16449, + "grad_norm": 1.3453266447923604, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 16449 + }, + { + "epoch": 0.1645, + "grad_norm": 1.20426945020544, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 16450 + }, + { + "epoch": 0.16451, + "grad_norm": 1.2960461000701875, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 16451 + }, + { + "epoch": 0.16452, + "grad_norm": 1.1882929055492957, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 16452 + }, + { + "epoch": 0.16453, + "grad_norm": 1.1786688126968292, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 16453 + }, + { + "epoch": 0.16454, + "grad_norm": 1.3048838114337977, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 16454 + }, + { + "epoch": 0.16455, + "grad_norm": 1.3505352827862118, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 16455 + }, + { + "epoch": 0.16456, + "grad_norm": 1.2457370763871507, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16456 + }, + { + "epoch": 0.16457, + "grad_norm": 1.1464300008417747, + "learning_rate": 0.003, + "loss": 4.034, + "step": 16457 + }, + { + "epoch": 0.16458, + "grad_norm": 1.268097477324259, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 16458 + }, + { + "epoch": 0.16459, + "grad_norm": 1.4368587052408366, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16459 + }, + { + "epoch": 0.1646, + "grad_norm": 1.2233996992334433, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 16460 + }, + { + "epoch": 0.16461, + "grad_norm": 1.1727679642417794, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 16461 + }, + { + "epoch": 0.16462, + "grad_norm": 1.3165209833868021, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 16462 + }, + { + "epoch": 0.16463, + "grad_norm": 1.2273269610329756, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 16463 + }, + { + "epoch": 0.16464, + "grad_norm": 1.2464394092353464, + "learning_rate": 0.003, + "loss": 3.993, + "step": 16464 + }, + { + "epoch": 0.16465, + "grad_norm": 1.3187530477988323, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 16465 + }, + { + "epoch": 0.16466, + "grad_norm": 1.1829192865232832, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 16466 + }, + { + "epoch": 0.16467, + "grad_norm": 1.1799295434341366, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16467 + }, + { + "epoch": 0.16468, + "grad_norm": 1.1506766706819855, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 16468 + }, + { + "epoch": 0.16469, + "grad_norm": 1.415062921477318, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 16469 + }, + { + "epoch": 0.1647, + "grad_norm": 1.472580170523846, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 16470 + }, + { + "epoch": 0.16471, + "grad_norm": 1.2214043721676051, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16471 + }, + { + "epoch": 0.16472, + "grad_norm": 1.2147205884895964, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 16472 + }, + { + "epoch": 0.16473, + "grad_norm": 1.1776590050316018, + "learning_rate": 0.003, + "loss": 4.023, + "step": 16473 + }, + { + "epoch": 0.16474, + "grad_norm": 1.2389656898506585, + "learning_rate": 0.003, + "loss": 4.015, + "step": 16474 + }, + { + "epoch": 0.16475, + "grad_norm": 1.176666306433308, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 16475 + }, + { + "epoch": 0.16476, + "grad_norm": 1.3147913418496244, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 16476 + }, + { + "epoch": 0.16477, + "grad_norm": 1.1587640003981066, + "learning_rate": 0.003, + "loss": 4.014, + "step": 16477 + }, + { + "epoch": 0.16478, + "grad_norm": 1.273223333688864, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 16478 + }, + { + "epoch": 0.16479, + "grad_norm": 1.2291506382173074, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16479 + }, + { + "epoch": 0.1648, + "grad_norm": 1.2322478623329385, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 16480 + }, + { + "epoch": 0.16481, + "grad_norm": 1.2174792441948241, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16481 + }, + { + "epoch": 0.16482, + "grad_norm": 1.4378399317818813, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 16482 + }, + { + "epoch": 0.16483, + "grad_norm": 1.1371049837514948, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 16483 + }, + { + "epoch": 0.16484, + "grad_norm": 1.3173930471537727, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 16484 + }, + { + "epoch": 0.16485, + "grad_norm": 1.1046764475930426, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 16485 + }, + { + "epoch": 0.16486, + "grad_norm": 1.239045577924919, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 16486 + }, + { + "epoch": 0.16487, + "grad_norm": 1.4686884132531135, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 16487 + }, + { + "epoch": 0.16488, + "grad_norm": 1.0800050156006988, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 16488 + }, + { + "epoch": 0.16489, + "grad_norm": 1.4134323871973382, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 16489 + }, + { + "epoch": 0.1649, + "grad_norm": 0.9682334441192832, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 16490 + }, + { + "epoch": 0.16491, + "grad_norm": 1.4200881733091222, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 16491 + }, + { + "epoch": 0.16492, + "grad_norm": 1.061585186787891, + "learning_rate": 0.003, + "loss": 3.995, + "step": 16492 + }, + { + "epoch": 0.16493, + "grad_norm": 1.7032178016253496, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 16493 + }, + { + "epoch": 0.16494, + "grad_norm": 1.1359795011220917, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16494 + }, + { + "epoch": 0.16495, + "grad_norm": 1.2483787020055124, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 16495 + }, + { + "epoch": 0.16496, + "grad_norm": 1.2910680574674094, + "learning_rate": 0.003, + "loss": 4.007, + "step": 16496 + }, + { + "epoch": 0.16497, + "grad_norm": 1.1862278580892176, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 16497 + }, + { + "epoch": 0.16498, + "grad_norm": 1.4112033344203716, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 16498 + }, + { + "epoch": 0.16499, + "grad_norm": 1.1855255839885497, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 16499 + }, + { + "epoch": 0.165, + "grad_norm": 1.4129010225149548, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 16500 + }, + { + "epoch": 0.16501, + "grad_norm": 0.9507245813429811, + "learning_rate": 0.003, + "loss": 4.026, + "step": 16501 + }, + { + "epoch": 0.16502, + "grad_norm": 1.3347143352886812, + "learning_rate": 0.003, + "loss": 4.023, + "step": 16502 + }, + { + "epoch": 0.16503, + "grad_norm": 1.0347509498235556, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 16503 + }, + { + "epoch": 0.16504, + "grad_norm": 1.423740368524998, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 16504 + }, + { + "epoch": 0.16505, + "grad_norm": 1.122722153433454, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 16505 + }, + { + "epoch": 0.16506, + "grad_norm": 1.294244218597045, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 16506 + }, + { + "epoch": 0.16507, + "grad_norm": 1.1443270649860016, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 16507 + }, + { + "epoch": 0.16508, + "grad_norm": 1.442432469846176, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 16508 + }, + { + "epoch": 0.16509, + "grad_norm": 1.0582122500043307, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 16509 + }, + { + "epoch": 0.1651, + "grad_norm": 1.6553330678007394, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 16510 + }, + { + "epoch": 0.16511, + "grad_norm": 1.2460014109083422, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 16511 + }, + { + "epoch": 0.16512, + "grad_norm": 1.4193887292079423, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 16512 + }, + { + "epoch": 0.16513, + "grad_norm": 0.9212556927907253, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 16513 + }, + { + "epoch": 0.16514, + "grad_norm": 1.2699018818052847, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 16514 + }, + { + "epoch": 0.16515, + "grad_norm": 1.106930909545978, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 16515 + }, + { + "epoch": 0.16516, + "grad_norm": 1.2259637741423628, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 16516 + }, + { + "epoch": 0.16517, + "grad_norm": 1.15824728988598, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 16517 + }, + { + "epoch": 0.16518, + "grad_norm": 1.3487176359210198, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 16518 + }, + { + "epoch": 0.16519, + "grad_norm": 1.0376620933785712, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 16519 + }, + { + "epoch": 0.1652, + "grad_norm": 1.3406151681112863, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 16520 + }, + { + "epoch": 0.16521, + "grad_norm": 1.237189948735632, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 16521 + }, + { + "epoch": 0.16522, + "grad_norm": 1.5870827299658874, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 16522 + }, + { + "epoch": 0.16523, + "grad_norm": 0.996710035102378, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 16523 + }, + { + "epoch": 0.16524, + "grad_norm": 1.3138563369561216, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 16524 + }, + { + "epoch": 0.16525, + "grad_norm": 0.9443266091429152, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 16525 + }, + { + "epoch": 0.16526, + "grad_norm": 1.3288852608092152, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 16526 + }, + { + "epoch": 0.16527, + "grad_norm": 1.1161681754946187, + "learning_rate": 0.003, + "loss": 4.035, + "step": 16527 + }, + { + "epoch": 0.16528, + "grad_norm": 1.2676851233683877, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 16528 + }, + { + "epoch": 0.16529, + "grad_norm": 1.5278489919144347, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 16529 + }, + { + "epoch": 0.1653, + "grad_norm": 1.3333473641523625, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 16530 + }, + { + "epoch": 0.16531, + "grad_norm": 1.0450767874846136, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 16531 + }, + { + "epoch": 0.16532, + "grad_norm": 1.276311673917285, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 16532 + }, + { + "epoch": 0.16533, + "grad_norm": 1.311758439221919, + "learning_rate": 0.003, + "loss": 4.017, + "step": 16533 + }, + { + "epoch": 0.16534, + "grad_norm": 1.251935222227861, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 16534 + }, + { + "epoch": 0.16535, + "grad_norm": 1.4219610470844877, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 16535 + }, + { + "epoch": 0.16536, + "grad_norm": 1.098314666492021, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 16536 + }, + { + "epoch": 0.16537, + "grad_norm": 1.2524466159575347, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 16537 + }, + { + "epoch": 0.16538, + "grad_norm": 1.1051257760069162, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 16538 + }, + { + "epoch": 0.16539, + "grad_norm": 1.4507028107549453, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 16539 + }, + { + "epoch": 0.1654, + "grad_norm": 1.2015000906469193, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16540 + }, + { + "epoch": 0.16541, + "grad_norm": 1.5193683138530925, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16541 + }, + { + "epoch": 0.16542, + "grad_norm": 0.9145876723834636, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 16542 + }, + { + "epoch": 0.16543, + "grad_norm": 1.1502530819129524, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 16543 + }, + { + "epoch": 0.16544, + "grad_norm": 1.3293026862371053, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 16544 + }, + { + "epoch": 0.16545, + "grad_norm": 1.0563595407032713, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 16545 + }, + { + "epoch": 0.16546, + "grad_norm": 1.4948166835840087, + "learning_rate": 0.003, + "loss": 4.011, + "step": 16546 + }, + { + "epoch": 0.16547, + "grad_norm": 0.9426733011174305, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 16547 + }, + { + "epoch": 0.16548, + "grad_norm": 1.3107429457937985, + "learning_rate": 0.003, + "loss": 4.003, + "step": 16548 + }, + { + "epoch": 0.16549, + "grad_norm": 1.3094077637021049, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 16549 + }, + { + "epoch": 0.1655, + "grad_norm": 1.073261489594188, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16550 + }, + { + "epoch": 0.16551, + "grad_norm": 1.354807864246254, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16551 + }, + { + "epoch": 0.16552, + "grad_norm": 1.0112854922645667, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 16552 + }, + { + "epoch": 0.16553, + "grad_norm": 1.59252432958257, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 16553 + }, + { + "epoch": 0.16554, + "grad_norm": 1.2108947227394369, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 16554 + }, + { + "epoch": 0.16555, + "grad_norm": 1.1690231164195901, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 16555 + }, + { + "epoch": 0.16556, + "grad_norm": 1.4264690518283973, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 16556 + }, + { + "epoch": 0.16557, + "grad_norm": 1.124470612180664, + "learning_rate": 0.003, + "loss": 4.02, + "step": 16557 + }, + { + "epoch": 0.16558, + "grad_norm": 1.4096201695258213, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 16558 + }, + { + "epoch": 0.16559, + "grad_norm": 1.1374025708558277, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 16559 + }, + { + "epoch": 0.1656, + "grad_norm": 1.4429274806112555, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 16560 + }, + { + "epoch": 0.16561, + "grad_norm": 1.3442276249320984, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 16561 + }, + { + "epoch": 0.16562, + "grad_norm": 1.480263293504643, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 16562 + }, + { + "epoch": 0.16563, + "grad_norm": 1.1138476514363325, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 16563 + }, + { + "epoch": 0.16564, + "grad_norm": 1.3431345573500961, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 16564 + }, + { + "epoch": 0.16565, + "grad_norm": 1.4352240332625645, + "learning_rate": 0.003, + "loss": 4.027, + "step": 16565 + }, + { + "epoch": 0.16566, + "grad_norm": 0.9837148136132222, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 16566 + }, + { + "epoch": 0.16567, + "grad_norm": 1.299108251594754, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 16567 + }, + { + "epoch": 0.16568, + "grad_norm": 1.1682095581834722, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 16568 + }, + { + "epoch": 0.16569, + "grad_norm": 1.1768012839694058, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16569 + }, + { + "epoch": 0.1657, + "grad_norm": 1.1031963756185752, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 16570 + }, + { + "epoch": 0.16571, + "grad_norm": 1.3860222274258607, + "learning_rate": 0.003, + "loss": 3.997, + "step": 16571 + }, + { + "epoch": 0.16572, + "grad_norm": 1.3969337609331909, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 16572 + }, + { + "epoch": 0.16573, + "grad_norm": 1.3063052614852653, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 16573 + }, + { + "epoch": 0.16574, + "grad_norm": 1.0464741451960757, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 16574 + }, + { + "epoch": 0.16575, + "grad_norm": 1.337615973734632, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 16575 + }, + { + "epoch": 0.16576, + "grad_norm": 1.1618907603870425, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 16576 + }, + { + "epoch": 0.16577, + "grad_norm": 1.3200946599609433, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 16577 + }, + { + "epoch": 0.16578, + "grad_norm": 1.0433429753006371, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 16578 + }, + { + "epoch": 0.16579, + "grad_norm": 1.662063329619664, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 16579 + }, + { + "epoch": 0.1658, + "grad_norm": 0.755601682306504, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 16580 + }, + { + "epoch": 0.16581, + "grad_norm": 1.0226406775980381, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 16581 + }, + { + "epoch": 0.16582, + "grad_norm": 1.3870489878903887, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16582 + }, + { + "epoch": 0.16583, + "grad_norm": 1.2087741859482437, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 16583 + }, + { + "epoch": 0.16584, + "grad_norm": 1.2152029624363838, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16584 + }, + { + "epoch": 0.16585, + "grad_norm": 1.3252901206130747, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 16585 + }, + { + "epoch": 0.16586, + "grad_norm": 1.1542687790472583, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 16586 + }, + { + "epoch": 0.16587, + "grad_norm": 1.3388694163458916, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 16587 + }, + { + "epoch": 0.16588, + "grad_norm": 1.3452678444742063, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 16588 + }, + { + "epoch": 0.16589, + "grad_norm": 1.1417025425422698, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 16589 + }, + { + "epoch": 0.1659, + "grad_norm": 1.1863637880560758, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 16590 + }, + { + "epoch": 0.16591, + "grad_norm": 1.4649363325974047, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 16591 + }, + { + "epoch": 0.16592, + "grad_norm": 1.1515827924904152, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 16592 + }, + { + "epoch": 0.16593, + "grad_norm": 1.2388456396257521, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 16593 + }, + { + "epoch": 0.16594, + "grad_norm": 1.232311058547632, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 16594 + }, + { + "epoch": 0.16595, + "grad_norm": 1.0999208653989847, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 16595 + }, + { + "epoch": 0.16596, + "grad_norm": 1.255290749615242, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16596 + }, + { + "epoch": 0.16597, + "grad_norm": 1.2889959301516716, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16597 + }, + { + "epoch": 0.16598, + "grad_norm": 1.3703577736789996, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 16598 + }, + { + "epoch": 0.16599, + "grad_norm": 1.2711906709860699, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 16599 + }, + { + "epoch": 0.166, + "grad_norm": 1.2141273117003566, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 16600 + }, + { + "epoch": 0.16601, + "grad_norm": 1.256337088331127, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 16601 + }, + { + "epoch": 0.16602, + "grad_norm": 1.2603912995095456, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 16602 + }, + { + "epoch": 0.16603, + "grad_norm": 1.2456616186407852, + "learning_rate": 0.003, + "loss": 4.039, + "step": 16603 + }, + { + "epoch": 0.16604, + "grad_norm": 1.415050378787449, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 16604 + }, + { + "epoch": 0.16605, + "grad_norm": 1.2547809965459051, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16605 + }, + { + "epoch": 0.16606, + "grad_norm": 1.3211577802082288, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 16606 + }, + { + "epoch": 0.16607, + "grad_norm": 1.3281001478920962, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16607 + }, + { + "epoch": 0.16608, + "grad_norm": 1.0475477901467427, + "learning_rate": 0.003, + "loss": 4.006, + "step": 16608 + }, + { + "epoch": 0.16609, + "grad_norm": 1.6382789949343959, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 16609 + }, + { + "epoch": 0.1661, + "grad_norm": 0.994687220833452, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 16610 + }, + { + "epoch": 0.16611, + "grad_norm": 1.5715788227614695, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 16611 + }, + { + "epoch": 0.16612, + "grad_norm": 1.0605613837990646, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 16612 + }, + { + "epoch": 0.16613, + "grad_norm": 1.2526526695670748, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 16613 + }, + { + "epoch": 0.16614, + "grad_norm": 1.4338328500378312, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 16614 + }, + { + "epoch": 0.16615, + "grad_norm": 1.0087894789702216, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 16615 + }, + { + "epoch": 0.16616, + "grad_norm": 1.3697634543246178, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16616 + }, + { + "epoch": 0.16617, + "grad_norm": 1.017535543821651, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 16617 + }, + { + "epoch": 0.16618, + "grad_norm": 1.3581523497537333, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 16618 + }, + { + "epoch": 0.16619, + "grad_norm": 1.1592790899911563, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 16619 + }, + { + "epoch": 0.1662, + "grad_norm": 1.3154600386381228, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 16620 + }, + { + "epoch": 0.16621, + "grad_norm": 1.4131419506857579, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 16621 + }, + { + "epoch": 0.16622, + "grad_norm": 1.120139961517329, + "learning_rate": 0.003, + "loss": 3.978, + "step": 16622 + }, + { + "epoch": 0.16623, + "grad_norm": 1.299129512003256, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 16623 + }, + { + "epoch": 0.16624, + "grad_norm": 1.0853732855861533, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 16624 + }, + { + "epoch": 0.16625, + "grad_norm": 1.2353134895344438, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 16625 + }, + { + "epoch": 0.16626, + "grad_norm": 1.0471620820993004, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 16626 + }, + { + "epoch": 0.16627, + "grad_norm": 1.5102483519391996, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 16627 + }, + { + "epoch": 0.16628, + "grad_norm": 1.3309274545091607, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 16628 + }, + { + "epoch": 0.16629, + "grad_norm": 1.3090873273971353, + "learning_rate": 0.003, + "loss": 4.055, + "step": 16629 + }, + { + "epoch": 0.1663, + "grad_norm": 1.3004116327526425, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 16630 + }, + { + "epoch": 0.16631, + "grad_norm": 1.0823361796499542, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 16631 + }, + { + "epoch": 0.16632, + "grad_norm": 1.2930715193846416, + "learning_rate": 0.003, + "loss": 4.009, + "step": 16632 + }, + { + "epoch": 0.16633, + "grad_norm": 1.424491690435542, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 16633 + }, + { + "epoch": 0.16634, + "grad_norm": 1.3834142236076787, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 16634 + }, + { + "epoch": 0.16635, + "grad_norm": 1.1052937514110142, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 16635 + }, + { + "epoch": 0.16636, + "grad_norm": 1.2612133046875074, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 16636 + }, + { + "epoch": 0.16637, + "grad_norm": 1.0980924188880004, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 16637 + }, + { + "epoch": 0.16638, + "grad_norm": 1.2526679697383083, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16638 + }, + { + "epoch": 0.16639, + "grad_norm": 1.452666129554067, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 16639 + }, + { + "epoch": 0.1664, + "grad_norm": 1.2623830226090003, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 16640 + }, + { + "epoch": 0.16641, + "grad_norm": 1.279212006893314, + "learning_rate": 0.003, + "loss": 4.039, + "step": 16641 + }, + { + "epoch": 0.16642, + "grad_norm": 1.247644798505221, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 16642 + }, + { + "epoch": 0.16643, + "grad_norm": 1.3550614752198638, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 16643 + }, + { + "epoch": 0.16644, + "grad_norm": 1.1944425316800222, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 16644 + }, + { + "epoch": 0.16645, + "grad_norm": 1.3047621724565892, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 16645 + }, + { + "epoch": 0.16646, + "grad_norm": 1.1236723102236015, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 16646 + }, + { + "epoch": 0.16647, + "grad_norm": 1.3102114573804153, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 16647 + }, + { + "epoch": 0.16648, + "grad_norm": 1.3853012906961333, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 16648 + }, + { + "epoch": 0.16649, + "grad_norm": 1.204186130236693, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 16649 + }, + { + "epoch": 0.1665, + "grad_norm": 1.2667647582380555, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 16650 + }, + { + "epoch": 0.16651, + "grad_norm": 1.1257390464916401, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 16651 + }, + { + "epoch": 0.16652, + "grad_norm": 1.186721837147446, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 16652 + }, + { + "epoch": 0.16653, + "grad_norm": 1.6462828654146497, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 16653 + }, + { + "epoch": 0.16654, + "grad_norm": 1.0411480294282478, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16654 + }, + { + "epoch": 0.16655, + "grad_norm": 1.4339898853354354, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 16655 + }, + { + "epoch": 0.16656, + "grad_norm": 1.1739844303369085, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 16656 + }, + { + "epoch": 0.16657, + "grad_norm": 1.317060669414251, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 16657 + }, + { + "epoch": 0.16658, + "grad_norm": 1.126918738365564, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 16658 + }, + { + "epoch": 0.16659, + "grad_norm": 1.3895895177863988, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 16659 + }, + { + "epoch": 0.1666, + "grad_norm": 1.096387786516118, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 16660 + }, + { + "epoch": 0.16661, + "grad_norm": 1.1898746539870575, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 16661 + }, + { + "epoch": 0.16662, + "grad_norm": 1.2605716405733234, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 16662 + }, + { + "epoch": 0.16663, + "grad_norm": 1.462237163983018, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 16663 + }, + { + "epoch": 0.16664, + "grad_norm": 1.3892155212314823, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 16664 + }, + { + "epoch": 0.16665, + "grad_norm": 1.340379892839075, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 16665 + }, + { + "epoch": 0.16666, + "grad_norm": 1.1584760250723478, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 16666 + }, + { + "epoch": 0.16667, + "grad_norm": 1.1991756211519113, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 16667 + }, + { + "epoch": 0.16668, + "grad_norm": 1.3689035096811366, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 16668 + }, + { + "epoch": 0.16669, + "grad_norm": 1.1232920358044445, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 16669 + }, + { + "epoch": 0.1667, + "grad_norm": 1.281615109502724, + "learning_rate": 0.003, + "loss": 4.004, + "step": 16670 + }, + { + "epoch": 0.16671, + "grad_norm": 1.1907936743828131, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 16671 + }, + { + "epoch": 0.16672, + "grad_norm": 1.4193075434273048, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 16672 + }, + { + "epoch": 0.16673, + "grad_norm": 1.0848940734064485, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 16673 + }, + { + "epoch": 0.16674, + "grad_norm": 1.5030624748585042, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 16674 + }, + { + "epoch": 0.16675, + "grad_norm": 1.0582258078957265, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 16675 + }, + { + "epoch": 0.16676, + "grad_norm": 1.4933079963776932, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 16676 + }, + { + "epoch": 0.16677, + "grad_norm": 1.0339320217688897, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 16677 + }, + { + "epoch": 0.16678, + "grad_norm": 1.2165205219680335, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 16678 + }, + { + "epoch": 0.16679, + "grad_norm": 1.2456381676389232, + "learning_rate": 0.003, + "loss": 4.026, + "step": 16679 + }, + { + "epoch": 0.1668, + "grad_norm": 1.1630329563160737, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 16680 + }, + { + "epoch": 0.16681, + "grad_norm": 1.3020999973173273, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 16681 + }, + { + "epoch": 0.16682, + "grad_norm": 1.0243414220786775, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 16682 + }, + { + "epoch": 0.16683, + "grad_norm": 1.302135134365088, + "learning_rate": 0.003, + "loss": 4.011, + "step": 16683 + }, + { + "epoch": 0.16684, + "grad_norm": 1.1169560404571857, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 16684 + }, + { + "epoch": 0.16685, + "grad_norm": 1.3256062969865654, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 16685 + }, + { + "epoch": 0.16686, + "grad_norm": 1.4763477134276877, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 16686 + }, + { + "epoch": 0.16687, + "grad_norm": 1.1346500313299308, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 16687 + }, + { + "epoch": 0.16688, + "grad_norm": 1.3655555799191739, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 16688 + }, + { + "epoch": 0.16689, + "grad_norm": 1.101426064051465, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 16689 + }, + { + "epoch": 0.1669, + "grad_norm": 1.3874465778661247, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 16690 + }, + { + "epoch": 0.16691, + "grad_norm": 1.0666013667829752, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 16691 + }, + { + "epoch": 0.16692, + "grad_norm": 1.294662136639945, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 16692 + }, + { + "epoch": 0.16693, + "grad_norm": 1.1594748536272677, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 16693 + }, + { + "epoch": 0.16694, + "grad_norm": 1.474360847049748, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 16694 + }, + { + "epoch": 0.16695, + "grad_norm": 1.11696559619686, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 16695 + }, + { + "epoch": 0.16696, + "grad_norm": 1.2671759382966818, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 16696 + }, + { + "epoch": 0.16697, + "grad_norm": 1.1882774659173905, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 16697 + }, + { + "epoch": 0.16698, + "grad_norm": 1.326883885143455, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 16698 + }, + { + "epoch": 0.16699, + "grad_norm": 1.0980149148747356, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 16699 + }, + { + "epoch": 0.167, + "grad_norm": 1.4241149247181872, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 16700 + }, + { + "epoch": 0.16701, + "grad_norm": 1.0844540501513107, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16701 + }, + { + "epoch": 0.16702, + "grad_norm": 1.5909286938176141, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 16702 + }, + { + "epoch": 0.16703, + "grad_norm": 1.1187009224227151, + "learning_rate": 0.003, + "loss": 3.98, + "step": 16703 + }, + { + "epoch": 0.16704, + "grad_norm": 1.3361911693628177, + "learning_rate": 0.003, + "loss": 4.046, + "step": 16704 + }, + { + "epoch": 0.16705, + "grad_norm": 1.1892842822492835, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 16705 + }, + { + "epoch": 0.16706, + "grad_norm": 1.3129804202129807, + "learning_rate": 0.003, + "loss": 4.005, + "step": 16706 + }, + { + "epoch": 0.16707, + "grad_norm": 1.2040177966034948, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 16707 + }, + { + "epoch": 0.16708, + "grad_norm": 1.3184689484922762, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 16708 + }, + { + "epoch": 0.16709, + "grad_norm": 1.1710476876758964, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 16709 + }, + { + "epoch": 0.1671, + "grad_norm": 1.482230330878662, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 16710 + }, + { + "epoch": 0.16711, + "grad_norm": 1.2272387432789482, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 16711 + }, + { + "epoch": 0.16712, + "grad_norm": 1.0283774893512463, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 16712 + }, + { + "epoch": 0.16713, + "grad_norm": 1.2397209300256942, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 16713 + }, + { + "epoch": 0.16714, + "grad_norm": 1.1778879363882633, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 16714 + }, + { + "epoch": 0.16715, + "grad_norm": 1.4236584642291445, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 16715 + }, + { + "epoch": 0.16716, + "grad_norm": 1.2435713003061704, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 16716 + }, + { + "epoch": 0.16717, + "grad_norm": 1.4285517568570933, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 16717 + }, + { + "epoch": 0.16718, + "grad_norm": 1.165974077606312, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 16718 + }, + { + "epoch": 0.16719, + "grad_norm": 1.2330315988108786, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 16719 + }, + { + "epoch": 0.1672, + "grad_norm": 1.1376708053144504, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 16720 + }, + { + "epoch": 0.16721, + "grad_norm": 1.4924178352557733, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 16721 + }, + { + "epoch": 0.16722, + "grad_norm": 1.1736703334094785, + "learning_rate": 0.003, + "loss": 3.993, + "step": 16722 + }, + { + "epoch": 0.16723, + "grad_norm": 1.4078887935894222, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 16723 + }, + { + "epoch": 0.16724, + "grad_norm": 1.0143058288575488, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 16724 + }, + { + "epoch": 0.16725, + "grad_norm": 1.5416517889434647, + "learning_rate": 0.003, + "loss": 4.049, + "step": 16725 + }, + { + "epoch": 0.16726, + "grad_norm": 1.277683984553053, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 16726 + }, + { + "epoch": 0.16727, + "grad_norm": 1.2943184716801348, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 16727 + }, + { + "epoch": 0.16728, + "grad_norm": 1.295832553200161, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 16728 + }, + { + "epoch": 0.16729, + "grad_norm": 1.0854698151175544, + "learning_rate": 0.003, + "loss": 4.012, + "step": 16729 + }, + { + "epoch": 0.1673, + "grad_norm": 1.2479114469022485, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 16730 + }, + { + "epoch": 0.16731, + "grad_norm": 1.2512266369895562, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 16731 + }, + { + "epoch": 0.16732, + "grad_norm": 1.3643541839221898, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 16732 + }, + { + "epoch": 0.16733, + "grad_norm": 1.2287892480363976, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 16733 + }, + { + "epoch": 0.16734, + "grad_norm": 1.225151030439417, + "learning_rate": 0.003, + "loss": 4.015, + "step": 16734 + }, + { + "epoch": 0.16735, + "grad_norm": 1.385564809223566, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 16735 + }, + { + "epoch": 0.16736, + "grad_norm": 1.195776919613441, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 16736 + }, + { + "epoch": 0.16737, + "grad_norm": 1.3884190285371376, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16737 + }, + { + "epoch": 0.16738, + "grad_norm": 1.3054771728390893, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 16738 + }, + { + "epoch": 0.16739, + "grad_norm": 1.46399542404328, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 16739 + }, + { + "epoch": 0.1674, + "grad_norm": 1.0718812275028755, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 16740 + }, + { + "epoch": 0.16741, + "grad_norm": 1.1745625466912526, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 16741 + }, + { + "epoch": 0.16742, + "grad_norm": 1.2075247790914365, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 16742 + }, + { + "epoch": 0.16743, + "grad_norm": 1.4867521718033896, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 16743 + }, + { + "epoch": 0.16744, + "grad_norm": 1.065952806673697, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 16744 + }, + { + "epoch": 0.16745, + "grad_norm": 1.3972026360932996, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 16745 + }, + { + "epoch": 0.16746, + "grad_norm": 0.9031064911642256, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 16746 + }, + { + "epoch": 0.16747, + "grad_norm": 1.2482872551552535, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 16747 + }, + { + "epoch": 0.16748, + "grad_norm": 1.0793757898817635, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 16748 + }, + { + "epoch": 0.16749, + "grad_norm": 1.277504587284281, + "learning_rate": 0.003, + "loss": 4.013, + "step": 16749 + }, + { + "epoch": 0.1675, + "grad_norm": 1.2997001547277538, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 16750 + }, + { + "epoch": 0.16751, + "grad_norm": 1.2662310141494997, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 16751 + }, + { + "epoch": 0.16752, + "grad_norm": 1.2715291357652396, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 16752 + }, + { + "epoch": 0.16753, + "grad_norm": 1.0903498454308935, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 16753 + }, + { + "epoch": 0.16754, + "grad_norm": 1.262414695269733, + "learning_rate": 0.003, + "loss": 4.022, + "step": 16754 + }, + { + "epoch": 0.16755, + "grad_norm": 1.2417549278845725, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 16755 + }, + { + "epoch": 0.16756, + "grad_norm": 1.2435194125675262, + "learning_rate": 0.003, + "loss": 3.976, + "step": 16756 + }, + { + "epoch": 0.16757, + "grad_norm": 1.146735908434143, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 16757 + }, + { + "epoch": 0.16758, + "grad_norm": 1.2602637595972412, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 16758 + }, + { + "epoch": 0.16759, + "grad_norm": 1.3309230491987023, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 16759 + }, + { + "epoch": 0.1676, + "grad_norm": 0.989748089374222, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16760 + }, + { + "epoch": 0.16761, + "grad_norm": 1.313745719423959, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 16761 + }, + { + "epoch": 0.16762, + "grad_norm": 1.1794458031105426, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 16762 + }, + { + "epoch": 0.16763, + "grad_norm": 1.5006657592196686, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 16763 + }, + { + "epoch": 0.16764, + "grad_norm": 1.0854875211858523, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 16764 + }, + { + "epoch": 0.16765, + "grad_norm": 1.2799489508034012, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 16765 + }, + { + "epoch": 0.16766, + "grad_norm": 1.5385809064040605, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 16766 + }, + { + "epoch": 0.16767, + "grad_norm": 0.9871757576558676, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 16767 + }, + { + "epoch": 0.16768, + "grad_norm": 1.517071967155638, + "learning_rate": 0.003, + "loss": 4.04, + "step": 16768 + }, + { + "epoch": 0.16769, + "grad_norm": 1.1359527581158686, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 16769 + }, + { + "epoch": 0.1677, + "grad_norm": 1.5020135445753076, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 16770 + }, + { + "epoch": 0.16771, + "grad_norm": 0.9101721493263855, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 16771 + }, + { + "epoch": 0.16772, + "grad_norm": 1.0621089618420638, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 16772 + }, + { + "epoch": 0.16773, + "grad_norm": 1.5564499611496503, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 16773 + }, + { + "epoch": 0.16774, + "grad_norm": 1.155688872689453, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 16774 + }, + { + "epoch": 0.16775, + "grad_norm": 1.3845128789778514, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 16775 + }, + { + "epoch": 0.16776, + "grad_norm": 1.360241301172709, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16776 + }, + { + "epoch": 0.16777, + "grad_norm": 1.4479593937643158, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 16777 + }, + { + "epoch": 0.16778, + "grad_norm": 1.2458349740823786, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 16778 + }, + { + "epoch": 0.16779, + "grad_norm": 1.259420554355854, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 16779 + }, + { + "epoch": 0.1678, + "grad_norm": 1.34903629188709, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 16780 + }, + { + "epoch": 0.16781, + "grad_norm": 1.286025725331968, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 16781 + }, + { + "epoch": 0.16782, + "grad_norm": 1.2291030746001255, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 16782 + }, + { + "epoch": 0.16783, + "grad_norm": 1.0875291955559694, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 16783 + }, + { + "epoch": 0.16784, + "grad_norm": 1.3159002512388083, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 16784 + }, + { + "epoch": 0.16785, + "grad_norm": 1.317527244791968, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 16785 + }, + { + "epoch": 0.16786, + "grad_norm": 1.4311664908854542, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 16786 + }, + { + "epoch": 0.16787, + "grad_norm": 1.1657445541783047, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 16787 + }, + { + "epoch": 0.16788, + "grad_norm": 1.1218297979981173, + "learning_rate": 0.003, + "loss": 4.039, + "step": 16788 + }, + { + "epoch": 0.16789, + "grad_norm": 1.2100040863439783, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 16789 + }, + { + "epoch": 0.1679, + "grad_norm": 1.2551242998126244, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 16790 + }, + { + "epoch": 0.16791, + "grad_norm": 1.1932540463897132, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 16791 + }, + { + "epoch": 0.16792, + "grad_norm": 1.2391779990331424, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 16792 + }, + { + "epoch": 0.16793, + "grad_norm": 1.0922719286288864, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 16793 + }, + { + "epoch": 0.16794, + "grad_norm": 1.27160864630517, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 16794 + }, + { + "epoch": 0.16795, + "grad_norm": 1.1913975588204553, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 16795 + }, + { + "epoch": 0.16796, + "grad_norm": 1.243028596016473, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 16796 + }, + { + "epoch": 0.16797, + "grad_norm": 1.0818591658569379, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16797 + }, + { + "epoch": 0.16798, + "grad_norm": 1.5692826015776473, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 16798 + }, + { + "epoch": 0.16799, + "grad_norm": 1.3033905320702153, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 16799 + }, + { + "epoch": 0.168, + "grad_norm": 1.270301971839668, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 16800 + }, + { + "epoch": 0.16801, + "grad_norm": 1.2606117354223052, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 16801 + }, + { + "epoch": 0.16802, + "grad_norm": 1.1290385301338544, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16802 + }, + { + "epoch": 0.16803, + "grad_norm": 1.3409426616809526, + "learning_rate": 0.003, + "loss": 4.051, + "step": 16803 + }, + { + "epoch": 0.16804, + "grad_norm": 1.1059978263354993, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 16804 + }, + { + "epoch": 0.16805, + "grad_norm": 1.3947787113601666, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 16805 + }, + { + "epoch": 0.16806, + "grad_norm": 0.9963101775114525, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 16806 + }, + { + "epoch": 0.16807, + "grad_norm": 1.5634406150316704, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 16807 + }, + { + "epoch": 0.16808, + "grad_norm": 1.2678649753363247, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16808 + }, + { + "epoch": 0.16809, + "grad_norm": 1.11171865107745, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 16809 + }, + { + "epoch": 0.1681, + "grad_norm": 1.2064696063825564, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 16810 + }, + { + "epoch": 0.16811, + "grad_norm": 1.2685610617105252, + "learning_rate": 0.003, + "loss": 4.039, + "step": 16811 + }, + { + "epoch": 0.16812, + "grad_norm": 1.242413676772281, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 16812 + }, + { + "epoch": 0.16813, + "grad_norm": 1.1500362337080925, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 16813 + }, + { + "epoch": 0.16814, + "grad_norm": 1.3192617974063514, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 16814 + }, + { + "epoch": 0.16815, + "grad_norm": 1.230514206708865, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 16815 + }, + { + "epoch": 0.16816, + "grad_norm": 1.4023838619816673, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 16816 + }, + { + "epoch": 0.16817, + "grad_norm": 1.09004068408355, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 16817 + }, + { + "epoch": 0.16818, + "grad_norm": 1.2087619042288684, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 16818 + }, + { + "epoch": 0.16819, + "grad_norm": 1.0940419536959587, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 16819 + }, + { + "epoch": 0.1682, + "grad_norm": 1.356788098447559, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 16820 + }, + { + "epoch": 0.16821, + "grad_norm": 1.1814260113034483, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 16821 + }, + { + "epoch": 0.16822, + "grad_norm": 1.083041113799058, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 16822 + }, + { + "epoch": 0.16823, + "grad_norm": 1.3845447100976924, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 16823 + }, + { + "epoch": 0.16824, + "grad_norm": 1.083888232716743, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 16824 + }, + { + "epoch": 0.16825, + "grad_norm": 1.5547189734994098, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 16825 + }, + { + "epoch": 0.16826, + "grad_norm": 1.1505415556402658, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 16826 + }, + { + "epoch": 0.16827, + "grad_norm": 1.516656079533665, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 16827 + }, + { + "epoch": 0.16828, + "grad_norm": 1.0255557130551507, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 16828 + }, + { + "epoch": 0.16829, + "grad_norm": 1.3575562934723178, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 16829 + }, + { + "epoch": 0.1683, + "grad_norm": 1.1988117222499541, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16830 + }, + { + "epoch": 0.16831, + "grad_norm": 1.4874532719939633, + "learning_rate": 0.003, + "loss": 3.994, + "step": 16831 + }, + { + "epoch": 0.16832, + "grad_norm": 1.1065575081516017, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 16832 + }, + { + "epoch": 0.16833, + "grad_norm": 1.361580708209208, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 16833 + }, + { + "epoch": 0.16834, + "grad_norm": 1.078641506329546, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 16834 + }, + { + "epoch": 0.16835, + "grad_norm": 1.2498711449421978, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 16835 + }, + { + "epoch": 0.16836, + "grad_norm": 1.2074394083708289, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 16836 + }, + { + "epoch": 0.16837, + "grad_norm": 1.4289220670388905, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 16837 + }, + { + "epoch": 0.16838, + "grad_norm": 1.1106440851025856, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 16838 + }, + { + "epoch": 0.16839, + "grad_norm": 1.4335390418677294, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 16839 + }, + { + "epoch": 0.1684, + "grad_norm": 1.0488019557253343, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 16840 + }, + { + "epoch": 0.16841, + "grad_norm": 1.3042148326398717, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16841 + }, + { + "epoch": 0.16842, + "grad_norm": 1.4603238658870208, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 16842 + }, + { + "epoch": 0.16843, + "grad_norm": 1.060291926461328, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 16843 + }, + { + "epoch": 0.16844, + "grad_norm": 1.405258245277971, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 16844 + }, + { + "epoch": 0.16845, + "grad_norm": 1.1607577698893636, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 16845 + }, + { + "epoch": 0.16846, + "grad_norm": 1.4783227381620572, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 16846 + }, + { + "epoch": 0.16847, + "grad_norm": 1.1648476762438198, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 16847 + }, + { + "epoch": 0.16848, + "grad_norm": 1.421633129727707, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16848 + }, + { + "epoch": 0.16849, + "grad_norm": 1.3381995418471722, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 16849 + }, + { + "epoch": 0.1685, + "grad_norm": 1.0817309204605705, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 16850 + }, + { + "epoch": 0.16851, + "grad_norm": 1.2280418994634874, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 16851 + }, + { + "epoch": 0.16852, + "grad_norm": 1.1433143202647051, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 16852 + }, + { + "epoch": 0.16853, + "grad_norm": 1.374223930656413, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 16853 + }, + { + "epoch": 0.16854, + "grad_norm": 1.1483790544696622, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 16854 + }, + { + "epoch": 0.16855, + "grad_norm": 1.3528775235819215, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 16855 + }, + { + "epoch": 0.16856, + "grad_norm": 1.177944326397307, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 16856 + }, + { + "epoch": 0.16857, + "grad_norm": 1.308171100592795, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 16857 + }, + { + "epoch": 0.16858, + "grad_norm": 1.1529652930812384, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 16858 + }, + { + "epoch": 0.16859, + "grad_norm": 1.2731702112081706, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 16859 + }, + { + "epoch": 0.1686, + "grad_norm": 1.31032159477305, + "learning_rate": 0.003, + "loss": 4.028, + "step": 16860 + }, + { + "epoch": 0.16861, + "grad_norm": 1.2036047048130094, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 16861 + }, + { + "epoch": 0.16862, + "grad_norm": 1.339073288144215, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16862 + }, + { + "epoch": 0.16863, + "grad_norm": 1.3109035788782717, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 16863 + }, + { + "epoch": 0.16864, + "grad_norm": 1.304758696665521, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 16864 + }, + { + "epoch": 0.16865, + "grad_norm": 1.2074240967263565, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 16865 + }, + { + "epoch": 0.16866, + "grad_norm": 1.1209272913654764, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 16866 + }, + { + "epoch": 0.16867, + "grad_norm": 1.4518110921716785, + "learning_rate": 0.003, + "loss": 4.009, + "step": 16867 + }, + { + "epoch": 0.16868, + "grad_norm": 1.0774290770428272, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16868 + }, + { + "epoch": 0.16869, + "grad_norm": 1.312204142314037, + "learning_rate": 0.003, + "loss": 4.009, + "step": 16869 + }, + { + "epoch": 0.1687, + "grad_norm": 1.1416925160063203, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 16870 + }, + { + "epoch": 0.16871, + "grad_norm": 1.5112168839952957, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 16871 + }, + { + "epoch": 0.16872, + "grad_norm": 1.1897944524848985, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 16872 + }, + { + "epoch": 0.16873, + "grad_norm": 1.446510519618927, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 16873 + }, + { + "epoch": 0.16874, + "grad_norm": 1.396135666890526, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 16874 + }, + { + "epoch": 0.16875, + "grad_norm": 1.433971567663403, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 16875 + }, + { + "epoch": 0.16876, + "grad_norm": 1.0044949337191382, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16876 + }, + { + "epoch": 0.16877, + "grad_norm": 1.4266779672159549, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 16877 + }, + { + "epoch": 0.16878, + "grad_norm": 1.29967644899772, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 16878 + }, + { + "epoch": 0.16879, + "grad_norm": 1.120701139574282, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 16879 + }, + { + "epoch": 0.1688, + "grad_norm": 1.2225522998809213, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 16880 + }, + { + "epoch": 0.16881, + "grad_norm": 1.2109024989470603, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 16881 + }, + { + "epoch": 0.16882, + "grad_norm": 1.4632096241684103, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 16882 + }, + { + "epoch": 0.16883, + "grad_norm": 1.2485484778905063, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 16883 + }, + { + "epoch": 0.16884, + "grad_norm": 1.1186898652467716, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 16884 + }, + { + "epoch": 0.16885, + "grad_norm": 1.24133162504021, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 16885 + }, + { + "epoch": 0.16886, + "grad_norm": 1.3050455876523348, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 16886 + }, + { + "epoch": 0.16887, + "grad_norm": 1.362781051037715, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 16887 + }, + { + "epoch": 0.16888, + "grad_norm": 1.3892189347450552, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 16888 + }, + { + "epoch": 0.16889, + "grad_norm": 1.1407710712129773, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16889 + }, + { + "epoch": 0.1689, + "grad_norm": 1.206006069080571, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 16890 + }, + { + "epoch": 0.16891, + "grad_norm": 1.1342045777811507, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 16891 + }, + { + "epoch": 0.16892, + "grad_norm": 1.2619875409492707, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 16892 + }, + { + "epoch": 0.16893, + "grad_norm": 1.2205092992378006, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 16893 + }, + { + "epoch": 0.16894, + "grad_norm": 1.244931242041419, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 16894 + }, + { + "epoch": 0.16895, + "grad_norm": 1.2651379871481943, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 16895 + }, + { + "epoch": 0.16896, + "grad_norm": 1.2366124296381513, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 16896 + }, + { + "epoch": 0.16897, + "grad_norm": 1.0590663890418193, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 16897 + }, + { + "epoch": 0.16898, + "grad_norm": 1.3479594975330957, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 16898 + }, + { + "epoch": 0.16899, + "grad_norm": 1.110658655779992, + "learning_rate": 0.003, + "loss": 4.017, + "step": 16899 + }, + { + "epoch": 0.169, + "grad_norm": 1.2805799269670264, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 16900 + }, + { + "epoch": 0.16901, + "grad_norm": 1.1028430432312288, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 16901 + }, + { + "epoch": 0.16902, + "grad_norm": 1.5531230378318743, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 16902 + }, + { + "epoch": 0.16903, + "grad_norm": 1.112848851912049, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 16903 + }, + { + "epoch": 0.16904, + "grad_norm": 1.3508659986287932, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 16904 + }, + { + "epoch": 0.16905, + "grad_norm": 1.3904795812482704, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 16905 + }, + { + "epoch": 0.16906, + "grad_norm": 1.3812560919461059, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 16906 + }, + { + "epoch": 0.16907, + "grad_norm": 1.2833538687555124, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 16907 + }, + { + "epoch": 0.16908, + "grad_norm": 1.2529061133847317, + "learning_rate": 0.003, + "loss": 4.012, + "step": 16908 + }, + { + "epoch": 0.16909, + "grad_norm": 1.1306501723012512, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 16909 + }, + { + "epoch": 0.1691, + "grad_norm": 1.145615530866832, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 16910 + }, + { + "epoch": 0.16911, + "grad_norm": 1.257679014582806, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 16911 + }, + { + "epoch": 0.16912, + "grad_norm": 1.1571753241793494, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 16912 + }, + { + "epoch": 0.16913, + "grad_norm": 1.2552940210106678, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 16913 + }, + { + "epoch": 0.16914, + "grad_norm": 1.384337884789972, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 16914 + }, + { + "epoch": 0.16915, + "grad_norm": 1.2197693395854996, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 16915 + }, + { + "epoch": 0.16916, + "grad_norm": 1.2880217635284237, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 16916 + }, + { + "epoch": 0.16917, + "grad_norm": 1.3384699793351602, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 16917 + }, + { + "epoch": 0.16918, + "grad_norm": 1.2616668365937014, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 16918 + }, + { + "epoch": 0.16919, + "grad_norm": 1.3222589677064205, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 16919 + }, + { + "epoch": 0.1692, + "grad_norm": 1.178864504225822, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 16920 + }, + { + "epoch": 0.16921, + "grad_norm": 1.2981216950894976, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 16921 + }, + { + "epoch": 0.16922, + "grad_norm": 1.5438283634887369, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 16922 + }, + { + "epoch": 0.16923, + "grad_norm": 1.152152700697969, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16923 + }, + { + "epoch": 0.16924, + "grad_norm": 1.2619466292301122, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 16924 + }, + { + "epoch": 0.16925, + "grad_norm": 1.2677460366996953, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 16925 + }, + { + "epoch": 0.16926, + "grad_norm": 1.0390056602589441, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 16926 + }, + { + "epoch": 0.16927, + "grad_norm": 1.3574529434547262, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 16927 + }, + { + "epoch": 0.16928, + "grad_norm": 1.0928639897909367, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 16928 + }, + { + "epoch": 0.16929, + "grad_norm": 1.3934850961846503, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 16929 + }, + { + "epoch": 0.1693, + "grad_norm": 1.285281464957926, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 16930 + }, + { + "epoch": 0.16931, + "grad_norm": 1.2455841870968274, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 16931 + }, + { + "epoch": 0.16932, + "grad_norm": 1.3341301746229626, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 16932 + }, + { + "epoch": 0.16933, + "grad_norm": 1.1781447497056439, + "learning_rate": 0.003, + "loss": 4.025, + "step": 16933 + }, + { + "epoch": 0.16934, + "grad_norm": 1.2428939057922601, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 16934 + }, + { + "epoch": 0.16935, + "grad_norm": 1.2483314679414697, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 16935 + }, + { + "epoch": 0.16936, + "grad_norm": 1.314712637897631, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 16936 + }, + { + "epoch": 0.16937, + "grad_norm": 1.0492346671508817, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 16937 + }, + { + "epoch": 0.16938, + "grad_norm": 1.3815854838892725, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 16938 + }, + { + "epoch": 0.16939, + "grad_norm": 1.2392019868274606, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 16939 + }, + { + "epoch": 0.1694, + "grad_norm": 1.5056024070904612, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 16940 + }, + { + "epoch": 0.16941, + "grad_norm": 1.2201284842797908, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 16941 + }, + { + "epoch": 0.16942, + "grad_norm": 1.3929057023313056, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 16942 + }, + { + "epoch": 0.16943, + "grad_norm": 1.2383122178551467, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 16943 + }, + { + "epoch": 0.16944, + "grad_norm": 1.101071882798595, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 16944 + }, + { + "epoch": 0.16945, + "grad_norm": 1.4592890179827827, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 16945 + }, + { + "epoch": 0.16946, + "grad_norm": 1.176261448397904, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 16946 + }, + { + "epoch": 0.16947, + "grad_norm": 1.1938508900306732, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 16947 + }, + { + "epoch": 0.16948, + "grad_norm": 1.2417706899331502, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 16948 + }, + { + "epoch": 0.16949, + "grad_norm": 1.119059964256359, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 16949 + }, + { + "epoch": 0.1695, + "grad_norm": 1.1741223880758327, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 16950 + }, + { + "epoch": 0.16951, + "grad_norm": 1.3448040518697508, + "learning_rate": 0.003, + "loss": 4.007, + "step": 16951 + }, + { + "epoch": 0.16952, + "grad_norm": 1.0488312214388276, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 16952 + }, + { + "epoch": 0.16953, + "grad_norm": 1.4062247086633848, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16953 + }, + { + "epoch": 0.16954, + "grad_norm": 1.2078394336426208, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 16954 + }, + { + "epoch": 0.16955, + "grad_norm": 1.2369747127025712, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 16955 + }, + { + "epoch": 0.16956, + "grad_norm": 1.2197357412428835, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16956 + }, + { + "epoch": 0.16957, + "grad_norm": 1.4609469443691683, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 16957 + }, + { + "epoch": 0.16958, + "grad_norm": 0.9557802790163199, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 16958 + }, + { + "epoch": 0.16959, + "grad_norm": 1.24108490074871, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 16959 + }, + { + "epoch": 0.1696, + "grad_norm": 1.4668437065833444, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16960 + }, + { + "epoch": 0.16961, + "grad_norm": 1.1743810113446638, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 16961 + }, + { + "epoch": 0.16962, + "grad_norm": 1.4978543000574625, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 16962 + }, + { + "epoch": 0.16963, + "grad_norm": 1.0906658346730245, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 16963 + }, + { + "epoch": 0.16964, + "grad_norm": 1.4468427659755054, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 16964 + }, + { + "epoch": 0.16965, + "grad_norm": 1.27416762346165, + "learning_rate": 0.003, + "loss": 3.988, + "step": 16965 + }, + { + "epoch": 0.16966, + "grad_norm": 1.331194166487556, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 16966 + }, + { + "epoch": 0.16967, + "grad_norm": 1.2625953737664501, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 16967 + }, + { + "epoch": 0.16968, + "grad_norm": 1.080363195070113, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 16968 + }, + { + "epoch": 0.16969, + "grad_norm": 1.5447983936107297, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 16969 + }, + { + "epoch": 0.1697, + "grad_norm": 1.2006316216642772, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 16970 + }, + { + "epoch": 0.16971, + "grad_norm": 1.3860194333678044, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 16971 + }, + { + "epoch": 0.16972, + "grad_norm": 1.2485380657128105, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 16972 + }, + { + "epoch": 0.16973, + "grad_norm": 1.1469982508816605, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 16973 + }, + { + "epoch": 0.16974, + "grad_norm": 1.2098045638378874, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 16974 + }, + { + "epoch": 0.16975, + "grad_norm": 1.2238529275774659, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 16975 + }, + { + "epoch": 0.16976, + "grad_norm": 1.0835686022976578, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 16976 + }, + { + "epoch": 0.16977, + "grad_norm": 1.5463250511671391, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 16977 + }, + { + "epoch": 0.16978, + "grad_norm": 1.156525371142893, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 16978 + }, + { + "epoch": 0.16979, + "grad_norm": 1.4567991308225594, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 16979 + }, + { + "epoch": 0.1698, + "grad_norm": 1.0656381636333407, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 16980 + }, + { + "epoch": 0.16981, + "grad_norm": 1.3190589658406147, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 16981 + }, + { + "epoch": 0.16982, + "grad_norm": 1.11372006687511, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 16982 + }, + { + "epoch": 0.16983, + "grad_norm": 1.363892577636926, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 16983 + }, + { + "epoch": 0.16984, + "grad_norm": 1.2820694465744142, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 16984 + }, + { + "epoch": 0.16985, + "grad_norm": 1.151936227728988, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 16985 + }, + { + "epoch": 0.16986, + "grad_norm": 1.2019003952022753, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 16986 + }, + { + "epoch": 0.16987, + "grad_norm": 1.2916031244486534, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 16987 + }, + { + "epoch": 0.16988, + "grad_norm": 1.3963732146288383, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 16988 + }, + { + "epoch": 0.16989, + "grad_norm": 1.2685891835699157, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 16989 + }, + { + "epoch": 0.1699, + "grad_norm": 1.2951575610634358, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 16990 + }, + { + "epoch": 0.16991, + "grad_norm": 1.2921496230383875, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 16991 + }, + { + "epoch": 0.16992, + "grad_norm": 1.1691614288218513, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 16992 + }, + { + "epoch": 0.16993, + "grad_norm": 1.3505549787048354, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 16993 + }, + { + "epoch": 0.16994, + "grad_norm": 1.234698424787448, + "learning_rate": 0.003, + "loss": 4.026, + "step": 16994 + }, + { + "epoch": 0.16995, + "grad_norm": 1.1635611711772356, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 16995 + }, + { + "epoch": 0.16996, + "grad_norm": 1.3932235075072121, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 16996 + }, + { + "epoch": 0.16997, + "grad_norm": 1.2329785004756628, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 16997 + }, + { + "epoch": 0.16998, + "grad_norm": 1.6130306233485228, + "learning_rate": 0.003, + "loss": 4.039, + "step": 16998 + }, + { + "epoch": 0.16999, + "grad_norm": 0.987298280415107, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 16999 + }, + { + "epoch": 0.17, + "grad_norm": 1.363353855910285, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 17000 + }, + { + "epoch": 0.17001, + "grad_norm": 1.072170500021738, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 17001 + }, + { + "epoch": 0.17002, + "grad_norm": 1.2286663236450293, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 17002 + }, + { + "epoch": 0.17003, + "grad_norm": 1.4430427644893247, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 17003 + }, + { + "epoch": 0.17004, + "grad_norm": 1.19606160541874, + "learning_rate": 0.003, + "loss": 4.011, + "step": 17004 + }, + { + "epoch": 0.17005, + "grad_norm": 1.1893825102433737, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 17005 + }, + { + "epoch": 0.17006, + "grad_norm": 1.4068763541598062, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 17006 + }, + { + "epoch": 0.17007, + "grad_norm": 1.2142864586823239, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 17007 + }, + { + "epoch": 0.17008, + "grad_norm": 1.2635706429106455, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 17008 + }, + { + "epoch": 0.17009, + "grad_norm": 1.1968596536847875, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 17009 + }, + { + "epoch": 0.1701, + "grad_norm": 1.2526564209820548, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 17010 + }, + { + "epoch": 0.17011, + "grad_norm": 1.1844228990235934, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 17011 + }, + { + "epoch": 0.17012, + "grad_norm": 1.1039154276178442, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 17012 + }, + { + "epoch": 0.17013, + "grad_norm": 1.54947273181041, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 17013 + }, + { + "epoch": 0.17014, + "grad_norm": 1.228717172082972, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 17014 + }, + { + "epoch": 0.17015, + "grad_norm": 1.4232491508453435, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 17015 + }, + { + "epoch": 0.17016, + "grad_norm": 1.330801365943393, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 17016 + }, + { + "epoch": 0.17017, + "grad_norm": 1.1359725998637589, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 17017 + }, + { + "epoch": 0.17018, + "grad_norm": 1.2227729610468783, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 17018 + }, + { + "epoch": 0.17019, + "grad_norm": 1.1216839963037502, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 17019 + }, + { + "epoch": 0.1702, + "grad_norm": 1.4395486441580458, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 17020 + }, + { + "epoch": 0.17021, + "grad_norm": 1.457913775333108, + "learning_rate": 0.003, + "loss": 4.005, + "step": 17021 + }, + { + "epoch": 0.17022, + "grad_norm": 1.2730261991335328, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17022 + }, + { + "epoch": 0.17023, + "grad_norm": 1.309025191561289, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 17023 + }, + { + "epoch": 0.17024, + "grad_norm": 1.222707520022976, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 17024 + }, + { + "epoch": 0.17025, + "grad_norm": 1.3516155725992163, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 17025 + }, + { + "epoch": 0.17026, + "grad_norm": 1.1329837729671406, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 17026 + }, + { + "epoch": 0.17027, + "grad_norm": 1.3633630624929551, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 17027 + }, + { + "epoch": 0.17028, + "grad_norm": 1.1815339013729904, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 17028 + }, + { + "epoch": 0.17029, + "grad_norm": 1.303776361036439, + "learning_rate": 0.003, + "loss": 4.031, + "step": 17029 + }, + { + "epoch": 0.1703, + "grad_norm": 1.1116647282472578, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 17030 + }, + { + "epoch": 0.17031, + "grad_norm": 1.6837890162486113, + "learning_rate": 0.003, + "loss": 4.033, + "step": 17031 + }, + { + "epoch": 0.17032, + "grad_norm": 1.0973423997807117, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 17032 + }, + { + "epoch": 0.17033, + "grad_norm": 1.4553041409424958, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 17033 + }, + { + "epoch": 0.17034, + "grad_norm": 1.067326827134977, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 17034 + }, + { + "epoch": 0.17035, + "grad_norm": 1.3964409257344448, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 17035 + }, + { + "epoch": 0.17036, + "grad_norm": 1.26893825218419, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 17036 + }, + { + "epoch": 0.17037, + "grad_norm": 1.1702590465709417, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 17037 + }, + { + "epoch": 0.17038, + "grad_norm": 1.2759678360880577, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 17038 + }, + { + "epoch": 0.17039, + "grad_norm": 1.2481788520067871, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 17039 + }, + { + "epoch": 0.1704, + "grad_norm": 1.3008871738106516, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 17040 + }, + { + "epoch": 0.17041, + "grad_norm": 0.993458540634306, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 17041 + }, + { + "epoch": 0.17042, + "grad_norm": 1.6483842502616208, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 17042 + }, + { + "epoch": 0.17043, + "grad_norm": 1.1823012906917088, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 17043 + }, + { + "epoch": 0.17044, + "grad_norm": 1.2834132877499729, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 17044 + }, + { + "epoch": 0.17045, + "grad_norm": 1.3017727577823928, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 17045 + }, + { + "epoch": 0.17046, + "grad_norm": 1.2997819468279899, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 17046 + }, + { + "epoch": 0.17047, + "grad_norm": 0.9669334493330405, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 17047 + }, + { + "epoch": 0.17048, + "grad_norm": 1.1175094646396293, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 17048 + }, + { + "epoch": 0.17049, + "grad_norm": 1.243380875464536, + "learning_rate": 0.003, + "loss": 3.985, + "step": 17049 + }, + { + "epoch": 0.1705, + "grad_norm": 1.2167661535688556, + "learning_rate": 0.003, + "loss": 4.005, + "step": 17050 + }, + { + "epoch": 0.17051, + "grad_norm": 1.4073761882344242, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 17051 + }, + { + "epoch": 0.17052, + "grad_norm": 1.2079374224606234, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 17052 + }, + { + "epoch": 0.17053, + "grad_norm": 1.1232919573416105, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 17053 + }, + { + "epoch": 0.17054, + "grad_norm": 1.6895509107983848, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 17054 + }, + { + "epoch": 0.17055, + "grad_norm": 1.1542580858819775, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 17055 + }, + { + "epoch": 0.17056, + "grad_norm": 1.208307073301473, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 17056 + }, + { + "epoch": 0.17057, + "grad_norm": 1.3242158599574259, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 17057 + }, + { + "epoch": 0.17058, + "grad_norm": 1.3255330368926068, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17058 + }, + { + "epoch": 0.17059, + "grad_norm": 1.1849168511846713, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 17059 + }, + { + "epoch": 0.1706, + "grad_norm": 1.3666009882279797, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 17060 + }, + { + "epoch": 0.17061, + "grad_norm": 1.3262779207828668, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 17061 + }, + { + "epoch": 0.17062, + "grad_norm": 1.0173614017135213, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 17062 + }, + { + "epoch": 0.17063, + "grad_norm": 1.2363760775920116, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 17063 + }, + { + "epoch": 0.17064, + "grad_norm": 1.0837262558950076, + "learning_rate": 0.003, + "loss": 4.01, + "step": 17064 + }, + { + "epoch": 0.17065, + "grad_norm": 1.3559097648251028, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17065 + }, + { + "epoch": 0.17066, + "grad_norm": 1.1819821915593431, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 17066 + }, + { + "epoch": 0.17067, + "grad_norm": 1.386641191079537, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 17067 + }, + { + "epoch": 0.17068, + "grad_norm": 1.1292876769780475, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 17068 + }, + { + "epoch": 0.17069, + "grad_norm": 1.3903157711241252, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 17069 + }, + { + "epoch": 0.1707, + "grad_norm": 1.136373967026405, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 17070 + }, + { + "epoch": 0.17071, + "grad_norm": 1.501179003007741, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 17071 + }, + { + "epoch": 0.17072, + "grad_norm": 0.9506830073519514, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 17072 + }, + { + "epoch": 0.17073, + "grad_norm": 1.3292716085551004, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 17073 + }, + { + "epoch": 0.17074, + "grad_norm": 1.3345983635658956, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 17074 + }, + { + "epoch": 0.17075, + "grad_norm": 1.390323079016144, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 17075 + }, + { + "epoch": 0.17076, + "grad_norm": 1.3772049791967669, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 17076 + }, + { + "epoch": 0.17077, + "grad_norm": 1.3685662444124322, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 17077 + }, + { + "epoch": 0.17078, + "grad_norm": 1.2048938743876443, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 17078 + }, + { + "epoch": 0.17079, + "grad_norm": 1.2550198188663826, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 17079 + }, + { + "epoch": 0.1708, + "grad_norm": 1.2509248960232697, + "learning_rate": 0.003, + "loss": 4.0, + "step": 17080 + }, + { + "epoch": 0.17081, + "grad_norm": 1.3155347659383203, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 17081 + }, + { + "epoch": 0.17082, + "grad_norm": 1.2297799179188096, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 17082 + }, + { + "epoch": 0.17083, + "grad_norm": 1.2401252434978838, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 17083 + }, + { + "epoch": 0.17084, + "grad_norm": 1.119392793832822, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 17084 + }, + { + "epoch": 0.17085, + "grad_norm": 1.1671328595938733, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 17085 + }, + { + "epoch": 0.17086, + "grad_norm": 1.384826264679687, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 17086 + }, + { + "epoch": 0.17087, + "grad_norm": 1.5263313659422078, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 17087 + }, + { + "epoch": 0.17088, + "grad_norm": 0.9182959568931038, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 17088 + }, + { + "epoch": 0.17089, + "grad_norm": 1.284208012928188, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 17089 + }, + { + "epoch": 0.1709, + "grad_norm": 1.2661764940608917, + "learning_rate": 0.003, + "loss": 4.005, + "step": 17090 + }, + { + "epoch": 0.17091, + "grad_norm": 1.280902038637925, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 17091 + }, + { + "epoch": 0.17092, + "grad_norm": 1.382288880310461, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 17092 + }, + { + "epoch": 0.17093, + "grad_norm": 1.1166438612003322, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 17093 + }, + { + "epoch": 0.17094, + "grad_norm": 1.5879639454360261, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 17094 + }, + { + "epoch": 0.17095, + "grad_norm": 1.242611008549483, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 17095 + }, + { + "epoch": 0.17096, + "grad_norm": 1.3686366406113646, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 17096 + }, + { + "epoch": 0.17097, + "grad_norm": 1.2406692703157454, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 17097 + }, + { + "epoch": 0.17098, + "grad_norm": 1.3392783702470064, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 17098 + }, + { + "epoch": 0.17099, + "grad_norm": 1.3207583531649192, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 17099 + }, + { + "epoch": 0.171, + "grad_norm": 0.9115212843188456, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 17100 + }, + { + "epoch": 0.17101, + "grad_norm": 1.4207831207188837, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17101 + }, + { + "epoch": 0.17102, + "grad_norm": 1.0106793121437063, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 17102 + }, + { + "epoch": 0.17103, + "grad_norm": 1.4870568757014018, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 17103 + }, + { + "epoch": 0.17104, + "grad_norm": 1.1297006257174922, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 17104 + }, + { + "epoch": 0.17105, + "grad_norm": 1.4574497121201375, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 17105 + }, + { + "epoch": 0.17106, + "grad_norm": 1.279203515550617, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 17106 + }, + { + "epoch": 0.17107, + "grad_norm": 1.120641556247558, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 17107 + }, + { + "epoch": 0.17108, + "grad_norm": 1.3611441573055538, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 17108 + }, + { + "epoch": 0.17109, + "grad_norm": 1.1597597254319019, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 17109 + }, + { + "epoch": 0.1711, + "grad_norm": 1.623200123412091, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 17110 + }, + { + "epoch": 0.17111, + "grad_norm": 1.0573921747614077, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 17111 + }, + { + "epoch": 0.17112, + "grad_norm": 1.3914494160558084, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 17112 + }, + { + "epoch": 0.17113, + "grad_norm": 1.0537510192201738, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 17113 + }, + { + "epoch": 0.17114, + "grad_norm": 1.4343479118895226, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 17114 + }, + { + "epoch": 0.17115, + "grad_norm": 1.0597295689465882, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 17115 + }, + { + "epoch": 0.17116, + "grad_norm": 1.4008441502145679, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 17116 + }, + { + "epoch": 0.17117, + "grad_norm": 1.1533115489494603, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 17117 + }, + { + "epoch": 0.17118, + "grad_norm": 1.3389173282580609, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 17118 + }, + { + "epoch": 0.17119, + "grad_norm": 1.0763942311886536, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 17119 + }, + { + "epoch": 0.1712, + "grad_norm": 1.3857008464880827, + "learning_rate": 0.003, + "loss": 4.014, + "step": 17120 + }, + { + "epoch": 0.17121, + "grad_norm": 1.1795518489267065, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 17121 + }, + { + "epoch": 0.17122, + "grad_norm": 1.2837531808444262, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 17122 + }, + { + "epoch": 0.17123, + "grad_norm": 1.213157552417634, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 17123 + }, + { + "epoch": 0.17124, + "grad_norm": 1.1134858549742377, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 17124 + }, + { + "epoch": 0.17125, + "grad_norm": 1.256476065981997, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 17125 + }, + { + "epoch": 0.17126, + "grad_norm": 1.2994744510899439, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17126 + }, + { + "epoch": 0.17127, + "grad_norm": 1.3261576115703024, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 17127 + }, + { + "epoch": 0.17128, + "grad_norm": 1.509759654632934, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 17128 + }, + { + "epoch": 0.17129, + "grad_norm": 1.157129471144684, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 17129 + }, + { + "epoch": 0.1713, + "grad_norm": 1.3506595690549228, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 17130 + }, + { + "epoch": 0.17131, + "grad_norm": 1.1810803692557885, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 17131 + }, + { + "epoch": 0.17132, + "grad_norm": 1.198309547444243, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17132 + }, + { + "epoch": 0.17133, + "grad_norm": 1.2633420447354877, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 17133 + }, + { + "epoch": 0.17134, + "grad_norm": 1.2609231995929508, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 17134 + }, + { + "epoch": 0.17135, + "grad_norm": 1.0653434516998657, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 17135 + }, + { + "epoch": 0.17136, + "grad_norm": 1.3862357041857178, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17136 + }, + { + "epoch": 0.17137, + "grad_norm": 1.1175033017808103, + "learning_rate": 0.003, + "loss": 4.015, + "step": 17137 + }, + { + "epoch": 0.17138, + "grad_norm": 1.5237743028823334, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 17138 + }, + { + "epoch": 0.17139, + "grad_norm": 1.175540536081767, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 17139 + }, + { + "epoch": 0.1714, + "grad_norm": 1.6728633199713923, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 17140 + }, + { + "epoch": 0.17141, + "grad_norm": 1.308263822363612, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 17141 + }, + { + "epoch": 0.17142, + "grad_norm": 1.1339564662254087, + "learning_rate": 0.003, + "loss": 3.984, + "step": 17142 + }, + { + "epoch": 0.17143, + "grad_norm": 1.2650050455155888, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 17143 + }, + { + "epoch": 0.17144, + "grad_norm": 1.163011141852609, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17144 + }, + { + "epoch": 0.17145, + "grad_norm": 1.5102254688904788, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 17145 + }, + { + "epoch": 0.17146, + "grad_norm": 1.2037448878305657, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 17146 + }, + { + "epoch": 0.17147, + "grad_norm": 1.5557240481772199, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 17147 + }, + { + "epoch": 0.17148, + "grad_norm": 1.0974950524747786, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 17148 + }, + { + "epoch": 0.17149, + "grad_norm": 1.3844767753715188, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 17149 + }, + { + "epoch": 0.1715, + "grad_norm": 1.319276395045874, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 17150 + }, + { + "epoch": 0.17151, + "grad_norm": 1.026204186557338, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 17151 + }, + { + "epoch": 0.17152, + "grad_norm": 1.2994904639201104, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 17152 + }, + { + "epoch": 0.17153, + "grad_norm": 1.1432778565248294, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17153 + }, + { + "epoch": 0.17154, + "grad_norm": 1.3697557622494088, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 17154 + }, + { + "epoch": 0.17155, + "grad_norm": 0.9894500157329774, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17155 + }, + { + "epoch": 0.17156, + "grad_norm": 1.3056644822045587, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 17156 + }, + { + "epoch": 0.17157, + "grad_norm": 1.282287938433088, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 17157 + }, + { + "epoch": 0.17158, + "grad_norm": 1.5061809099266898, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 17158 + }, + { + "epoch": 0.17159, + "grad_norm": 1.1371383950913954, + "learning_rate": 0.003, + "loss": 4.03, + "step": 17159 + }, + { + "epoch": 0.1716, + "grad_norm": 1.1832037804357762, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 17160 + }, + { + "epoch": 0.17161, + "grad_norm": 1.2545340918402512, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 17161 + }, + { + "epoch": 0.17162, + "grad_norm": 1.250357599105719, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 17162 + }, + { + "epoch": 0.17163, + "grad_norm": 1.4133447120518978, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 17163 + }, + { + "epoch": 0.17164, + "grad_norm": 1.1332442236888076, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 17164 + }, + { + "epoch": 0.17165, + "grad_norm": 1.5546016808686114, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 17165 + }, + { + "epoch": 0.17166, + "grad_norm": 1.0412075628559774, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 17166 + }, + { + "epoch": 0.17167, + "grad_norm": 1.4810738774004917, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 17167 + }, + { + "epoch": 0.17168, + "grad_norm": 1.0665666006787518, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 17168 + }, + { + "epoch": 0.17169, + "grad_norm": 1.3356993977397127, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 17169 + }, + { + "epoch": 0.1717, + "grad_norm": 1.35007975105471, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 17170 + }, + { + "epoch": 0.17171, + "grad_norm": 1.2221539025536585, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 17171 + }, + { + "epoch": 0.17172, + "grad_norm": 1.2277033051888775, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 17172 + }, + { + "epoch": 0.17173, + "grad_norm": 1.2042206000526587, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 17173 + }, + { + "epoch": 0.17174, + "grad_norm": 1.2356537217268861, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 17174 + }, + { + "epoch": 0.17175, + "grad_norm": 1.2937051183249795, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 17175 + }, + { + "epoch": 0.17176, + "grad_norm": 1.1750530246637987, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 17176 + }, + { + "epoch": 0.17177, + "grad_norm": 1.1435867677060834, + "learning_rate": 0.003, + "loss": 3.985, + "step": 17177 + }, + { + "epoch": 0.17178, + "grad_norm": 1.1244967318359802, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 17178 + }, + { + "epoch": 0.17179, + "grad_norm": 1.4152304529583006, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 17179 + }, + { + "epoch": 0.1718, + "grad_norm": 1.1261904961191211, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 17180 + }, + { + "epoch": 0.17181, + "grad_norm": 1.5335060142886194, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 17181 + }, + { + "epoch": 0.17182, + "grad_norm": 1.2391843231048232, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 17182 + }, + { + "epoch": 0.17183, + "grad_norm": 1.2088367421460817, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 17183 + }, + { + "epoch": 0.17184, + "grad_norm": 1.267094285490875, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 17184 + }, + { + "epoch": 0.17185, + "grad_norm": 1.0561147685321686, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 17185 + }, + { + "epoch": 0.17186, + "grad_norm": 1.3648888221954985, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 17186 + }, + { + "epoch": 0.17187, + "grad_norm": 1.514129993387543, + "learning_rate": 0.003, + "loss": 4.031, + "step": 17187 + }, + { + "epoch": 0.17188, + "grad_norm": 1.096699626012865, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 17188 + }, + { + "epoch": 0.17189, + "grad_norm": 1.4513064755452996, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 17189 + }, + { + "epoch": 0.1719, + "grad_norm": 1.0785963548641389, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 17190 + }, + { + "epoch": 0.17191, + "grad_norm": 1.3173078321818739, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 17191 + }, + { + "epoch": 0.17192, + "grad_norm": 1.221197625265541, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 17192 + }, + { + "epoch": 0.17193, + "grad_norm": 1.1767196120016654, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 17193 + }, + { + "epoch": 0.17194, + "grad_norm": 1.2755338735482418, + "learning_rate": 0.003, + "loss": 4.001, + "step": 17194 + }, + { + "epoch": 0.17195, + "grad_norm": 1.1647728885087472, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 17195 + }, + { + "epoch": 0.17196, + "grad_norm": 1.3491590116637575, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 17196 + }, + { + "epoch": 0.17197, + "grad_norm": 1.0154907883507833, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 17197 + }, + { + "epoch": 0.17198, + "grad_norm": 1.6514837011453636, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 17198 + }, + { + "epoch": 0.17199, + "grad_norm": 1.2249600301431616, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17199 + }, + { + "epoch": 0.172, + "grad_norm": 1.6335846575515012, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 17200 + }, + { + "epoch": 0.17201, + "grad_norm": 0.970957332720887, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 17201 + }, + { + "epoch": 0.17202, + "grad_norm": 1.4736329370686672, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 17202 + }, + { + "epoch": 0.17203, + "grad_norm": 1.1642100779363067, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 17203 + }, + { + "epoch": 0.17204, + "grad_norm": 1.3738026949477042, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17204 + }, + { + "epoch": 0.17205, + "grad_norm": 1.3653211359108215, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 17205 + }, + { + "epoch": 0.17206, + "grad_norm": 1.1810839310821635, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 17206 + }, + { + "epoch": 0.17207, + "grad_norm": 1.2788381610917945, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 17207 + }, + { + "epoch": 0.17208, + "grad_norm": 1.343528758148611, + "learning_rate": 0.003, + "loss": 4.028, + "step": 17208 + }, + { + "epoch": 0.17209, + "grad_norm": 1.2044259761732348, + "learning_rate": 0.003, + "loss": 4.012, + "step": 17209 + }, + { + "epoch": 0.1721, + "grad_norm": 1.282205491394408, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 17210 + }, + { + "epoch": 0.17211, + "grad_norm": 1.441980895914016, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 17211 + }, + { + "epoch": 0.17212, + "grad_norm": 1.050487950193999, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 17212 + }, + { + "epoch": 0.17213, + "grad_norm": 1.5050463177550548, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 17213 + }, + { + "epoch": 0.17214, + "grad_norm": 1.1403063493006569, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 17214 + }, + { + "epoch": 0.17215, + "grad_norm": 1.3687090498085275, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 17215 + }, + { + "epoch": 0.17216, + "grad_norm": 1.149331732564463, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 17216 + }, + { + "epoch": 0.17217, + "grad_norm": 1.4725985013833562, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 17217 + }, + { + "epoch": 0.17218, + "grad_norm": 1.1358280974247414, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 17218 + }, + { + "epoch": 0.17219, + "grad_norm": 1.3384339930647065, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 17219 + }, + { + "epoch": 0.1722, + "grad_norm": 1.0902169071833985, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 17220 + }, + { + "epoch": 0.17221, + "grad_norm": 1.473353818398544, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 17221 + }, + { + "epoch": 0.17222, + "grad_norm": 0.904154848913335, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 17222 + }, + { + "epoch": 0.17223, + "grad_norm": 1.233038973013818, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 17223 + }, + { + "epoch": 0.17224, + "grad_norm": 1.2218626090563223, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 17224 + }, + { + "epoch": 0.17225, + "grad_norm": 1.5852766020655138, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 17225 + }, + { + "epoch": 0.17226, + "grad_norm": 1.0144098208355075, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 17226 + }, + { + "epoch": 0.17227, + "grad_norm": 1.6856679111747674, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 17227 + }, + { + "epoch": 0.17228, + "grad_norm": 1.1674976534137596, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 17228 + }, + { + "epoch": 0.17229, + "grad_norm": 1.1846419527574505, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 17229 + }, + { + "epoch": 0.1723, + "grad_norm": 1.3372635062550227, + "learning_rate": 0.003, + "loss": 4.029, + "step": 17230 + }, + { + "epoch": 0.17231, + "grad_norm": 1.4360328784576093, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17231 + }, + { + "epoch": 0.17232, + "grad_norm": 1.167948325233245, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 17232 + }, + { + "epoch": 0.17233, + "grad_norm": 1.5103516999920048, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 17233 + }, + { + "epoch": 0.17234, + "grad_norm": 1.1718884631666613, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 17234 + }, + { + "epoch": 0.17235, + "grad_norm": 1.256899195330911, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 17235 + }, + { + "epoch": 0.17236, + "grad_norm": 1.3231041800701269, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 17236 + }, + { + "epoch": 0.17237, + "grad_norm": 1.2089819685493517, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17237 + }, + { + "epoch": 0.17238, + "grad_norm": 1.3938837032385776, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 17238 + }, + { + "epoch": 0.17239, + "grad_norm": 1.0986323428112195, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17239 + }, + { + "epoch": 0.1724, + "grad_norm": 1.4926533238792659, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 17240 + }, + { + "epoch": 0.17241, + "grad_norm": 0.9565770312520956, + "learning_rate": 0.003, + "loss": 4.025, + "step": 17241 + }, + { + "epoch": 0.17242, + "grad_norm": 1.4964969318330428, + "learning_rate": 0.003, + "loss": 4.009, + "step": 17242 + }, + { + "epoch": 0.17243, + "grad_norm": 1.078719932273665, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 17243 + }, + { + "epoch": 0.17244, + "grad_norm": 1.4107399149627897, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 17244 + }, + { + "epoch": 0.17245, + "grad_norm": 1.205665776851779, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 17245 + }, + { + "epoch": 0.17246, + "grad_norm": 1.4098420085442245, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 17246 + }, + { + "epoch": 0.17247, + "grad_norm": 1.1738218195027936, + "learning_rate": 0.003, + "loss": 4.017, + "step": 17247 + }, + { + "epoch": 0.17248, + "grad_norm": 1.5534211984480004, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 17248 + }, + { + "epoch": 0.17249, + "grad_norm": 1.1059230496993113, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 17249 + }, + { + "epoch": 0.1725, + "grad_norm": 1.2068057270289108, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 17250 + }, + { + "epoch": 0.17251, + "grad_norm": 1.3653391554834, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17251 + }, + { + "epoch": 0.17252, + "grad_norm": 0.9652249172910172, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 17252 + }, + { + "epoch": 0.17253, + "grad_norm": 1.3883368862028067, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17253 + }, + { + "epoch": 0.17254, + "grad_norm": 1.1475381505515283, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 17254 + }, + { + "epoch": 0.17255, + "grad_norm": 1.1818737148016185, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 17255 + }, + { + "epoch": 0.17256, + "grad_norm": 1.3480812015205677, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 17256 + }, + { + "epoch": 0.17257, + "grad_norm": 1.3232272920505788, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 17257 + }, + { + "epoch": 0.17258, + "grad_norm": 1.5173980234137427, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 17258 + }, + { + "epoch": 0.17259, + "grad_norm": 1.2723554794602032, + "learning_rate": 0.003, + "loss": 4.027, + "step": 17259 + }, + { + "epoch": 0.1726, + "grad_norm": 1.1888805114386718, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 17260 + }, + { + "epoch": 0.17261, + "grad_norm": 1.2247012755221713, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 17261 + }, + { + "epoch": 0.17262, + "grad_norm": 1.148262136354316, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 17262 + }, + { + "epoch": 0.17263, + "grad_norm": 1.3775928289484927, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 17263 + }, + { + "epoch": 0.17264, + "grad_norm": 1.1251068551429848, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 17264 + }, + { + "epoch": 0.17265, + "grad_norm": 1.3824170647249685, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 17265 + }, + { + "epoch": 0.17266, + "grad_norm": 0.871810097084379, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 17266 + }, + { + "epoch": 0.17267, + "grad_norm": 0.9651045548460154, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 17267 + }, + { + "epoch": 0.17268, + "grad_norm": 1.2767085120092714, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 17268 + }, + { + "epoch": 0.17269, + "grad_norm": 1.1721368271222345, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 17269 + }, + { + "epoch": 0.1727, + "grad_norm": 1.1223776476237222, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 17270 + }, + { + "epoch": 0.17271, + "grad_norm": 1.311417321913556, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 17271 + }, + { + "epoch": 0.17272, + "grad_norm": 1.27141658246143, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 17272 + }, + { + "epoch": 0.17273, + "grad_norm": 1.4113108900497515, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 17273 + }, + { + "epoch": 0.17274, + "grad_norm": 1.3858642060179733, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 17274 + }, + { + "epoch": 0.17275, + "grad_norm": 1.1843334908306122, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 17275 + }, + { + "epoch": 0.17276, + "grad_norm": 1.331721774774919, + "learning_rate": 0.003, + "loss": 4.009, + "step": 17276 + }, + { + "epoch": 0.17277, + "grad_norm": 1.2843503356977932, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 17277 + }, + { + "epoch": 0.17278, + "grad_norm": 1.1975980849144978, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 17278 + }, + { + "epoch": 0.17279, + "grad_norm": 1.4014082743567955, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 17279 + }, + { + "epoch": 0.1728, + "grad_norm": 1.06715525384562, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 17280 + }, + { + "epoch": 0.17281, + "grad_norm": 1.62746983560525, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 17281 + }, + { + "epoch": 0.17282, + "grad_norm": 0.8849603978508457, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 17282 + }, + { + "epoch": 0.17283, + "grad_norm": 1.3812091147021786, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 17283 + }, + { + "epoch": 0.17284, + "grad_norm": 1.2985208598972184, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 17284 + }, + { + "epoch": 0.17285, + "grad_norm": 1.228938514266029, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 17285 + }, + { + "epoch": 0.17286, + "grad_norm": 1.3800775260071658, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 17286 + }, + { + "epoch": 0.17287, + "grad_norm": 1.1335184142844406, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 17287 + }, + { + "epoch": 0.17288, + "grad_norm": 1.6187836779866542, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 17288 + }, + { + "epoch": 0.17289, + "grad_norm": 1.0880921795295901, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 17289 + }, + { + "epoch": 0.1729, + "grad_norm": 1.3183469402744639, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 17290 + }, + { + "epoch": 0.17291, + "grad_norm": 1.1312108474649964, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 17291 + }, + { + "epoch": 0.17292, + "grad_norm": 1.2583469188890024, + "learning_rate": 0.003, + "loss": 3.994, + "step": 17292 + }, + { + "epoch": 0.17293, + "grad_norm": 1.0036171406305894, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 17293 + }, + { + "epoch": 0.17294, + "grad_norm": 1.2717671981000425, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 17294 + }, + { + "epoch": 0.17295, + "grad_norm": 1.3573026646066382, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 17295 + }, + { + "epoch": 0.17296, + "grad_norm": 1.1595736297044072, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 17296 + }, + { + "epoch": 0.17297, + "grad_norm": 1.3391544056038518, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 17297 + }, + { + "epoch": 0.17298, + "grad_norm": 1.3052310544504753, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 17298 + }, + { + "epoch": 0.17299, + "grad_norm": 1.3962524411100379, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 17299 + }, + { + "epoch": 0.173, + "grad_norm": 1.313060365360365, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 17300 + }, + { + "epoch": 0.17301, + "grad_norm": 1.283124319072189, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 17301 + }, + { + "epoch": 0.17302, + "grad_norm": 1.2925301300118133, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 17302 + }, + { + "epoch": 0.17303, + "grad_norm": 1.1853360077635136, + "learning_rate": 0.003, + "loss": 4.009, + "step": 17303 + }, + { + "epoch": 0.17304, + "grad_norm": 1.20723607947912, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 17304 + }, + { + "epoch": 0.17305, + "grad_norm": 1.3688554411898706, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 17305 + }, + { + "epoch": 0.17306, + "grad_norm": 1.2597179043885396, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 17306 + }, + { + "epoch": 0.17307, + "grad_norm": 1.3381299531733732, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 17307 + }, + { + "epoch": 0.17308, + "grad_norm": 1.0221588315818542, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 17308 + }, + { + "epoch": 0.17309, + "grad_norm": 1.4406904812727543, + "learning_rate": 0.003, + "loss": 4.04, + "step": 17309 + }, + { + "epoch": 0.1731, + "grad_norm": 1.0715062331314942, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 17310 + }, + { + "epoch": 0.17311, + "grad_norm": 1.4638372254172154, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 17311 + }, + { + "epoch": 0.17312, + "grad_norm": 1.0568515584991716, + "learning_rate": 0.003, + "loss": 4.004, + "step": 17312 + }, + { + "epoch": 0.17313, + "grad_norm": 1.39433827191643, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 17313 + }, + { + "epoch": 0.17314, + "grad_norm": 0.9855272687414648, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 17314 + }, + { + "epoch": 0.17315, + "grad_norm": 1.2762421632760321, + "learning_rate": 0.003, + "loss": 4.006, + "step": 17315 + }, + { + "epoch": 0.17316, + "grad_norm": 1.077189428086061, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 17316 + }, + { + "epoch": 0.17317, + "grad_norm": 1.667953493782491, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 17317 + }, + { + "epoch": 0.17318, + "grad_norm": 1.3129642827230545, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 17318 + }, + { + "epoch": 0.17319, + "grad_norm": 1.225283913272662, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 17319 + }, + { + "epoch": 0.1732, + "grad_norm": 1.4212125329003817, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 17320 + }, + { + "epoch": 0.17321, + "grad_norm": 1.3433281763166263, + "learning_rate": 0.003, + "loss": 4.029, + "step": 17321 + }, + { + "epoch": 0.17322, + "grad_norm": 1.2880076072785773, + "learning_rate": 0.003, + "loss": 4.018, + "step": 17322 + }, + { + "epoch": 0.17323, + "grad_norm": 1.2690146848678605, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17323 + }, + { + "epoch": 0.17324, + "grad_norm": 1.1182734624267912, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 17324 + }, + { + "epoch": 0.17325, + "grad_norm": 1.2846627438473532, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 17325 + }, + { + "epoch": 0.17326, + "grad_norm": 1.4021564989948474, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 17326 + }, + { + "epoch": 0.17327, + "grad_norm": 1.0692247719863897, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 17327 + }, + { + "epoch": 0.17328, + "grad_norm": 1.3707062697557013, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 17328 + }, + { + "epoch": 0.17329, + "grad_norm": 1.0903184574512028, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 17329 + }, + { + "epoch": 0.1733, + "grad_norm": 1.359809657985758, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 17330 + }, + { + "epoch": 0.17331, + "grad_norm": 1.276241963281705, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 17331 + }, + { + "epoch": 0.17332, + "grad_norm": 1.447065159401617, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 17332 + }, + { + "epoch": 0.17333, + "grad_norm": 1.0959714650664985, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 17333 + }, + { + "epoch": 0.17334, + "grad_norm": 1.2519159045010237, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 17334 + }, + { + "epoch": 0.17335, + "grad_norm": 1.6111332627475863, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 17335 + }, + { + "epoch": 0.17336, + "grad_norm": 1.0054733076695719, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 17336 + }, + { + "epoch": 0.17337, + "grad_norm": 1.4945123114552656, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 17337 + }, + { + "epoch": 0.17338, + "grad_norm": 1.0235002207606916, + "learning_rate": 0.003, + "loss": 3.974, + "step": 17338 + }, + { + "epoch": 0.17339, + "grad_norm": 1.4223634327208732, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 17339 + }, + { + "epoch": 0.1734, + "grad_norm": 1.1891548110046788, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 17340 + }, + { + "epoch": 0.17341, + "grad_norm": 1.2678570960019357, + "learning_rate": 0.003, + "loss": 3.997, + "step": 17341 + }, + { + "epoch": 0.17342, + "grad_norm": 1.1946809146193087, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 17342 + }, + { + "epoch": 0.17343, + "grad_norm": 1.3417607658892123, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 17343 + }, + { + "epoch": 0.17344, + "grad_norm": 1.279236075035203, + "learning_rate": 0.003, + "loss": 4.017, + "step": 17344 + }, + { + "epoch": 0.17345, + "grad_norm": 1.6873037813875194, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 17345 + }, + { + "epoch": 0.17346, + "grad_norm": 1.0906193772457748, + "learning_rate": 0.003, + "loss": 4.026, + "step": 17346 + }, + { + "epoch": 0.17347, + "grad_norm": 1.301669503890121, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 17347 + }, + { + "epoch": 0.17348, + "grad_norm": 1.1201010663394788, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 17348 + }, + { + "epoch": 0.17349, + "grad_norm": 1.2822773260180442, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 17349 + }, + { + "epoch": 0.1735, + "grad_norm": 1.269482955451877, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 17350 + }, + { + "epoch": 0.17351, + "grad_norm": 1.2707305975348144, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 17351 + }, + { + "epoch": 0.17352, + "grad_norm": 1.1494463357967402, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 17352 + }, + { + "epoch": 0.17353, + "grad_norm": 1.0918376495855586, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 17353 + }, + { + "epoch": 0.17354, + "grad_norm": 1.2566020425973183, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 17354 + }, + { + "epoch": 0.17355, + "grad_norm": 1.2542980099734522, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 17355 + }, + { + "epoch": 0.17356, + "grad_norm": 1.4794482403087317, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 17356 + }, + { + "epoch": 0.17357, + "grad_norm": 1.1625413074001234, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 17357 + }, + { + "epoch": 0.17358, + "grad_norm": 1.2878581349328948, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 17358 + }, + { + "epoch": 0.17359, + "grad_norm": 1.1218207169793912, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 17359 + }, + { + "epoch": 0.1736, + "grad_norm": 1.2486176787250105, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 17360 + }, + { + "epoch": 0.17361, + "grad_norm": 1.266343297913411, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 17361 + }, + { + "epoch": 0.17362, + "grad_norm": 1.479733028324367, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 17362 + }, + { + "epoch": 0.17363, + "grad_norm": 1.174357803273249, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 17363 + }, + { + "epoch": 0.17364, + "grad_norm": 1.3289916018824397, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 17364 + }, + { + "epoch": 0.17365, + "grad_norm": 1.2266454626377468, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 17365 + }, + { + "epoch": 0.17366, + "grad_norm": 1.1761585123971194, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 17366 + }, + { + "epoch": 0.17367, + "grad_norm": 1.3180785258166257, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 17367 + }, + { + "epoch": 0.17368, + "grad_norm": 1.1512397103307213, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 17368 + }, + { + "epoch": 0.17369, + "grad_norm": 1.3030786067405742, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 17369 + }, + { + "epoch": 0.1737, + "grad_norm": 1.4600400342286277, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 17370 + }, + { + "epoch": 0.17371, + "grad_norm": 0.9789559321499843, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 17371 + }, + { + "epoch": 0.17372, + "grad_norm": 1.406150001021043, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 17372 + }, + { + "epoch": 0.17373, + "grad_norm": 1.1831169481340245, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 17373 + }, + { + "epoch": 0.17374, + "grad_norm": 1.4770654324579267, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 17374 + }, + { + "epoch": 0.17375, + "grad_norm": 1.1501125608044909, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 17375 + }, + { + "epoch": 0.17376, + "grad_norm": 1.2110112945636147, + "learning_rate": 0.003, + "loss": 3.998, + "step": 17376 + }, + { + "epoch": 0.17377, + "grad_norm": 1.3738972981009299, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 17377 + }, + { + "epoch": 0.17378, + "grad_norm": 1.117572575502471, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 17378 + }, + { + "epoch": 0.17379, + "grad_norm": 1.3860507283902452, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 17379 + }, + { + "epoch": 0.1738, + "grad_norm": 1.0779684375512073, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 17380 + }, + { + "epoch": 0.17381, + "grad_norm": 1.5955300999519237, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 17381 + }, + { + "epoch": 0.17382, + "grad_norm": 1.2292224660826188, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 17382 + }, + { + "epoch": 0.17383, + "grad_norm": 1.3205397866165893, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 17383 + }, + { + "epoch": 0.17384, + "grad_norm": 1.1634386312620115, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 17384 + }, + { + "epoch": 0.17385, + "grad_norm": 1.33842741762835, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 17385 + }, + { + "epoch": 0.17386, + "grad_norm": 1.0875714130899277, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17386 + }, + { + "epoch": 0.17387, + "grad_norm": 1.3582376011663713, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 17387 + }, + { + "epoch": 0.17388, + "grad_norm": 1.1678374075460847, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 17388 + }, + { + "epoch": 0.17389, + "grad_norm": 1.211558868854294, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 17389 + }, + { + "epoch": 0.1739, + "grad_norm": 1.4528085360953462, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 17390 + }, + { + "epoch": 0.17391, + "grad_norm": 1.119391271340527, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 17391 + }, + { + "epoch": 0.17392, + "grad_norm": 1.3227918665408323, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 17392 + }, + { + "epoch": 0.17393, + "grad_norm": 0.9623420610117401, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 17393 + }, + { + "epoch": 0.17394, + "grad_norm": 1.4037951528357682, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 17394 + }, + { + "epoch": 0.17395, + "grad_norm": 1.058862036609257, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 17395 + }, + { + "epoch": 0.17396, + "grad_norm": 1.5177044219698903, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 17396 + }, + { + "epoch": 0.17397, + "grad_norm": 1.0067233592822795, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 17397 + }, + { + "epoch": 0.17398, + "grad_norm": 1.440925491596526, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 17398 + }, + { + "epoch": 0.17399, + "grad_norm": 1.1403915015967954, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 17399 + }, + { + "epoch": 0.174, + "grad_norm": 1.243404881400955, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 17400 + }, + { + "epoch": 0.17401, + "grad_norm": 1.4799671525618545, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 17401 + }, + { + "epoch": 0.17402, + "grad_norm": 1.2513926420361863, + "learning_rate": 0.003, + "loss": 4.007, + "step": 17402 + }, + { + "epoch": 0.17403, + "grad_norm": 1.4293955535980087, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 17403 + }, + { + "epoch": 0.17404, + "grad_norm": 1.0335425855082576, + "learning_rate": 0.003, + "loss": 4.007, + "step": 17404 + }, + { + "epoch": 0.17405, + "grad_norm": 1.1741411409646767, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 17405 + }, + { + "epoch": 0.17406, + "grad_norm": 1.3652937364359996, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 17406 + }, + { + "epoch": 0.17407, + "grad_norm": 1.2124474143939583, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 17407 + }, + { + "epoch": 0.17408, + "grad_norm": 1.565002645615215, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 17408 + }, + { + "epoch": 0.17409, + "grad_norm": 1.239366136435791, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 17409 + }, + { + "epoch": 0.1741, + "grad_norm": 1.0812481404551977, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 17410 + }, + { + "epoch": 0.17411, + "grad_norm": 1.5086957512968893, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 17411 + }, + { + "epoch": 0.17412, + "grad_norm": 1.1652147415973524, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 17412 + }, + { + "epoch": 0.17413, + "grad_norm": 1.1927128455337468, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 17413 + }, + { + "epoch": 0.17414, + "grad_norm": 1.3528521306498578, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 17414 + }, + { + "epoch": 0.17415, + "grad_norm": 1.0960703425326428, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 17415 + }, + { + "epoch": 0.17416, + "grad_norm": 1.4602162168314847, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 17416 + }, + { + "epoch": 0.17417, + "grad_norm": 1.0368443511701424, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 17417 + }, + { + "epoch": 0.17418, + "grad_norm": 1.6099977197910769, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 17418 + }, + { + "epoch": 0.17419, + "grad_norm": 1.1972675032513773, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 17419 + }, + { + "epoch": 0.1742, + "grad_norm": 1.3184638228001029, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 17420 + }, + { + "epoch": 0.17421, + "grad_norm": 1.176368382451766, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 17421 + }, + { + "epoch": 0.17422, + "grad_norm": 1.093838392847663, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 17422 + }, + { + "epoch": 0.17423, + "grad_norm": 1.3771944160059182, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 17423 + }, + { + "epoch": 0.17424, + "grad_norm": 1.2426775725792576, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 17424 + }, + { + "epoch": 0.17425, + "grad_norm": 1.3163419783944523, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 17425 + }, + { + "epoch": 0.17426, + "grad_norm": 0.97312063332047, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 17426 + }, + { + "epoch": 0.17427, + "grad_norm": 1.463229691284514, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 17427 + }, + { + "epoch": 0.17428, + "grad_norm": 1.2014268675563655, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 17428 + }, + { + "epoch": 0.17429, + "grad_norm": 1.2928611370309657, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 17429 + }, + { + "epoch": 0.1743, + "grad_norm": 1.236739863097322, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 17430 + }, + { + "epoch": 0.17431, + "grad_norm": 1.2581378071175113, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 17431 + }, + { + "epoch": 0.17432, + "grad_norm": 1.2825682371586564, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 17432 + }, + { + "epoch": 0.17433, + "grad_norm": 1.3083920982067194, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 17433 + }, + { + "epoch": 0.17434, + "grad_norm": 1.1835437320930857, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17434 + }, + { + "epoch": 0.17435, + "grad_norm": 1.2615839587476694, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 17435 + }, + { + "epoch": 0.17436, + "grad_norm": 1.3765442849644418, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 17436 + }, + { + "epoch": 0.17437, + "grad_norm": 1.0665678127620197, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 17437 + }, + { + "epoch": 0.17438, + "grad_norm": 1.6294484139660101, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 17438 + }, + { + "epoch": 0.17439, + "grad_norm": 1.2135317553416847, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 17439 + }, + { + "epoch": 0.1744, + "grad_norm": 1.3014420389996388, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 17440 + }, + { + "epoch": 0.17441, + "grad_norm": 1.1441838771751138, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 17441 + }, + { + "epoch": 0.17442, + "grad_norm": 1.5053710572281074, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 17442 + }, + { + "epoch": 0.17443, + "grad_norm": 0.9045722850562977, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 17443 + }, + { + "epoch": 0.17444, + "grad_norm": 1.154807886610162, + "learning_rate": 0.003, + "loss": 3.972, + "step": 17444 + }, + { + "epoch": 0.17445, + "grad_norm": 1.4581057868595215, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17445 + }, + { + "epoch": 0.17446, + "grad_norm": 1.102586921173529, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 17446 + }, + { + "epoch": 0.17447, + "grad_norm": 1.7142103095973686, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 17447 + }, + { + "epoch": 0.17448, + "grad_norm": 1.0883226149353684, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 17448 + }, + { + "epoch": 0.17449, + "grad_norm": 1.439175568795504, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 17449 + }, + { + "epoch": 0.1745, + "grad_norm": 0.9894201968795465, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 17450 + }, + { + "epoch": 0.17451, + "grad_norm": 1.1319851110611714, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 17451 + }, + { + "epoch": 0.17452, + "grad_norm": 1.4101783964755321, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 17452 + }, + { + "epoch": 0.17453, + "grad_norm": 1.1950631496714745, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 17453 + }, + { + "epoch": 0.17454, + "grad_norm": 1.2678178704996652, + "learning_rate": 0.003, + "loss": 4.014, + "step": 17454 + }, + { + "epoch": 0.17455, + "grad_norm": 1.29183275761241, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17455 + }, + { + "epoch": 0.17456, + "grad_norm": 1.3880097914219025, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 17456 + }, + { + "epoch": 0.17457, + "grad_norm": 1.4516062258836981, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 17457 + }, + { + "epoch": 0.17458, + "grad_norm": 1.0864160692418074, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 17458 + }, + { + "epoch": 0.17459, + "grad_norm": 1.1726149063068008, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 17459 + }, + { + "epoch": 0.1746, + "grad_norm": 1.0890544630370935, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 17460 + }, + { + "epoch": 0.17461, + "grad_norm": 1.3667948040505313, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 17461 + }, + { + "epoch": 0.17462, + "grad_norm": 1.1645584883877051, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 17462 + }, + { + "epoch": 0.17463, + "grad_norm": 1.503741330222054, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 17463 + }, + { + "epoch": 0.17464, + "grad_norm": 1.0489494837596345, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 17464 + }, + { + "epoch": 0.17465, + "grad_norm": 1.4163168678260742, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 17465 + }, + { + "epoch": 0.17466, + "grad_norm": 1.0905468833071308, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 17466 + }, + { + "epoch": 0.17467, + "grad_norm": 1.561498753185741, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 17467 + }, + { + "epoch": 0.17468, + "grad_norm": 1.2436964093941147, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 17468 + }, + { + "epoch": 0.17469, + "grad_norm": 1.3442079014440735, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 17469 + }, + { + "epoch": 0.1747, + "grad_norm": 1.3412844922898968, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 17470 + }, + { + "epoch": 0.17471, + "grad_norm": 1.5212025627883503, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 17471 + }, + { + "epoch": 0.17472, + "grad_norm": 1.1359504116512613, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 17472 + }, + { + "epoch": 0.17473, + "grad_norm": 1.189587217761058, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 17473 + }, + { + "epoch": 0.17474, + "grad_norm": 1.1467561240598758, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 17474 + }, + { + "epoch": 0.17475, + "grad_norm": 1.2190623306021742, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 17475 + }, + { + "epoch": 0.17476, + "grad_norm": 1.2077612593166749, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 17476 + }, + { + "epoch": 0.17477, + "grad_norm": 1.2805042241733537, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 17477 + }, + { + "epoch": 0.17478, + "grad_norm": 1.1128725501592887, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 17478 + }, + { + "epoch": 0.17479, + "grad_norm": 1.24886952802986, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 17479 + }, + { + "epoch": 0.1748, + "grad_norm": 1.2882811283564244, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 17480 + }, + { + "epoch": 0.17481, + "grad_norm": 1.3820476542413211, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 17481 + }, + { + "epoch": 0.17482, + "grad_norm": 1.4295014325350368, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 17482 + }, + { + "epoch": 0.17483, + "grad_norm": 1.239118088638119, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 17483 + }, + { + "epoch": 0.17484, + "grad_norm": 1.2489244256582073, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 17484 + }, + { + "epoch": 0.17485, + "grad_norm": 1.2527953693531593, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17485 + }, + { + "epoch": 0.17486, + "grad_norm": 1.2442497188525592, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 17486 + }, + { + "epoch": 0.17487, + "grad_norm": 1.2567200106459853, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 17487 + }, + { + "epoch": 0.17488, + "grad_norm": 1.3224776601648938, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 17488 + }, + { + "epoch": 0.17489, + "grad_norm": 1.0544687569406892, + "learning_rate": 0.003, + "loss": 3.987, + "step": 17489 + }, + { + "epoch": 0.1749, + "grad_norm": 1.3644368776681581, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 17490 + }, + { + "epoch": 0.17491, + "grad_norm": 1.170964249779947, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 17491 + }, + { + "epoch": 0.17492, + "grad_norm": 1.570199409081135, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 17492 + }, + { + "epoch": 0.17493, + "grad_norm": 1.1791374142942461, + "learning_rate": 0.003, + "loss": 4.015, + "step": 17493 + }, + { + "epoch": 0.17494, + "grad_norm": 1.4209200136606746, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17494 + }, + { + "epoch": 0.17495, + "grad_norm": 1.2564179478971467, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 17495 + }, + { + "epoch": 0.17496, + "grad_norm": 1.3155050931655181, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 17496 + }, + { + "epoch": 0.17497, + "grad_norm": 1.328719821056356, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 17497 + }, + { + "epoch": 0.17498, + "grad_norm": 1.4774943449101552, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 17498 + }, + { + "epoch": 0.17499, + "grad_norm": 1.2361239694490804, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 17499 + }, + { + "epoch": 0.175, + "grad_norm": 1.26961012920436, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 17500 + }, + { + "epoch": 0.17501, + "grad_norm": 1.168143621476018, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 17501 + }, + { + "epoch": 0.17502, + "grad_norm": 1.371389114647039, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 17502 + }, + { + "epoch": 0.17503, + "grad_norm": 1.21480838583015, + "learning_rate": 0.003, + "loss": 4.035, + "step": 17503 + }, + { + "epoch": 0.17504, + "grad_norm": 1.174257813679406, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17504 + }, + { + "epoch": 0.17505, + "grad_norm": 1.290775859106054, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 17505 + }, + { + "epoch": 0.17506, + "grad_norm": 1.3943836445178055, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 17506 + }, + { + "epoch": 0.17507, + "grad_norm": 1.1928197149111355, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 17507 + }, + { + "epoch": 0.17508, + "grad_norm": 1.3179143177195527, + "learning_rate": 0.003, + "loss": 4.001, + "step": 17508 + }, + { + "epoch": 0.17509, + "grad_norm": 1.106568451592994, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 17509 + }, + { + "epoch": 0.1751, + "grad_norm": 1.2945594796110989, + "learning_rate": 0.003, + "loss": 3.994, + "step": 17510 + }, + { + "epoch": 0.17511, + "grad_norm": 1.1046731165335164, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 17511 + }, + { + "epoch": 0.17512, + "grad_norm": 1.452680110654347, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 17512 + }, + { + "epoch": 0.17513, + "grad_norm": 1.2021451899954632, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 17513 + }, + { + "epoch": 0.17514, + "grad_norm": 1.2475477973757747, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 17514 + }, + { + "epoch": 0.17515, + "grad_norm": 1.26448299868792, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 17515 + }, + { + "epoch": 0.17516, + "grad_norm": 1.346066366740584, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 17516 + }, + { + "epoch": 0.17517, + "grad_norm": 1.2380737111410907, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 17517 + }, + { + "epoch": 0.17518, + "grad_norm": 1.2012239141263896, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 17518 + }, + { + "epoch": 0.17519, + "grad_norm": 1.2286300179862844, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 17519 + }, + { + "epoch": 0.1752, + "grad_norm": 1.1285470362368863, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 17520 + }, + { + "epoch": 0.17521, + "grad_norm": 1.4469269417899426, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 17521 + }, + { + "epoch": 0.17522, + "grad_norm": 0.9874807206339904, + "learning_rate": 0.003, + "loss": 3.99, + "step": 17522 + }, + { + "epoch": 0.17523, + "grad_norm": 1.4122407383518634, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 17523 + }, + { + "epoch": 0.17524, + "grad_norm": 1.0231215239865685, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 17524 + }, + { + "epoch": 0.17525, + "grad_norm": 1.390824344503427, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 17525 + }, + { + "epoch": 0.17526, + "grad_norm": 1.2058258237116968, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 17526 + }, + { + "epoch": 0.17527, + "grad_norm": 1.4580280986425835, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 17527 + }, + { + "epoch": 0.17528, + "grad_norm": 1.0776378794974155, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 17528 + }, + { + "epoch": 0.17529, + "grad_norm": 1.3549429661530912, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 17529 + }, + { + "epoch": 0.1753, + "grad_norm": 1.0128705323649436, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 17530 + }, + { + "epoch": 0.17531, + "grad_norm": 1.5774919045495472, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 17531 + }, + { + "epoch": 0.17532, + "grad_norm": 1.0305305958749693, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 17532 + }, + { + "epoch": 0.17533, + "grad_norm": 1.6209730948269785, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 17533 + }, + { + "epoch": 0.17534, + "grad_norm": 1.1751955702817969, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 17534 + }, + { + "epoch": 0.17535, + "grad_norm": 1.1649296906823114, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 17535 + }, + { + "epoch": 0.17536, + "grad_norm": 1.2174031909483718, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 17536 + }, + { + "epoch": 0.17537, + "grad_norm": 1.185420170654492, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 17537 + }, + { + "epoch": 0.17538, + "grad_norm": 1.3201442908408842, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 17538 + }, + { + "epoch": 0.17539, + "grad_norm": 1.3588723981014663, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 17539 + }, + { + "epoch": 0.1754, + "grad_norm": 1.4270005102548036, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 17540 + }, + { + "epoch": 0.17541, + "grad_norm": 1.1309033485992048, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 17541 + }, + { + "epoch": 0.17542, + "grad_norm": 1.1601972477484055, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 17542 + }, + { + "epoch": 0.17543, + "grad_norm": 1.2719466286625902, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 17543 + }, + { + "epoch": 0.17544, + "grad_norm": 1.2934184007525624, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 17544 + }, + { + "epoch": 0.17545, + "grad_norm": 1.1443382594276104, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 17545 + }, + { + "epoch": 0.17546, + "grad_norm": 1.484698575855083, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 17546 + }, + { + "epoch": 0.17547, + "grad_norm": 1.443084888027566, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 17547 + }, + { + "epoch": 0.17548, + "grad_norm": 1.3578247263605012, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 17548 + }, + { + "epoch": 0.17549, + "grad_norm": 1.1222740359797407, + "learning_rate": 0.003, + "loss": 3.993, + "step": 17549 + }, + { + "epoch": 0.1755, + "grad_norm": 1.2217045827974942, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 17550 + }, + { + "epoch": 0.17551, + "grad_norm": 1.2599905587835294, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 17551 + }, + { + "epoch": 0.17552, + "grad_norm": 1.3173881636461955, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 17552 + }, + { + "epoch": 0.17553, + "grad_norm": 1.1985283675532807, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17553 + }, + { + "epoch": 0.17554, + "grad_norm": 1.3535779532008505, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 17554 + }, + { + "epoch": 0.17555, + "grad_norm": 1.1593943577563153, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 17555 + }, + { + "epoch": 0.17556, + "grad_norm": 1.2212030190780252, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17556 + }, + { + "epoch": 0.17557, + "grad_norm": 1.1953193688900052, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 17557 + }, + { + "epoch": 0.17558, + "grad_norm": 1.284028278474116, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 17558 + }, + { + "epoch": 0.17559, + "grad_norm": 1.3281165223541755, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 17559 + }, + { + "epoch": 0.1756, + "grad_norm": 0.9785727433834605, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 17560 + }, + { + "epoch": 0.17561, + "grad_norm": 1.2432156776124126, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 17561 + }, + { + "epoch": 0.17562, + "grad_norm": 1.3246982368270601, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 17562 + }, + { + "epoch": 0.17563, + "grad_norm": 1.249921604351906, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 17563 + }, + { + "epoch": 0.17564, + "grad_norm": 1.2465591420404474, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 17564 + }, + { + "epoch": 0.17565, + "grad_norm": 1.2093675545638525, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 17565 + }, + { + "epoch": 0.17566, + "grad_norm": 1.421791354142211, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 17566 + }, + { + "epoch": 0.17567, + "grad_norm": 1.237781468795189, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 17567 + }, + { + "epoch": 0.17568, + "grad_norm": 1.0586259261472677, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 17568 + }, + { + "epoch": 0.17569, + "grad_norm": 1.3867036041741054, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 17569 + }, + { + "epoch": 0.1757, + "grad_norm": 1.3645783708200951, + "learning_rate": 0.003, + "loss": 4.004, + "step": 17570 + }, + { + "epoch": 0.17571, + "grad_norm": 1.362526636972413, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 17571 + }, + { + "epoch": 0.17572, + "grad_norm": 1.169807271156517, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 17572 + }, + { + "epoch": 0.17573, + "grad_norm": 1.2493866528248445, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 17573 + }, + { + "epoch": 0.17574, + "grad_norm": 1.2984770898637006, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 17574 + }, + { + "epoch": 0.17575, + "grad_norm": 1.0356437941386805, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 17575 + }, + { + "epoch": 0.17576, + "grad_norm": 1.4871383560855997, + "learning_rate": 0.003, + "loss": 4.028, + "step": 17576 + }, + { + "epoch": 0.17577, + "grad_norm": 1.150557329959256, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 17577 + }, + { + "epoch": 0.17578, + "grad_norm": 1.4133663757588073, + "learning_rate": 0.003, + "loss": 4.03, + "step": 17578 + }, + { + "epoch": 0.17579, + "grad_norm": 1.2581433405946318, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17579 + }, + { + "epoch": 0.1758, + "grad_norm": 1.4213042433507315, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 17580 + }, + { + "epoch": 0.17581, + "grad_norm": 1.2669235513462596, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 17581 + }, + { + "epoch": 0.17582, + "grad_norm": 1.366659090525752, + "learning_rate": 0.003, + "loss": 4.045, + "step": 17582 + }, + { + "epoch": 0.17583, + "grad_norm": 1.5443151445880576, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 17583 + }, + { + "epoch": 0.17584, + "grad_norm": 1.187717678251351, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 17584 + }, + { + "epoch": 0.17585, + "grad_norm": 1.566622961243868, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 17585 + }, + { + "epoch": 0.17586, + "grad_norm": 1.235649476915203, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 17586 + }, + { + "epoch": 0.17587, + "grad_norm": 1.379215757629583, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 17587 + }, + { + "epoch": 0.17588, + "grad_norm": 1.0614341038953146, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 17588 + }, + { + "epoch": 0.17589, + "grad_norm": 1.2648484455968334, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 17589 + }, + { + "epoch": 0.1759, + "grad_norm": 1.1576910924641934, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 17590 + }, + { + "epoch": 0.17591, + "grad_norm": 1.1431254154227177, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 17591 + }, + { + "epoch": 0.17592, + "grad_norm": 1.3161147509011362, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 17592 + }, + { + "epoch": 0.17593, + "grad_norm": 1.2696360298337392, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 17593 + }, + { + "epoch": 0.17594, + "grad_norm": 1.4224208341519127, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 17594 + }, + { + "epoch": 0.17595, + "grad_norm": 0.99950016326709, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 17595 + }, + { + "epoch": 0.17596, + "grad_norm": 1.5338785658460057, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 17596 + }, + { + "epoch": 0.17597, + "grad_norm": 1.0377250441976649, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 17597 + }, + { + "epoch": 0.17598, + "grad_norm": 1.6791282982424311, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 17598 + }, + { + "epoch": 0.17599, + "grad_norm": 1.0560648540651831, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 17599 + }, + { + "epoch": 0.176, + "grad_norm": 1.2152626989221356, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 17600 + }, + { + "epoch": 0.17601, + "grad_norm": 1.3408770196110984, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 17601 + }, + { + "epoch": 0.17602, + "grad_norm": 1.135109841606531, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 17602 + }, + { + "epoch": 0.17603, + "grad_norm": 1.243107289474913, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 17603 + }, + { + "epoch": 0.17604, + "grad_norm": 1.2852511433011993, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 17604 + }, + { + "epoch": 0.17605, + "grad_norm": 0.956126693243154, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 17605 + }, + { + "epoch": 0.17606, + "grad_norm": 1.635243932239004, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 17606 + }, + { + "epoch": 0.17607, + "grad_norm": 1.1330955529107827, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 17607 + }, + { + "epoch": 0.17608, + "grad_norm": 1.5430115729514313, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17608 + }, + { + "epoch": 0.17609, + "grad_norm": 1.08412975084303, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 17609 + }, + { + "epoch": 0.1761, + "grad_norm": 1.377042344730517, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 17610 + }, + { + "epoch": 0.17611, + "grad_norm": 1.1638370823196682, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17611 + }, + { + "epoch": 0.17612, + "grad_norm": 1.493808977247851, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 17612 + }, + { + "epoch": 0.17613, + "grad_norm": 1.3228465169531465, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 17613 + }, + { + "epoch": 0.17614, + "grad_norm": 1.3087121810410431, + "learning_rate": 0.003, + "loss": 3.989, + "step": 17614 + }, + { + "epoch": 0.17615, + "grad_norm": 1.2068750677649611, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 17615 + }, + { + "epoch": 0.17616, + "grad_norm": 1.0696833716922531, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 17616 + }, + { + "epoch": 0.17617, + "grad_norm": 1.2759221671837773, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 17617 + }, + { + "epoch": 0.17618, + "grad_norm": 1.2043041905130851, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 17618 + }, + { + "epoch": 0.17619, + "grad_norm": 1.4037105214642187, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 17619 + }, + { + "epoch": 0.1762, + "grad_norm": 1.2242585181629013, + "learning_rate": 0.003, + "loss": 3.994, + "step": 17620 + }, + { + "epoch": 0.17621, + "grad_norm": 1.2206287301130856, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 17621 + }, + { + "epoch": 0.17622, + "grad_norm": 1.2835961455791998, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 17622 + }, + { + "epoch": 0.17623, + "grad_norm": 1.2070454470336767, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 17623 + }, + { + "epoch": 0.17624, + "grad_norm": 1.313277425758526, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 17624 + }, + { + "epoch": 0.17625, + "grad_norm": 1.2724185930197756, + "learning_rate": 0.003, + "loss": 3.997, + "step": 17625 + }, + { + "epoch": 0.17626, + "grad_norm": 1.1092201178149645, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 17626 + }, + { + "epoch": 0.17627, + "grad_norm": 1.402151390353694, + "learning_rate": 0.003, + "loss": 4.037, + "step": 17627 + }, + { + "epoch": 0.17628, + "grad_norm": 1.1428631040156145, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 17628 + }, + { + "epoch": 0.17629, + "grad_norm": 1.2710406949237625, + "learning_rate": 0.003, + "loss": 4.031, + "step": 17629 + }, + { + "epoch": 0.1763, + "grad_norm": 1.2241399705227562, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 17630 + }, + { + "epoch": 0.17631, + "grad_norm": 1.2815206513869735, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 17631 + }, + { + "epoch": 0.17632, + "grad_norm": 1.1123932464141089, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 17632 + }, + { + "epoch": 0.17633, + "grad_norm": 1.3194960667717541, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 17633 + }, + { + "epoch": 0.17634, + "grad_norm": 1.1513525106952893, + "learning_rate": 0.003, + "loss": 3.998, + "step": 17634 + }, + { + "epoch": 0.17635, + "grad_norm": 1.508444451922507, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 17635 + }, + { + "epoch": 0.17636, + "grad_norm": 1.2828155918425586, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 17636 + }, + { + "epoch": 0.17637, + "grad_norm": 1.275723512265431, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 17637 + }, + { + "epoch": 0.17638, + "grad_norm": 1.3655858171298656, + "learning_rate": 0.003, + "loss": 4.033, + "step": 17638 + }, + { + "epoch": 0.17639, + "grad_norm": 1.2186927399922851, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 17639 + }, + { + "epoch": 0.1764, + "grad_norm": 1.3689522848466271, + "learning_rate": 0.003, + "loss": 4.02, + "step": 17640 + }, + { + "epoch": 0.17641, + "grad_norm": 1.4312717569589986, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 17641 + }, + { + "epoch": 0.17642, + "grad_norm": 1.1798880942001364, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 17642 + }, + { + "epoch": 0.17643, + "grad_norm": 1.5216060480329352, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 17643 + }, + { + "epoch": 0.17644, + "grad_norm": 1.0725669386174144, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 17644 + }, + { + "epoch": 0.17645, + "grad_norm": 1.2059907916524846, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 17645 + }, + { + "epoch": 0.17646, + "grad_norm": 1.089964308292804, + "learning_rate": 0.003, + "loss": 3.991, + "step": 17646 + }, + { + "epoch": 0.17647, + "grad_norm": 1.2652122674975332, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 17647 + }, + { + "epoch": 0.17648, + "grad_norm": 1.0873367578770192, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 17648 + }, + { + "epoch": 0.17649, + "grad_norm": 1.350335271643384, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 17649 + }, + { + "epoch": 0.1765, + "grad_norm": 1.1466414457032206, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 17650 + }, + { + "epoch": 0.17651, + "grad_norm": 1.6318275207641635, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 17651 + }, + { + "epoch": 0.17652, + "grad_norm": 1.0730130344316913, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 17652 + }, + { + "epoch": 0.17653, + "grad_norm": 1.4448135906604729, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 17653 + }, + { + "epoch": 0.17654, + "grad_norm": 0.995886658610207, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 17654 + }, + { + "epoch": 0.17655, + "grad_norm": 1.2610663984889068, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 17655 + }, + { + "epoch": 0.17656, + "grad_norm": 1.1872689563076624, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 17656 + }, + { + "epoch": 0.17657, + "grad_norm": 1.3197271867164788, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 17657 + }, + { + "epoch": 0.17658, + "grad_norm": 1.1312799689004187, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 17658 + }, + { + "epoch": 0.17659, + "grad_norm": 1.4613108130626726, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 17659 + }, + { + "epoch": 0.1766, + "grad_norm": 1.215305872947806, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 17660 + }, + { + "epoch": 0.17661, + "grad_norm": 1.3428648707045394, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 17661 + }, + { + "epoch": 0.17662, + "grad_norm": 1.0548112155389986, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 17662 + }, + { + "epoch": 0.17663, + "grad_norm": 1.38585228216765, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 17663 + }, + { + "epoch": 0.17664, + "grad_norm": 1.1616798399752524, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 17664 + }, + { + "epoch": 0.17665, + "grad_norm": 1.5132186238889027, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 17665 + }, + { + "epoch": 0.17666, + "grad_norm": 1.2126955483808322, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 17666 + }, + { + "epoch": 0.17667, + "grad_norm": 1.1756236455801312, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 17667 + }, + { + "epoch": 0.17668, + "grad_norm": 1.1492162473406433, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 17668 + }, + { + "epoch": 0.17669, + "grad_norm": 1.1566973144214787, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 17669 + }, + { + "epoch": 0.1767, + "grad_norm": 1.506338186573076, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 17670 + }, + { + "epoch": 0.17671, + "grad_norm": 1.3894656365989166, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 17671 + }, + { + "epoch": 0.17672, + "grad_norm": 1.3466288883358835, + "learning_rate": 0.003, + "loss": 4.018, + "step": 17672 + }, + { + "epoch": 0.17673, + "grad_norm": 1.186810242645646, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 17673 + }, + { + "epoch": 0.17674, + "grad_norm": 1.275311566356402, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 17674 + }, + { + "epoch": 0.17675, + "grad_norm": 1.2940798901031692, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 17675 + }, + { + "epoch": 0.17676, + "grad_norm": 1.3305715960327655, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 17676 + }, + { + "epoch": 0.17677, + "grad_norm": 1.1855274683196175, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 17677 + }, + { + "epoch": 0.17678, + "grad_norm": 1.5154701258114815, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 17678 + }, + { + "epoch": 0.17679, + "grad_norm": 1.091549182112842, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 17679 + }, + { + "epoch": 0.1768, + "grad_norm": 1.2864130044198197, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 17680 + }, + { + "epoch": 0.17681, + "grad_norm": 1.16301276367685, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 17681 + }, + { + "epoch": 0.17682, + "grad_norm": 1.4227728662833272, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 17682 + }, + { + "epoch": 0.17683, + "grad_norm": 1.1344882161162675, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 17683 + }, + { + "epoch": 0.17684, + "grad_norm": 1.4008816348134472, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 17684 + }, + { + "epoch": 0.17685, + "grad_norm": 1.231251621514772, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 17685 + }, + { + "epoch": 0.17686, + "grad_norm": 1.2301141486401743, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 17686 + }, + { + "epoch": 0.17687, + "grad_norm": 1.1821536828911614, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 17687 + }, + { + "epoch": 0.17688, + "grad_norm": 1.6340806933485859, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 17688 + }, + { + "epoch": 0.17689, + "grad_norm": 1.291422498596477, + "learning_rate": 0.003, + "loss": 4.002, + "step": 17689 + }, + { + "epoch": 0.1769, + "grad_norm": 1.370574438743369, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 17690 + }, + { + "epoch": 0.17691, + "grad_norm": 1.2254311160033806, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 17691 + }, + { + "epoch": 0.17692, + "grad_norm": 1.2974026858514667, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 17692 + }, + { + "epoch": 0.17693, + "grad_norm": 1.272721143710752, + "learning_rate": 0.003, + "loss": 4.027, + "step": 17693 + }, + { + "epoch": 0.17694, + "grad_norm": 1.1896694512983272, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 17694 + }, + { + "epoch": 0.17695, + "grad_norm": 1.2918362518647468, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 17695 + }, + { + "epoch": 0.17696, + "grad_norm": 1.2716963200505105, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 17696 + }, + { + "epoch": 0.17697, + "grad_norm": 1.4537290270714847, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 17697 + }, + { + "epoch": 0.17698, + "grad_norm": 1.1057124277034496, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 17698 + }, + { + "epoch": 0.17699, + "grad_norm": 1.422077767087793, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 17699 + }, + { + "epoch": 0.177, + "grad_norm": 0.9764022016301912, + "learning_rate": 0.003, + "loss": 4.004, + "step": 17700 + }, + { + "epoch": 0.17701, + "grad_norm": 1.3305469891418082, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 17701 + }, + { + "epoch": 0.17702, + "grad_norm": 1.0302701942137433, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 17702 + }, + { + "epoch": 0.17703, + "grad_norm": 1.4584854604802886, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 17703 + }, + { + "epoch": 0.17704, + "grad_norm": 1.1897371995285562, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 17704 + }, + { + "epoch": 0.17705, + "grad_norm": 1.3886917458309411, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 17705 + }, + { + "epoch": 0.17706, + "grad_norm": 1.1999671477506535, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 17706 + }, + { + "epoch": 0.17707, + "grad_norm": 1.2901270427968135, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17707 + }, + { + "epoch": 0.17708, + "grad_norm": 1.2304546999945096, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 17708 + }, + { + "epoch": 0.17709, + "grad_norm": 1.2430163275046657, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 17709 + }, + { + "epoch": 0.1771, + "grad_norm": 1.2656647666393261, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 17710 + }, + { + "epoch": 0.17711, + "grad_norm": 1.3360788684996352, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17711 + }, + { + "epoch": 0.17712, + "grad_norm": 1.2041746388333983, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 17712 + }, + { + "epoch": 0.17713, + "grad_norm": 1.280418735699287, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 17713 + }, + { + "epoch": 0.17714, + "grad_norm": 1.075661513736821, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 17714 + }, + { + "epoch": 0.17715, + "grad_norm": 1.4788091330140924, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 17715 + }, + { + "epoch": 0.17716, + "grad_norm": 1.2800608738801205, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 17716 + }, + { + "epoch": 0.17717, + "grad_norm": 1.287082989265906, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 17717 + }, + { + "epoch": 0.17718, + "grad_norm": 1.1821351252449808, + "learning_rate": 0.003, + "loss": 4.037, + "step": 17718 + }, + { + "epoch": 0.17719, + "grad_norm": 1.4291892459728956, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 17719 + }, + { + "epoch": 0.1772, + "grad_norm": 1.0418877738975607, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 17720 + }, + { + "epoch": 0.17721, + "grad_norm": 1.5396247215856047, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 17721 + }, + { + "epoch": 0.17722, + "grad_norm": 1.2578079509816884, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 17722 + }, + { + "epoch": 0.17723, + "grad_norm": 1.5959123237774604, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 17723 + }, + { + "epoch": 0.17724, + "grad_norm": 1.2700442301751993, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 17724 + }, + { + "epoch": 0.17725, + "grad_norm": 1.2688808066870647, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 17725 + }, + { + "epoch": 0.17726, + "grad_norm": 1.411334765064551, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 17726 + }, + { + "epoch": 0.17727, + "grad_norm": 1.3186517008825351, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17727 + }, + { + "epoch": 0.17728, + "grad_norm": 1.209822516658508, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 17728 + }, + { + "epoch": 0.17729, + "grad_norm": 1.4961920570514124, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17729 + }, + { + "epoch": 0.1773, + "grad_norm": 1.27328898760038, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 17730 + }, + { + "epoch": 0.17731, + "grad_norm": 1.2122259626114444, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17731 + }, + { + "epoch": 0.17732, + "grad_norm": 1.2777247490895498, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 17732 + }, + { + "epoch": 0.17733, + "grad_norm": 1.1962870920111894, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 17733 + }, + { + "epoch": 0.17734, + "grad_norm": 1.3076328325875022, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 17734 + }, + { + "epoch": 0.17735, + "grad_norm": 1.2468178873207434, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 17735 + }, + { + "epoch": 0.17736, + "grad_norm": 1.4154689015219308, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 17736 + }, + { + "epoch": 0.17737, + "grad_norm": 1.1167695404121292, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 17737 + }, + { + "epoch": 0.17738, + "grad_norm": 1.4217078305436055, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 17738 + }, + { + "epoch": 0.17739, + "grad_norm": 1.133246619258175, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 17739 + }, + { + "epoch": 0.1774, + "grad_norm": 1.2347033258411095, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 17740 + }, + { + "epoch": 0.17741, + "grad_norm": 1.1348099403562892, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 17741 + }, + { + "epoch": 0.17742, + "grad_norm": 1.3785597373138432, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 17742 + }, + { + "epoch": 0.17743, + "grad_norm": 1.2690463645671801, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 17743 + }, + { + "epoch": 0.17744, + "grad_norm": 1.3338291965253706, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 17744 + }, + { + "epoch": 0.17745, + "grad_norm": 1.12328921255648, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 17745 + }, + { + "epoch": 0.17746, + "grad_norm": 1.5091144446382325, + "learning_rate": 0.003, + "loss": 4.029, + "step": 17746 + }, + { + "epoch": 0.17747, + "grad_norm": 1.1177715836119348, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 17747 + }, + { + "epoch": 0.17748, + "grad_norm": 1.3387225070034643, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 17748 + }, + { + "epoch": 0.17749, + "grad_norm": 1.233881653208535, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 17749 + }, + { + "epoch": 0.1775, + "grad_norm": 1.2206411609276537, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 17750 + }, + { + "epoch": 0.17751, + "grad_norm": 1.2803136770513195, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 17751 + }, + { + "epoch": 0.17752, + "grad_norm": 1.281486086260054, + "learning_rate": 0.003, + "loss": 4.002, + "step": 17752 + }, + { + "epoch": 0.17753, + "grad_norm": 1.1452833113701208, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 17753 + }, + { + "epoch": 0.17754, + "grad_norm": 1.3107700497422121, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 17754 + }, + { + "epoch": 0.17755, + "grad_norm": 1.2631770254982326, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 17755 + }, + { + "epoch": 0.17756, + "grad_norm": 1.4179454778251528, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 17756 + }, + { + "epoch": 0.17757, + "grad_norm": 1.209455850454635, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 17757 + }, + { + "epoch": 0.17758, + "grad_norm": 1.2430503248091411, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 17758 + }, + { + "epoch": 0.17759, + "grad_norm": 1.2258695046015262, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 17759 + }, + { + "epoch": 0.1776, + "grad_norm": 1.3720321833058833, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 17760 + }, + { + "epoch": 0.17761, + "grad_norm": 1.1157879586813826, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 17761 + }, + { + "epoch": 0.17762, + "grad_norm": 1.379835831092119, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 17762 + }, + { + "epoch": 0.17763, + "grad_norm": 1.074832416938527, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 17763 + }, + { + "epoch": 0.17764, + "grad_norm": 1.4779325456764498, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 17764 + }, + { + "epoch": 0.17765, + "grad_norm": 1.0737916991954493, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 17765 + }, + { + "epoch": 0.17766, + "grad_norm": 1.144441928094557, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 17766 + }, + { + "epoch": 0.17767, + "grad_norm": 1.3573965743376548, + "learning_rate": 0.003, + "loss": 4.026, + "step": 17767 + }, + { + "epoch": 0.17768, + "grad_norm": 1.4348911192679688, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 17768 + }, + { + "epoch": 0.17769, + "grad_norm": 1.288248592955297, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 17769 + }, + { + "epoch": 0.1777, + "grad_norm": 1.3009077629177082, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 17770 + }, + { + "epoch": 0.17771, + "grad_norm": 1.1001977329301142, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 17771 + }, + { + "epoch": 0.17772, + "grad_norm": 1.3781231544392432, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 17772 + }, + { + "epoch": 0.17773, + "grad_norm": 1.2070000918672394, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 17773 + }, + { + "epoch": 0.17774, + "grad_norm": 1.3584805334694108, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 17774 + }, + { + "epoch": 0.17775, + "grad_norm": 1.3257674589255668, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 17775 + }, + { + "epoch": 0.17776, + "grad_norm": 1.1440316911369965, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 17776 + }, + { + "epoch": 0.17777, + "grad_norm": 1.2951993538101918, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 17777 + }, + { + "epoch": 0.17778, + "grad_norm": 1.219670726338208, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 17778 + }, + { + "epoch": 0.17779, + "grad_norm": 1.2576854830100939, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 17779 + }, + { + "epoch": 0.1778, + "grad_norm": 1.2969554118521678, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 17780 + }, + { + "epoch": 0.17781, + "grad_norm": 1.395480473701788, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 17781 + }, + { + "epoch": 0.17782, + "grad_norm": 1.1210870129212929, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17782 + }, + { + "epoch": 0.17783, + "grad_norm": 1.3925910127657615, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 17783 + }, + { + "epoch": 0.17784, + "grad_norm": 1.1606521282473876, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 17784 + }, + { + "epoch": 0.17785, + "grad_norm": 1.3082696149032056, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 17785 + }, + { + "epoch": 0.17786, + "grad_norm": 1.055551621576099, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 17786 + }, + { + "epoch": 0.17787, + "grad_norm": 1.437729945871911, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 17787 + }, + { + "epoch": 0.17788, + "grad_norm": 1.0769162266838206, + "learning_rate": 0.003, + "loss": 4.018, + "step": 17788 + }, + { + "epoch": 0.17789, + "grad_norm": 1.4482118142629796, + "learning_rate": 0.003, + "loss": 4.005, + "step": 17789 + }, + { + "epoch": 0.1779, + "grad_norm": 1.3582494783626673, + "learning_rate": 0.003, + "loss": 3.988, + "step": 17790 + }, + { + "epoch": 0.17791, + "grad_norm": 1.2590550485047474, + "learning_rate": 0.003, + "loss": 4.006, + "step": 17791 + }, + { + "epoch": 0.17792, + "grad_norm": 1.5233881061352947, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 17792 + }, + { + "epoch": 0.17793, + "grad_norm": 1.2476714509933917, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 17793 + }, + { + "epoch": 0.17794, + "grad_norm": 1.1457363146513437, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 17794 + }, + { + "epoch": 0.17795, + "grad_norm": 1.285982846103947, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 17795 + }, + { + "epoch": 0.17796, + "grad_norm": 0.9674368897959768, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 17796 + }, + { + "epoch": 0.17797, + "grad_norm": 1.3847835487943505, + "learning_rate": 0.003, + "loss": 4.027, + "step": 17797 + }, + { + "epoch": 0.17798, + "grad_norm": 1.388823923849772, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 17798 + }, + { + "epoch": 0.17799, + "grad_norm": 1.1850443723939679, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 17799 + }, + { + "epoch": 0.178, + "grad_norm": 1.239507130863008, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 17800 + }, + { + "epoch": 0.17801, + "grad_norm": 1.1649600820146366, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 17801 + }, + { + "epoch": 0.17802, + "grad_norm": 1.1564667995275404, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 17802 + }, + { + "epoch": 0.17803, + "grad_norm": 1.2422393675117376, + "learning_rate": 0.003, + "loss": 4.021, + "step": 17803 + }, + { + "epoch": 0.17804, + "grad_norm": 1.270593768355618, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 17804 + }, + { + "epoch": 0.17805, + "grad_norm": 1.3871869687322365, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 17805 + }, + { + "epoch": 0.17806, + "grad_norm": 1.2303533481624622, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 17806 + }, + { + "epoch": 0.17807, + "grad_norm": 1.3022480992793621, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 17807 + }, + { + "epoch": 0.17808, + "grad_norm": 1.0142472820069428, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 17808 + }, + { + "epoch": 0.17809, + "grad_norm": 1.2972177297703724, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 17809 + }, + { + "epoch": 0.1781, + "grad_norm": 1.0214519579537469, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 17810 + }, + { + "epoch": 0.17811, + "grad_norm": 1.4056262049180404, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 17811 + }, + { + "epoch": 0.17812, + "grad_norm": 0.9613461804577055, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 17812 + }, + { + "epoch": 0.17813, + "grad_norm": 1.3987265467581167, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 17813 + }, + { + "epoch": 0.17814, + "grad_norm": 1.1919738807894389, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 17814 + }, + { + "epoch": 0.17815, + "grad_norm": 1.3973972286087633, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 17815 + }, + { + "epoch": 0.17816, + "grad_norm": 1.0827310417582043, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 17816 + }, + { + "epoch": 0.17817, + "grad_norm": 1.462587451628352, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 17817 + }, + { + "epoch": 0.17818, + "grad_norm": 1.3370237425179567, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17818 + }, + { + "epoch": 0.17819, + "grad_norm": 1.3710553095240827, + "learning_rate": 0.003, + "loss": 4.032, + "step": 17819 + }, + { + "epoch": 0.1782, + "grad_norm": 1.3646700731882089, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 17820 + }, + { + "epoch": 0.17821, + "grad_norm": 1.2610373738435834, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 17821 + }, + { + "epoch": 0.17822, + "grad_norm": 1.2254092006565505, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 17822 + }, + { + "epoch": 0.17823, + "grad_norm": 1.4283800539534146, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17823 + }, + { + "epoch": 0.17824, + "grad_norm": 1.1670433064995493, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 17824 + }, + { + "epoch": 0.17825, + "grad_norm": 1.5167128526422318, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 17825 + }, + { + "epoch": 0.17826, + "grad_norm": 1.2188122274167377, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 17826 + }, + { + "epoch": 0.17827, + "grad_norm": 1.300553278058531, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 17827 + }, + { + "epoch": 0.17828, + "grad_norm": 1.0737184150438381, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 17828 + }, + { + "epoch": 0.17829, + "grad_norm": 1.281079446383483, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17829 + }, + { + "epoch": 0.1783, + "grad_norm": 1.2856173908121271, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 17830 + }, + { + "epoch": 0.17831, + "grad_norm": 1.1935793636552723, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 17831 + }, + { + "epoch": 0.17832, + "grad_norm": 1.205178492606702, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 17832 + }, + { + "epoch": 0.17833, + "grad_norm": 1.1944403357477134, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 17833 + }, + { + "epoch": 0.17834, + "grad_norm": 1.3908171744977453, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 17834 + }, + { + "epoch": 0.17835, + "grad_norm": 1.1405338280337045, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 17835 + }, + { + "epoch": 0.17836, + "grad_norm": 1.6468845041136069, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 17836 + }, + { + "epoch": 0.17837, + "grad_norm": 1.0256403167460948, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 17837 + }, + { + "epoch": 0.17838, + "grad_norm": 1.7529827411633745, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 17838 + }, + { + "epoch": 0.17839, + "grad_norm": 1.0343519324796442, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 17839 + }, + { + "epoch": 0.1784, + "grad_norm": 1.3088197200383436, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 17840 + }, + { + "epoch": 0.17841, + "grad_norm": 1.391166098418807, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 17841 + }, + { + "epoch": 0.17842, + "grad_norm": 1.0313034724517778, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 17842 + }, + { + "epoch": 0.17843, + "grad_norm": 1.5906309328849493, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 17843 + }, + { + "epoch": 0.17844, + "grad_norm": 1.2887138037584314, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 17844 + }, + { + "epoch": 0.17845, + "grad_norm": 1.3157215102744646, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 17845 + }, + { + "epoch": 0.17846, + "grad_norm": 1.25794672180459, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 17846 + }, + { + "epoch": 0.17847, + "grad_norm": 1.2592578115929949, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 17847 + }, + { + "epoch": 0.17848, + "grad_norm": 1.2275976027820696, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17848 + }, + { + "epoch": 0.17849, + "grad_norm": 1.414525973723539, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 17849 + }, + { + "epoch": 0.1785, + "grad_norm": 1.3275192594843719, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 17850 + }, + { + "epoch": 0.17851, + "grad_norm": 1.1369183603220658, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 17851 + }, + { + "epoch": 0.17852, + "grad_norm": 1.223063913334217, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 17852 + }, + { + "epoch": 0.17853, + "grad_norm": 1.156306213694106, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 17853 + }, + { + "epoch": 0.17854, + "grad_norm": 1.230513390835216, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 17854 + }, + { + "epoch": 0.17855, + "grad_norm": 1.3467261543697027, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 17855 + }, + { + "epoch": 0.17856, + "grad_norm": 1.4479018385666143, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 17856 + }, + { + "epoch": 0.17857, + "grad_norm": 1.096494157687689, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 17857 + }, + { + "epoch": 0.17858, + "grad_norm": 1.37154842305199, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 17858 + }, + { + "epoch": 0.17859, + "grad_norm": 1.1382470176153237, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 17859 + }, + { + "epoch": 0.1786, + "grad_norm": 1.4260573661969602, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 17860 + }, + { + "epoch": 0.17861, + "grad_norm": 1.2749053697764088, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 17861 + }, + { + "epoch": 0.17862, + "grad_norm": 1.2813180513464035, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 17862 + }, + { + "epoch": 0.17863, + "grad_norm": 1.1378759447463387, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 17863 + }, + { + "epoch": 0.17864, + "grad_norm": 1.1581966288327354, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 17864 + }, + { + "epoch": 0.17865, + "grad_norm": 1.2251666216073551, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 17865 + }, + { + "epoch": 0.17866, + "grad_norm": 1.5888960876025273, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 17866 + }, + { + "epoch": 0.17867, + "grad_norm": 1.2610834406615368, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 17867 + }, + { + "epoch": 0.17868, + "grad_norm": 1.2981799021894602, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 17868 + }, + { + "epoch": 0.17869, + "grad_norm": 1.22439715024633, + "learning_rate": 0.003, + "loss": 4.021, + "step": 17869 + }, + { + "epoch": 0.1787, + "grad_norm": 1.285613960652768, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 17870 + }, + { + "epoch": 0.17871, + "grad_norm": 1.39587116721041, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 17871 + }, + { + "epoch": 0.17872, + "grad_norm": 1.0331956935241802, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 17872 + }, + { + "epoch": 0.17873, + "grad_norm": 1.5222012103196498, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 17873 + }, + { + "epoch": 0.17874, + "grad_norm": 1.109330647963323, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 17874 + }, + { + "epoch": 0.17875, + "grad_norm": 1.3513786080210313, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 17875 + }, + { + "epoch": 0.17876, + "grad_norm": 1.1377725636560931, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 17876 + }, + { + "epoch": 0.17877, + "grad_norm": 1.1081239090394368, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 17877 + }, + { + "epoch": 0.17878, + "grad_norm": 1.3504881207612687, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 17878 + }, + { + "epoch": 0.17879, + "grad_norm": 1.2202818156011814, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 17879 + }, + { + "epoch": 0.1788, + "grad_norm": 1.4659793559370045, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 17880 + }, + { + "epoch": 0.17881, + "grad_norm": 0.9991861962378518, + "learning_rate": 0.003, + "loss": 4.015, + "step": 17881 + }, + { + "epoch": 0.17882, + "grad_norm": 1.2840130566269725, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 17882 + }, + { + "epoch": 0.17883, + "grad_norm": 1.223462630513146, + "learning_rate": 0.003, + "loss": 4.008, + "step": 17883 + }, + { + "epoch": 0.17884, + "grad_norm": 1.1745549627144207, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 17884 + }, + { + "epoch": 0.17885, + "grad_norm": 1.3288648291234721, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 17885 + }, + { + "epoch": 0.17886, + "grad_norm": 1.2275248459800174, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 17886 + }, + { + "epoch": 0.17887, + "grad_norm": 1.2840299146550205, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 17887 + }, + { + "epoch": 0.17888, + "grad_norm": 1.1109153791822728, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 17888 + }, + { + "epoch": 0.17889, + "grad_norm": 1.4154726301866394, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 17889 + }, + { + "epoch": 0.1789, + "grad_norm": 1.3765358041628772, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 17890 + }, + { + "epoch": 0.17891, + "grad_norm": 1.356781113514337, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 17891 + }, + { + "epoch": 0.17892, + "grad_norm": 1.2083142687013617, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 17892 + }, + { + "epoch": 0.17893, + "grad_norm": 1.3886682337620795, + "learning_rate": 0.003, + "loss": 4.004, + "step": 17893 + }, + { + "epoch": 0.17894, + "grad_norm": 1.087255529710916, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 17894 + }, + { + "epoch": 0.17895, + "grad_norm": 1.2379106089715426, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 17895 + }, + { + "epoch": 0.17896, + "grad_norm": 1.329775546400975, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 17896 + }, + { + "epoch": 0.17897, + "grad_norm": 1.2250169685520114, + "learning_rate": 0.003, + "loss": 4.031, + "step": 17897 + }, + { + "epoch": 0.17898, + "grad_norm": 1.2887533824006294, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 17898 + }, + { + "epoch": 0.17899, + "grad_norm": 1.1142232587310008, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 17899 + }, + { + "epoch": 0.179, + "grad_norm": 1.5125094557705192, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 17900 + }, + { + "epoch": 0.17901, + "grad_norm": 0.8959437293219716, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 17901 + }, + { + "epoch": 0.17902, + "grad_norm": 1.481590620348304, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 17902 + }, + { + "epoch": 0.17903, + "grad_norm": 1.4513535134662827, + "learning_rate": 0.003, + "loss": 4.018, + "step": 17903 + }, + { + "epoch": 0.17904, + "grad_norm": 1.2684392163155676, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 17904 + }, + { + "epoch": 0.17905, + "grad_norm": 1.3435859991089862, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 17905 + }, + { + "epoch": 0.17906, + "grad_norm": 1.14005741793786, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 17906 + }, + { + "epoch": 0.17907, + "grad_norm": 1.2647231228841296, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 17907 + }, + { + "epoch": 0.17908, + "grad_norm": 1.0303130185661538, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 17908 + }, + { + "epoch": 0.17909, + "grad_norm": 1.3362370493898243, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 17909 + }, + { + "epoch": 0.1791, + "grad_norm": 1.3376456562964527, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 17910 + }, + { + "epoch": 0.17911, + "grad_norm": 1.2894298745129902, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 17911 + }, + { + "epoch": 0.17912, + "grad_norm": 1.1938509311080707, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 17912 + }, + { + "epoch": 0.17913, + "grad_norm": 1.2953870550115025, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 17913 + }, + { + "epoch": 0.17914, + "grad_norm": 1.3673196371585163, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 17914 + }, + { + "epoch": 0.17915, + "grad_norm": 1.2180323020321726, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 17915 + }, + { + "epoch": 0.17916, + "grad_norm": 1.335540244149917, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 17916 + }, + { + "epoch": 0.17917, + "grad_norm": 1.1782735032392786, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 17917 + }, + { + "epoch": 0.17918, + "grad_norm": 1.361398393966456, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 17918 + }, + { + "epoch": 0.17919, + "grad_norm": 1.176221574671349, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 17919 + }, + { + "epoch": 0.1792, + "grad_norm": 1.397010968398075, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 17920 + }, + { + "epoch": 0.17921, + "grad_norm": 1.156606009121643, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 17921 + }, + { + "epoch": 0.17922, + "grad_norm": 1.32125183249408, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 17922 + }, + { + "epoch": 0.17923, + "grad_norm": 1.1974488910447372, + "learning_rate": 0.003, + "loss": 4.005, + "step": 17923 + }, + { + "epoch": 0.17924, + "grad_norm": 1.4585490204785243, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 17924 + }, + { + "epoch": 0.17925, + "grad_norm": 1.1094529111745237, + "learning_rate": 0.003, + "loss": 3.992, + "step": 17925 + }, + { + "epoch": 0.17926, + "grad_norm": 1.309982676010241, + "learning_rate": 0.003, + "loss": 3.99, + "step": 17926 + }, + { + "epoch": 0.17927, + "grad_norm": 1.1844851359692525, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 17927 + }, + { + "epoch": 0.17928, + "grad_norm": 1.4815185736791534, + "learning_rate": 0.003, + "loss": 4.02, + "step": 17928 + }, + { + "epoch": 0.17929, + "grad_norm": 1.1605835647717957, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 17929 + }, + { + "epoch": 0.1793, + "grad_norm": 1.320507640873421, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 17930 + }, + { + "epoch": 0.17931, + "grad_norm": 1.274046317213217, + "learning_rate": 0.003, + "loss": 4.004, + "step": 17931 + }, + { + "epoch": 0.17932, + "grad_norm": 1.2972089424586997, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 17932 + }, + { + "epoch": 0.17933, + "grad_norm": 1.2441857571769512, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 17933 + }, + { + "epoch": 0.17934, + "grad_norm": 1.2009046967225345, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 17934 + }, + { + "epoch": 0.17935, + "grad_norm": 1.068183752609735, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 17935 + }, + { + "epoch": 0.17936, + "grad_norm": 1.3524304151950766, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 17936 + }, + { + "epoch": 0.17937, + "grad_norm": 1.2006139512947491, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 17937 + }, + { + "epoch": 0.17938, + "grad_norm": 1.5279550747071753, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 17938 + }, + { + "epoch": 0.17939, + "grad_norm": 1.0707218523923578, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 17939 + }, + { + "epoch": 0.1794, + "grad_norm": 1.470373557818703, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 17940 + }, + { + "epoch": 0.17941, + "grad_norm": 1.283217845435159, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 17941 + }, + { + "epoch": 0.17942, + "grad_norm": 1.2524743350909433, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 17942 + }, + { + "epoch": 0.17943, + "grad_norm": 1.3631818977429888, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 17943 + }, + { + "epoch": 0.17944, + "grad_norm": 1.198417068713717, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 17944 + }, + { + "epoch": 0.17945, + "grad_norm": 1.3364558732265843, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 17945 + }, + { + "epoch": 0.17946, + "grad_norm": 1.1993189564378368, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 17946 + }, + { + "epoch": 0.17947, + "grad_norm": 1.2724891851218023, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 17947 + }, + { + "epoch": 0.17948, + "grad_norm": 1.329903192112161, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 17948 + }, + { + "epoch": 0.17949, + "grad_norm": 1.2493252737844187, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 17949 + }, + { + "epoch": 0.1795, + "grad_norm": 1.4258413169140705, + "learning_rate": 0.003, + "loss": 4.031, + "step": 17950 + }, + { + "epoch": 0.17951, + "grad_norm": 1.0947045287903696, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 17951 + }, + { + "epoch": 0.17952, + "grad_norm": 1.2386797285364355, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 17952 + }, + { + "epoch": 0.17953, + "grad_norm": 1.14594781761, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 17953 + }, + { + "epoch": 0.17954, + "grad_norm": 1.297250203912752, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 17954 + }, + { + "epoch": 0.17955, + "grad_norm": 1.4061598745408204, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 17955 + }, + { + "epoch": 0.17956, + "grad_norm": 1.1411124734017042, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 17956 + }, + { + "epoch": 0.17957, + "grad_norm": 1.4739424177374514, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 17957 + }, + { + "epoch": 0.17958, + "grad_norm": 1.1333136046265506, + "learning_rate": 0.003, + "loss": 4.003, + "step": 17958 + }, + { + "epoch": 0.17959, + "grad_norm": 1.3103038254148236, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 17959 + }, + { + "epoch": 0.1796, + "grad_norm": 1.119448017728659, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 17960 + }, + { + "epoch": 0.17961, + "grad_norm": 1.2547630506354666, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 17961 + }, + { + "epoch": 0.17962, + "grad_norm": 1.2326847898269968, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 17962 + }, + { + "epoch": 0.17963, + "grad_norm": 1.3243312318037208, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 17963 + }, + { + "epoch": 0.17964, + "grad_norm": 1.2676590854248495, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 17964 + }, + { + "epoch": 0.17965, + "grad_norm": 1.144462546866322, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 17965 + }, + { + "epoch": 0.17966, + "grad_norm": 1.316130726919737, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 17966 + }, + { + "epoch": 0.17967, + "grad_norm": 1.2412240478500736, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 17967 + }, + { + "epoch": 0.17968, + "grad_norm": 1.2352631669776275, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 17968 + }, + { + "epoch": 0.17969, + "grad_norm": 1.1924883625361797, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 17969 + }, + { + "epoch": 0.1797, + "grad_norm": 1.1305434477854719, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 17970 + }, + { + "epoch": 0.17971, + "grad_norm": 1.511192269977118, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 17971 + }, + { + "epoch": 0.17972, + "grad_norm": 1.2017320818124952, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 17972 + }, + { + "epoch": 0.17973, + "grad_norm": 1.4131206724996497, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 17973 + }, + { + "epoch": 0.17974, + "grad_norm": 1.1541400742377594, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 17974 + }, + { + "epoch": 0.17975, + "grad_norm": 1.5212433825381124, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 17975 + }, + { + "epoch": 0.17976, + "grad_norm": 1.0915852560086072, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 17976 + }, + { + "epoch": 0.17977, + "grad_norm": 1.4541941510471958, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17977 + }, + { + "epoch": 0.17978, + "grad_norm": 1.1637687489354096, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17978 + }, + { + "epoch": 0.17979, + "grad_norm": 1.6518764770465473, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 17979 + }, + { + "epoch": 0.1798, + "grad_norm": 1.226657411711691, + "learning_rate": 0.003, + "loss": 4.0, + "step": 17980 + }, + { + "epoch": 0.17981, + "grad_norm": 1.4672932830010585, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 17981 + }, + { + "epoch": 0.17982, + "grad_norm": 1.0789539113969855, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 17982 + }, + { + "epoch": 0.17983, + "grad_norm": 1.3077465429252888, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 17983 + }, + { + "epoch": 0.17984, + "grad_norm": 1.3217963244269073, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 17984 + }, + { + "epoch": 0.17985, + "grad_norm": 1.4177608416518557, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 17985 + }, + { + "epoch": 0.17986, + "grad_norm": 1.2630185518715435, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 17986 + }, + { + "epoch": 0.17987, + "grad_norm": 1.1224916779924532, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 17987 + }, + { + "epoch": 0.17988, + "grad_norm": 1.223872692574195, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 17988 + }, + { + "epoch": 0.17989, + "grad_norm": 1.1846536337950013, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 17989 + }, + { + "epoch": 0.1799, + "grad_norm": 1.3396779385326112, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 17990 + }, + { + "epoch": 0.17991, + "grad_norm": 1.0887540493867556, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17991 + }, + { + "epoch": 0.17992, + "grad_norm": 1.2346943418605258, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 17992 + }, + { + "epoch": 0.17993, + "grad_norm": 1.4320850253519977, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17993 + }, + { + "epoch": 0.17994, + "grad_norm": 1.218821459511451, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 17994 + }, + { + "epoch": 0.17995, + "grad_norm": 1.3859837355781857, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 17995 + }, + { + "epoch": 0.17996, + "grad_norm": 1.1344698861221325, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 17996 + }, + { + "epoch": 0.17997, + "grad_norm": 1.3849875752526728, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 17997 + }, + { + "epoch": 0.17998, + "grad_norm": 1.251398295605249, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 17998 + }, + { + "epoch": 0.17999, + "grad_norm": 1.2819503563394226, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 17999 + }, + { + "epoch": 0.18, + "grad_norm": 1.3741654746676373, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 18000 + }, + { + "epoch": 0.18001, + "grad_norm": 1.1933635475479498, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 18001 + }, + { + "epoch": 0.18002, + "grad_norm": 1.3098748972850278, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 18002 + }, + { + "epoch": 0.18003, + "grad_norm": 1.212517051786349, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 18003 + }, + { + "epoch": 0.18004, + "grad_norm": 1.392106308722888, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18004 + }, + { + "epoch": 0.18005, + "grad_norm": 1.424687540303722, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 18005 + }, + { + "epoch": 0.18006, + "grad_norm": 1.246009503653905, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 18006 + }, + { + "epoch": 0.18007, + "grad_norm": 1.099120441540404, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 18007 + }, + { + "epoch": 0.18008, + "grad_norm": 1.4627704259168404, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 18008 + }, + { + "epoch": 0.18009, + "grad_norm": 1.224397106357738, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 18009 + }, + { + "epoch": 0.1801, + "grad_norm": 1.472982585937009, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 18010 + }, + { + "epoch": 0.18011, + "grad_norm": 1.3218278772814305, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 18011 + }, + { + "epoch": 0.18012, + "grad_norm": 1.2015626900018146, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 18012 + }, + { + "epoch": 0.18013, + "grad_norm": 1.3141744379305316, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 18013 + }, + { + "epoch": 0.18014, + "grad_norm": 1.305760597878548, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 18014 + }, + { + "epoch": 0.18015, + "grad_norm": 1.2904157889554446, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 18015 + }, + { + "epoch": 0.18016, + "grad_norm": 1.0936669246650228, + "learning_rate": 0.003, + "loss": 4.016, + "step": 18016 + }, + { + "epoch": 0.18017, + "grad_norm": 1.4724170169853559, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 18017 + }, + { + "epoch": 0.18018, + "grad_norm": 1.1737041272683193, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 18018 + }, + { + "epoch": 0.18019, + "grad_norm": 1.29288969785492, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 18019 + }, + { + "epoch": 0.1802, + "grad_norm": 1.0685343810456742, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18020 + }, + { + "epoch": 0.18021, + "grad_norm": 1.3399528333597222, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 18021 + }, + { + "epoch": 0.18022, + "grad_norm": 1.2301345998275353, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 18022 + }, + { + "epoch": 0.18023, + "grad_norm": 1.4377926684604674, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 18023 + }, + { + "epoch": 0.18024, + "grad_norm": 1.4485503664499548, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 18024 + }, + { + "epoch": 0.18025, + "grad_norm": 1.108201695546443, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 18025 + }, + { + "epoch": 0.18026, + "grad_norm": 1.2249896527161086, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 18026 + }, + { + "epoch": 0.18027, + "grad_norm": 1.3631836057201774, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 18027 + }, + { + "epoch": 0.18028, + "grad_norm": 1.2128407493827715, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 18028 + }, + { + "epoch": 0.18029, + "grad_norm": 1.4377792927247532, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 18029 + }, + { + "epoch": 0.1803, + "grad_norm": 1.1196044734687098, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 18030 + }, + { + "epoch": 0.18031, + "grad_norm": 1.3337921773230739, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 18031 + }, + { + "epoch": 0.18032, + "grad_norm": 1.2361892568593928, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 18032 + }, + { + "epoch": 0.18033, + "grad_norm": 1.2633246942423886, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 18033 + }, + { + "epoch": 0.18034, + "grad_norm": 1.284347096442261, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 18034 + }, + { + "epoch": 0.18035, + "grad_norm": 1.2429609759207414, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 18035 + }, + { + "epoch": 0.18036, + "grad_norm": 1.3922648571191765, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 18036 + }, + { + "epoch": 0.18037, + "grad_norm": 1.1520110539034873, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18037 + }, + { + "epoch": 0.18038, + "grad_norm": 1.4640431772660716, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18038 + }, + { + "epoch": 0.18039, + "grad_norm": 1.194106552299273, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 18039 + }, + { + "epoch": 0.1804, + "grad_norm": 1.4627313146103296, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 18040 + }, + { + "epoch": 0.18041, + "grad_norm": 1.1027984597618095, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 18041 + }, + { + "epoch": 0.18042, + "grad_norm": 1.3924257301363576, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 18042 + }, + { + "epoch": 0.18043, + "grad_norm": 1.1401766419096124, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 18043 + }, + { + "epoch": 0.18044, + "grad_norm": 1.2731268679660979, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 18044 + }, + { + "epoch": 0.18045, + "grad_norm": 1.2319214886258891, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 18045 + }, + { + "epoch": 0.18046, + "grad_norm": 1.1969308560875676, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 18046 + }, + { + "epoch": 0.18047, + "grad_norm": 1.1623865443335577, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 18047 + }, + { + "epoch": 0.18048, + "grad_norm": 1.347600280707264, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18048 + }, + { + "epoch": 0.18049, + "grad_norm": 1.3095982082710382, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 18049 + }, + { + "epoch": 0.1805, + "grad_norm": 1.4009570741734567, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 18050 + }, + { + "epoch": 0.18051, + "grad_norm": 1.350897678706278, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 18051 + }, + { + "epoch": 0.18052, + "grad_norm": 1.1555449767293169, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 18052 + }, + { + "epoch": 0.18053, + "grad_norm": 1.201864879613062, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 18053 + }, + { + "epoch": 0.18054, + "grad_norm": 1.3632552345487847, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 18054 + }, + { + "epoch": 0.18055, + "grad_norm": 1.3748051053858183, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 18055 + }, + { + "epoch": 0.18056, + "grad_norm": 1.368423119049279, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 18056 + }, + { + "epoch": 0.18057, + "grad_norm": 1.4737393844383795, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 18057 + }, + { + "epoch": 0.18058, + "grad_norm": 1.180383393312585, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 18058 + }, + { + "epoch": 0.18059, + "grad_norm": 1.220989213819937, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 18059 + }, + { + "epoch": 0.1806, + "grad_norm": 1.3790859140402394, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 18060 + }, + { + "epoch": 0.18061, + "grad_norm": 1.3230364801202141, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 18061 + }, + { + "epoch": 0.18062, + "grad_norm": 1.2205140368731702, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 18062 + }, + { + "epoch": 0.18063, + "grad_norm": 1.1794690741742826, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 18063 + }, + { + "epoch": 0.18064, + "grad_norm": 1.3874867718840194, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 18064 + }, + { + "epoch": 0.18065, + "grad_norm": 1.2158102941757782, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 18065 + }, + { + "epoch": 0.18066, + "grad_norm": 1.3971232563203826, + "learning_rate": 0.003, + "loss": 3.997, + "step": 18066 + }, + { + "epoch": 0.18067, + "grad_norm": 1.1537258202681013, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 18067 + }, + { + "epoch": 0.18068, + "grad_norm": 1.317011717519727, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18068 + }, + { + "epoch": 0.18069, + "grad_norm": 1.4116033534763732, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 18069 + }, + { + "epoch": 0.1807, + "grad_norm": 1.2352646362071453, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 18070 + }, + { + "epoch": 0.18071, + "grad_norm": 1.0614122175823477, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 18071 + }, + { + "epoch": 0.18072, + "grad_norm": 1.4037585678071596, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 18072 + }, + { + "epoch": 0.18073, + "grad_norm": 1.2619816077975499, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 18073 + }, + { + "epoch": 0.18074, + "grad_norm": 1.279158743537198, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 18074 + }, + { + "epoch": 0.18075, + "grad_norm": 1.0144218595931895, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 18075 + }, + { + "epoch": 0.18076, + "grad_norm": 1.3932923689905843, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 18076 + }, + { + "epoch": 0.18077, + "grad_norm": 1.2381797948281783, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 18077 + }, + { + "epoch": 0.18078, + "grad_norm": 1.2064732572389434, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 18078 + }, + { + "epoch": 0.18079, + "grad_norm": 1.2309520334545374, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18079 + }, + { + "epoch": 0.1808, + "grad_norm": 1.4472036453649932, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 18080 + }, + { + "epoch": 0.18081, + "grad_norm": 1.0872125792142273, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 18081 + }, + { + "epoch": 0.18082, + "grad_norm": 1.7513922460165103, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 18082 + }, + { + "epoch": 0.18083, + "grad_norm": 0.901507635320083, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 18083 + }, + { + "epoch": 0.18084, + "grad_norm": 1.1722118986780032, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 18084 + }, + { + "epoch": 0.18085, + "grad_norm": 1.3065034468877128, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 18085 + }, + { + "epoch": 0.18086, + "grad_norm": 1.3000990175300853, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 18086 + }, + { + "epoch": 0.18087, + "grad_norm": 1.176289481273397, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18087 + }, + { + "epoch": 0.18088, + "grad_norm": 1.474747785249665, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 18088 + }, + { + "epoch": 0.18089, + "grad_norm": 1.15869905445471, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 18089 + }, + { + "epoch": 0.1809, + "grad_norm": 1.531752376162567, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 18090 + }, + { + "epoch": 0.18091, + "grad_norm": 1.3081067249732787, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 18091 + }, + { + "epoch": 0.18092, + "grad_norm": 1.1556575754200102, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 18092 + }, + { + "epoch": 0.18093, + "grad_norm": 1.311386628742779, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 18093 + }, + { + "epoch": 0.18094, + "grad_norm": 1.2743429778379944, + "learning_rate": 0.003, + "loss": 3.997, + "step": 18094 + }, + { + "epoch": 0.18095, + "grad_norm": 1.1268201413256642, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 18095 + }, + { + "epoch": 0.18096, + "grad_norm": 1.533802254578135, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 18096 + }, + { + "epoch": 0.18097, + "grad_norm": 1.143733177519895, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 18097 + }, + { + "epoch": 0.18098, + "grad_norm": 1.6237064413171671, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 18098 + }, + { + "epoch": 0.18099, + "grad_norm": 1.2066381421671812, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18099 + }, + { + "epoch": 0.181, + "grad_norm": 1.3431374559869123, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 18100 + }, + { + "epoch": 0.18101, + "grad_norm": 1.202812996875956, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 18101 + }, + { + "epoch": 0.18102, + "grad_norm": 1.1698554573987243, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 18102 + }, + { + "epoch": 0.18103, + "grad_norm": 1.2428386400393254, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 18103 + }, + { + "epoch": 0.18104, + "grad_norm": 1.1208181772818042, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18104 + }, + { + "epoch": 0.18105, + "grad_norm": 1.4972398021134263, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 18105 + }, + { + "epoch": 0.18106, + "grad_norm": 1.115878346566044, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 18106 + }, + { + "epoch": 0.18107, + "grad_norm": 1.4277133733715102, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18107 + }, + { + "epoch": 0.18108, + "grad_norm": 1.4102164927486251, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 18108 + }, + { + "epoch": 0.18109, + "grad_norm": 1.1152660397268048, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 18109 + }, + { + "epoch": 0.1811, + "grad_norm": 1.2872868106598139, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 18110 + }, + { + "epoch": 0.18111, + "grad_norm": 1.394418514138197, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 18111 + }, + { + "epoch": 0.18112, + "grad_norm": 1.1602986408748321, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 18112 + }, + { + "epoch": 0.18113, + "grad_norm": 1.405413427856654, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 18113 + }, + { + "epoch": 0.18114, + "grad_norm": 1.2248450601215533, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 18114 + }, + { + "epoch": 0.18115, + "grad_norm": 1.3294455281217434, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 18115 + }, + { + "epoch": 0.18116, + "grad_norm": 1.4136512393647949, + "learning_rate": 0.003, + "loss": 4.053, + "step": 18116 + }, + { + "epoch": 0.18117, + "grad_norm": 1.115576619747742, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 18117 + }, + { + "epoch": 0.18118, + "grad_norm": 1.4550200396045296, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 18118 + }, + { + "epoch": 0.18119, + "grad_norm": 1.178333624432399, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 18119 + }, + { + "epoch": 0.1812, + "grad_norm": 1.53289204822824, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 18120 + }, + { + "epoch": 0.18121, + "grad_norm": 1.165712719335234, + "learning_rate": 0.003, + "loss": 3.99, + "step": 18121 + }, + { + "epoch": 0.18122, + "grad_norm": 1.1536296849685932, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 18122 + }, + { + "epoch": 0.18123, + "grad_norm": 1.3061571896655713, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 18123 + }, + { + "epoch": 0.18124, + "grad_norm": 1.2293773622678938, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 18124 + }, + { + "epoch": 0.18125, + "grad_norm": 1.2751405136705993, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18125 + }, + { + "epoch": 0.18126, + "grad_norm": 1.108491319706021, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 18126 + }, + { + "epoch": 0.18127, + "grad_norm": 1.3837138363424657, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 18127 + }, + { + "epoch": 0.18128, + "grad_norm": 0.9960471378550317, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 18128 + }, + { + "epoch": 0.18129, + "grad_norm": 1.5916971542444212, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 18129 + }, + { + "epoch": 0.1813, + "grad_norm": 1.058993336455924, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 18130 + }, + { + "epoch": 0.18131, + "grad_norm": 1.5209413438626223, + "learning_rate": 0.003, + "loss": 4.022, + "step": 18131 + }, + { + "epoch": 0.18132, + "grad_norm": 1.1303170062062597, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 18132 + }, + { + "epoch": 0.18133, + "grad_norm": 1.5315039018604342, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 18133 + }, + { + "epoch": 0.18134, + "grad_norm": 1.1804779996899288, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 18134 + }, + { + "epoch": 0.18135, + "grad_norm": 1.3623009834521842, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 18135 + }, + { + "epoch": 0.18136, + "grad_norm": 1.184765040963404, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 18136 + }, + { + "epoch": 0.18137, + "grad_norm": 1.248538813627692, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 18137 + }, + { + "epoch": 0.18138, + "grad_norm": 1.1882430006443279, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 18138 + }, + { + "epoch": 0.18139, + "grad_norm": 1.2704498041616852, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 18139 + }, + { + "epoch": 0.1814, + "grad_norm": 1.3659019087493167, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 18140 + }, + { + "epoch": 0.18141, + "grad_norm": 1.2346528099879523, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 18141 + }, + { + "epoch": 0.18142, + "grad_norm": 1.3299592955694377, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 18142 + }, + { + "epoch": 0.18143, + "grad_norm": 1.2893952830730906, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18143 + }, + { + "epoch": 0.18144, + "grad_norm": 1.2444686259253939, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 18144 + }, + { + "epoch": 0.18145, + "grad_norm": 1.1147178423921968, + "learning_rate": 0.003, + "loss": 4.006, + "step": 18145 + }, + { + "epoch": 0.18146, + "grad_norm": 1.3555438914152864, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 18146 + }, + { + "epoch": 0.18147, + "grad_norm": 1.1190174607528691, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 18147 + }, + { + "epoch": 0.18148, + "grad_norm": 1.4753459405725098, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 18148 + }, + { + "epoch": 0.18149, + "grad_norm": 1.2181917350244813, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 18149 + }, + { + "epoch": 0.1815, + "grad_norm": 1.3655624174024366, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 18150 + }, + { + "epoch": 0.18151, + "grad_norm": 1.2763174805966657, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 18151 + }, + { + "epoch": 0.18152, + "grad_norm": 1.444889682698498, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 18152 + }, + { + "epoch": 0.18153, + "grad_norm": 1.2085691193360635, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 18153 + }, + { + "epoch": 0.18154, + "grad_norm": 1.0908245865605166, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 18154 + }, + { + "epoch": 0.18155, + "grad_norm": 1.5099350769802962, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 18155 + }, + { + "epoch": 0.18156, + "grad_norm": 1.1255545526946735, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 18156 + }, + { + "epoch": 0.18157, + "grad_norm": 1.3686218230504699, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 18157 + }, + { + "epoch": 0.18158, + "grad_norm": 1.014531954026706, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 18158 + }, + { + "epoch": 0.18159, + "grad_norm": 1.5630287215197785, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 18159 + }, + { + "epoch": 0.1816, + "grad_norm": 1.1951031627857596, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 18160 + }, + { + "epoch": 0.18161, + "grad_norm": 1.4440238058794297, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 18161 + }, + { + "epoch": 0.18162, + "grad_norm": 1.2055848111490026, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 18162 + }, + { + "epoch": 0.18163, + "grad_norm": 1.3472878879325914, + "learning_rate": 0.003, + "loss": 4.012, + "step": 18163 + }, + { + "epoch": 0.18164, + "grad_norm": 1.261742286541519, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 18164 + }, + { + "epoch": 0.18165, + "grad_norm": 1.3588301712393138, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 18165 + }, + { + "epoch": 0.18166, + "grad_norm": 1.1585767016001705, + "learning_rate": 0.003, + "loss": 3.995, + "step": 18166 + }, + { + "epoch": 0.18167, + "grad_norm": 1.3441348595731184, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 18167 + }, + { + "epoch": 0.18168, + "grad_norm": 1.173183522395183, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 18168 + }, + { + "epoch": 0.18169, + "grad_norm": 1.533036177909173, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 18169 + }, + { + "epoch": 0.1817, + "grad_norm": 1.2600044816955933, + "learning_rate": 0.003, + "loss": 4.001, + "step": 18170 + }, + { + "epoch": 0.18171, + "grad_norm": 1.3100149997239376, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 18171 + }, + { + "epoch": 0.18172, + "grad_norm": 1.2136390858732689, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 18172 + }, + { + "epoch": 0.18173, + "grad_norm": 1.2607335650878135, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 18173 + }, + { + "epoch": 0.18174, + "grad_norm": 1.036453337144555, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 18174 + }, + { + "epoch": 0.18175, + "grad_norm": 1.4523655146583063, + "learning_rate": 0.003, + "loss": 4.033, + "step": 18175 + }, + { + "epoch": 0.18176, + "grad_norm": 1.048780729980621, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 18176 + }, + { + "epoch": 0.18177, + "grad_norm": 1.5229064863778266, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 18177 + }, + { + "epoch": 0.18178, + "grad_norm": 1.1509516821016428, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 18178 + }, + { + "epoch": 0.18179, + "grad_norm": 1.759995085427657, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 18179 + }, + { + "epoch": 0.1818, + "grad_norm": 1.0552996974550792, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 18180 + }, + { + "epoch": 0.18181, + "grad_norm": 1.30115622957003, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 18181 + }, + { + "epoch": 0.18182, + "grad_norm": 1.3024146378378039, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 18182 + }, + { + "epoch": 0.18183, + "grad_norm": 1.2110209860613366, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 18183 + }, + { + "epoch": 0.18184, + "grad_norm": 1.3931575048972484, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 18184 + }, + { + "epoch": 0.18185, + "grad_norm": 1.3834328466363612, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 18185 + }, + { + "epoch": 0.18186, + "grad_norm": 1.2536345227137324, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 18186 + }, + { + "epoch": 0.18187, + "grad_norm": 1.1041603580401191, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 18187 + }, + { + "epoch": 0.18188, + "grad_norm": 1.2536065909520349, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 18188 + }, + { + "epoch": 0.18189, + "grad_norm": 1.117385507479427, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 18189 + }, + { + "epoch": 0.1819, + "grad_norm": 1.2983498459511376, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18190 + }, + { + "epoch": 0.18191, + "grad_norm": 1.2184459973642885, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18191 + }, + { + "epoch": 0.18192, + "grad_norm": 1.1509796342701153, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 18192 + }, + { + "epoch": 0.18193, + "grad_norm": 1.2928667659153972, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 18193 + }, + { + "epoch": 0.18194, + "grad_norm": 1.3229778827050338, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18194 + }, + { + "epoch": 0.18195, + "grad_norm": 1.2935896839958851, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 18195 + }, + { + "epoch": 0.18196, + "grad_norm": 1.188892636841423, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 18196 + }, + { + "epoch": 0.18197, + "grad_norm": 1.2851214383382479, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 18197 + }, + { + "epoch": 0.18198, + "grad_norm": 1.3852784994657443, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 18198 + }, + { + "epoch": 0.18199, + "grad_norm": 1.4143040656970995, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 18199 + }, + { + "epoch": 0.182, + "grad_norm": 1.350309120327819, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18200 + }, + { + "epoch": 0.18201, + "grad_norm": 1.2407464281534468, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 18201 + }, + { + "epoch": 0.18202, + "grad_norm": 1.2119305968921505, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 18202 + }, + { + "epoch": 0.18203, + "grad_norm": 1.2101898464577356, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 18203 + }, + { + "epoch": 0.18204, + "grad_norm": 1.4768420872114572, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 18204 + }, + { + "epoch": 0.18205, + "grad_norm": 1.0963125232780624, + "learning_rate": 0.003, + "loss": 3.976, + "step": 18205 + }, + { + "epoch": 0.18206, + "grad_norm": 1.4907463450281417, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 18206 + }, + { + "epoch": 0.18207, + "grad_norm": 1.08251403467361, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 18207 + }, + { + "epoch": 0.18208, + "grad_norm": 1.3712794594338669, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 18208 + }, + { + "epoch": 0.18209, + "grad_norm": 0.9978160782565596, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 18209 + }, + { + "epoch": 0.1821, + "grad_norm": 1.3462577806423313, + "learning_rate": 0.003, + "loss": 4.016, + "step": 18210 + }, + { + "epoch": 0.18211, + "grad_norm": 1.0638576140205505, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 18211 + }, + { + "epoch": 0.18212, + "grad_norm": 1.487066531389715, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 18212 + }, + { + "epoch": 0.18213, + "grad_norm": 1.2345978757776293, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 18213 + }, + { + "epoch": 0.18214, + "grad_norm": 1.3367698913746935, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 18214 + }, + { + "epoch": 0.18215, + "grad_norm": 1.2982170491306446, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 18215 + }, + { + "epoch": 0.18216, + "grad_norm": 1.4710501586205065, + "learning_rate": 0.003, + "loss": 3.991, + "step": 18216 + }, + { + "epoch": 0.18217, + "grad_norm": 0.9724905612179099, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 18217 + }, + { + "epoch": 0.18218, + "grad_norm": 1.243080895345441, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 18218 + }, + { + "epoch": 0.18219, + "grad_norm": 1.1583332300816789, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 18219 + }, + { + "epoch": 0.1822, + "grad_norm": 1.4061740171975021, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 18220 + }, + { + "epoch": 0.18221, + "grad_norm": 1.3347184015308404, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 18221 + }, + { + "epoch": 0.18222, + "grad_norm": 1.197649749847447, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 18222 + }, + { + "epoch": 0.18223, + "grad_norm": 1.1559623563922403, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 18223 + }, + { + "epoch": 0.18224, + "grad_norm": 1.3258862057335294, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 18224 + }, + { + "epoch": 0.18225, + "grad_norm": 1.3489691972208808, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 18225 + }, + { + "epoch": 0.18226, + "grad_norm": 1.4596546168913027, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 18226 + }, + { + "epoch": 0.18227, + "grad_norm": 1.1011420292857328, + "learning_rate": 0.003, + "loss": 4.035, + "step": 18227 + }, + { + "epoch": 0.18228, + "grad_norm": 1.4691879728382518, + "learning_rate": 0.003, + "loss": 4.029, + "step": 18228 + }, + { + "epoch": 0.18229, + "grad_norm": 1.0600027382272688, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 18229 + }, + { + "epoch": 0.1823, + "grad_norm": 1.91440668806715, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 18230 + }, + { + "epoch": 0.18231, + "grad_norm": 1.087272659097562, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 18231 + }, + { + "epoch": 0.18232, + "grad_norm": 1.3160489393275667, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 18232 + }, + { + "epoch": 0.18233, + "grad_norm": 1.205997450393259, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 18233 + }, + { + "epoch": 0.18234, + "grad_norm": 1.2852537514127869, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 18234 + }, + { + "epoch": 0.18235, + "grad_norm": 1.3131140085067154, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 18235 + }, + { + "epoch": 0.18236, + "grad_norm": 1.3322776741884512, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 18236 + }, + { + "epoch": 0.18237, + "grad_norm": 1.2484259749190518, + "learning_rate": 0.003, + "loss": 4.015, + "step": 18237 + }, + { + "epoch": 0.18238, + "grad_norm": 1.1687486871042316, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 18238 + }, + { + "epoch": 0.18239, + "grad_norm": 1.069335222221601, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18239 + }, + { + "epoch": 0.1824, + "grad_norm": 1.555965179687311, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 18240 + }, + { + "epoch": 0.18241, + "grad_norm": 1.172706123428621, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 18241 + }, + { + "epoch": 0.18242, + "grad_norm": 1.6096084141251266, + "learning_rate": 0.003, + "loss": 3.984, + "step": 18242 + }, + { + "epoch": 0.18243, + "grad_norm": 1.1914307307720635, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 18243 + }, + { + "epoch": 0.18244, + "grad_norm": 1.3138377738427773, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 18244 + }, + { + "epoch": 0.18245, + "grad_norm": 1.3413909707937954, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 18245 + }, + { + "epoch": 0.18246, + "grad_norm": 1.0939883780957658, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 18246 + }, + { + "epoch": 0.18247, + "grad_norm": 1.2312089126378756, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 18247 + }, + { + "epoch": 0.18248, + "grad_norm": 1.3704760427576983, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 18248 + }, + { + "epoch": 0.18249, + "grad_norm": 1.1121469068817809, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 18249 + }, + { + "epoch": 0.1825, + "grad_norm": 1.2455867737055366, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 18250 + }, + { + "epoch": 0.18251, + "grad_norm": 1.5324873960709116, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 18251 + }, + { + "epoch": 0.18252, + "grad_norm": 1.0015715031516774, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 18252 + }, + { + "epoch": 0.18253, + "grad_norm": 1.130718883969915, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 18253 + }, + { + "epoch": 0.18254, + "grad_norm": 1.3137672179215365, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 18254 + }, + { + "epoch": 0.18255, + "grad_norm": 1.3104229198211335, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 18255 + }, + { + "epoch": 0.18256, + "grad_norm": 0.9945094424588728, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 18256 + }, + { + "epoch": 0.18257, + "grad_norm": 1.7183461639952, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 18257 + }, + { + "epoch": 0.18258, + "grad_norm": 1.1880604901138898, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 18258 + }, + { + "epoch": 0.18259, + "grad_norm": 1.1040386368554849, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 18259 + }, + { + "epoch": 0.1826, + "grad_norm": 1.5306036978635968, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 18260 + }, + { + "epoch": 0.18261, + "grad_norm": 1.3150678557370818, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 18261 + }, + { + "epoch": 0.18262, + "grad_norm": 1.409168557615195, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 18262 + }, + { + "epoch": 0.18263, + "grad_norm": 1.3989714258024877, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 18263 + }, + { + "epoch": 0.18264, + "grad_norm": 1.344570123510531, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 18264 + }, + { + "epoch": 0.18265, + "grad_norm": 1.123239125575655, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 18265 + }, + { + "epoch": 0.18266, + "grad_norm": 1.3037603359249763, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 18266 + }, + { + "epoch": 0.18267, + "grad_norm": 1.2861279634682024, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18267 + }, + { + "epoch": 0.18268, + "grad_norm": 1.3954523620095873, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 18268 + }, + { + "epoch": 0.18269, + "grad_norm": 1.232497391571598, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 18269 + }, + { + "epoch": 0.1827, + "grad_norm": 1.2787868742776072, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 18270 + }, + { + "epoch": 0.18271, + "grad_norm": 1.2320432357390205, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 18271 + }, + { + "epoch": 0.18272, + "grad_norm": 1.3320307314330602, + "learning_rate": 0.003, + "loss": 4.01, + "step": 18272 + }, + { + "epoch": 0.18273, + "grad_norm": 1.3135210430179733, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 18273 + }, + { + "epoch": 0.18274, + "grad_norm": 1.1245856238003318, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 18274 + }, + { + "epoch": 0.18275, + "grad_norm": 1.447129762793166, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 18275 + }, + { + "epoch": 0.18276, + "grad_norm": 1.1511065778756517, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 18276 + }, + { + "epoch": 0.18277, + "grad_norm": 1.4510257946581455, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 18277 + }, + { + "epoch": 0.18278, + "grad_norm": 1.268295122455787, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 18278 + }, + { + "epoch": 0.18279, + "grad_norm": 1.2968887235526405, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 18279 + }, + { + "epoch": 0.1828, + "grad_norm": 1.3690968920432314, + "learning_rate": 0.003, + "loss": 4.02, + "step": 18280 + }, + { + "epoch": 0.18281, + "grad_norm": 1.2938902793547082, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 18281 + }, + { + "epoch": 0.18282, + "grad_norm": 1.1835470553092178, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 18282 + }, + { + "epoch": 0.18283, + "grad_norm": 1.1957265377601571, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 18283 + }, + { + "epoch": 0.18284, + "grad_norm": 1.391943196815868, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 18284 + }, + { + "epoch": 0.18285, + "grad_norm": 1.107369458028556, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 18285 + }, + { + "epoch": 0.18286, + "grad_norm": 1.1283881078129234, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 18286 + }, + { + "epoch": 0.18287, + "grad_norm": 1.4298950698020385, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 18287 + }, + { + "epoch": 0.18288, + "grad_norm": 1.0464529595741836, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18288 + }, + { + "epoch": 0.18289, + "grad_norm": 1.5449404053034377, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 18289 + }, + { + "epoch": 0.1829, + "grad_norm": 1.1599745083921265, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 18290 + }, + { + "epoch": 0.18291, + "grad_norm": 1.3878360335531605, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 18291 + }, + { + "epoch": 0.18292, + "grad_norm": 1.128394557326477, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 18292 + }, + { + "epoch": 0.18293, + "grad_norm": 1.2015910000500716, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 18293 + }, + { + "epoch": 0.18294, + "grad_norm": 1.495744620608908, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 18294 + }, + { + "epoch": 0.18295, + "grad_norm": 1.188194593122797, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 18295 + }, + { + "epoch": 0.18296, + "grad_norm": 1.5503573027787667, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 18296 + }, + { + "epoch": 0.18297, + "grad_norm": 1.0139028798028031, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 18297 + }, + { + "epoch": 0.18298, + "grad_norm": 1.5246622301862984, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 18298 + }, + { + "epoch": 0.18299, + "grad_norm": 0.971631792285627, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 18299 + }, + { + "epoch": 0.183, + "grad_norm": 1.2882693419149442, + "learning_rate": 0.003, + "loss": 4.025, + "step": 18300 + }, + { + "epoch": 0.18301, + "grad_norm": 1.295287202260425, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18301 + }, + { + "epoch": 0.18302, + "grad_norm": 1.2151581662507234, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 18302 + }, + { + "epoch": 0.18303, + "grad_norm": 1.2384063271278432, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 18303 + }, + { + "epoch": 0.18304, + "grad_norm": 1.383694473007289, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 18304 + }, + { + "epoch": 0.18305, + "grad_norm": 1.1996409587123917, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 18305 + }, + { + "epoch": 0.18306, + "grad_norm": 1.2824445651452854, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 18306 + }, + { + "epoch": 0.18307, + "grad_norm": 1.2826231555044272, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 18307 + }, + { + "epoch": 0.18308, + "grad_norm": 1.3253483334986864, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 18308 + }, + { + "epoch": 0.18309, + "grad_norm": 1.3339497408438519, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 18309 + }, + { + "epoch": 0.1831, + "grad_norm": 1.335636097592298, + "learning_rate": 0.003, + "loss": 4.021, + "step": 18310 + }, + { + "epoch": 0.18311, + "grad_norm": 1.201242319528529, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 18311 + }, + { + "epoch": 0.18312, + "grad_norm": 1.202663178066936, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 18312 + }, + { + "epoch": 0.18313, + "grad_norm": 1.0739801360247447, + "learning_rate": 0.003, + "loss": 3.9366, + "step": 18313 + }, + { + "epoch": 0.18314, + "grad_norm": 1.3549682080063843, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 18314 + }, + { + "epoch": 0.18315, + "grad_norm": 1.2451047465446101, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 18315 + }, + { + "epoch": 0.18316, + "grad_norm": 1.1503352607657136, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 18316 + }, + { + "epoch": 0.18317, + "grad_norm": 1.4214259749166296, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 18317 + }, + { + "epoch": 0.18318, + "grad_norm": 1.1636299959381882, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 18318 + }, + { + "epoch": 0.18319, + "grad_norm": 1.637779383067376, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 18319 + }, + { + "epoch": 0.1832, + "grad_norm": 1.1294077294081348, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 18320 + }, + { + "epoch": 0.18321, + "grad_norm": 1.595618162969652, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 18321 + }, + { + "epoch": 0.18322, + "grad_norm": 1.385932302929975, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 18322 + }, + { + "epoch": 0.18323, + "grad_norm": 1.148155339367707, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 18323 + }, + { + "epoch": 0.18324, + "grad_norm": 1.2766032145718194, + "learning_rate": 0.003, + "loss": 4.022, + "step": 18324 + }, + { + "epoch": 0.18325, + "grad_norm": 1.348066500177508, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 18325 + }, + { + "epoch": 0.18326, + "grad_norm": 1.1033829098676295, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 18326 + }, + { + "epoch": 0.18327, + "grad_norm": 1.594138909184963, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 18327 + }, + { + "epoch": 0.18328, + "grad_norm": 0.9913174047152843, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 18328 + }, + { + "epoch": 0.18329, + "grad_norm": 1.451935960013849, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 18329 + }, + { + "epoch": 0.1833, + "grad_norm": 1.1632710985775636, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 18330 + }, + { + "epoch": 0.18331, + "grad_norm": 1.4401507582279762, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 18331 + }, + { + "epoch": 0.18332, + "grad_norm": 1.134330387221328, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18332 + }, + { + "epoch": 0.18333, + "grad_norm": 1.406834878988515, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 18333 + }, + { + "epoch": 0.18334, + "grad_norm": 1.2369064747229477, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 18334 + }, + { + "epoch": 0.18335, + "grad_norm": 1.1604113149220496, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 18335 + }, + { + "epoch": 0.18336, + "grad_norm": 1.3460719303584274, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 18336 + }, + { + "epoch": 0.18337, + "grad_norm": 1.2055253712252243, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 18337 + }, + { + "epoch": 0.18338, + "grad_norm": 1.252502635335824, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 18338 + }, + { + "epoch": 0.18339, + "grad_norm": 1.1572220045392678, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 18339 + }, + { + "epoch": 0.1834, + "grad_norm": 1.3235373092444123, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18340 + }, + { + "epoch": 0.18341, + "grad_norm": 1.2542925521645085, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 18341 + }, + { + "epoch": 0.18342, + "grad_norm": 1.2525317962625822, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 18342 + }, + { + "epoch": 0.18343, + "grad_norm": 1.23293947973228, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 18343 + }, + { + "epoch": 0.18344, + "grad_norm": 1.2771981048908279, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 18344 + }, + { + "epoch": 0.18345, + "grad_norm": 1.24303441554185, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 18345 + }, + { + "epoch": 0.18346, + "grad_norm": 1.1896657466868845, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 18346 + }, + { + "epoch": 0.18347, + "grad_norm": 1.3315160070580436, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 18347 + }, + { + "epoch": 0.18348, + "grad_norm": 1.1130315426915258, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 18348 + }, + { + "epoch": 0.18349, + "grad_norm": 1.534654876215884, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 18349 + }, + { + "epoch": 0.1835, + "grad_norm": 1.091582673242823, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 18350 + }, + { + "epoch": 0.18351, + "grad_norm": 1.236814118415645, + "learning_rate": 0.003, + "loss": 4.01, + "step": 18351 + }, + { + "epoch": 0.18352, + "grad_norm": 1.1432737189332145, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 18352 + }, + { + "epoch": 0.18353, + "grad_norm": 1.3846882584393398, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 18353 + }, + { + "epoch": 0.18354, + "grad_norm": 1.3679455609199003, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 18354 + }, + { + "epoch": 0.18355, + "grad_norm": 1.2603503593575665, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 18355 + }, + { + "epoch": 0.18356, + "grad_norm": 1.1296201645828623, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 18356 + }, + { + "epoch": 0.18357, + "grad_norm": 1.3538616687265028, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 18357 + }, + { + "epoch": 0.18358, + "grad_norm": 1.3092564527192359, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 18358 + }, + { + "epoch": 0.18359, + "grad_norm": 1.4992074553152586, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 18359 + }, + { + "epoch": 0.1836, + "grad_norm": 0.9900264088805295, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 18360 + }, + { + "epoch": 0.18361, + "grad_norm": 1.4370949498241343, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 18361 + }, + { + "epoch": 0.18362, + "grad_norm": 1.1886338472278208, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 18362 + }, + { + "epoch": 0.18363, + "grad_norm": 1.2032449147560684, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 18363 + }, + { + "epoch": 0.18364, + "grad_norm": 1.4954201193748469, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 18364 + }, + { + "epoch": 0.18365, + "grad_norm": 1.1318946680178235, + "learning_rate": 0.003, + "loss": 4.003, + "step": 18365 + }, + { + "epoch": 0.18366, + "grad_norm": 1.4399574101203345, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 18366 + }, + { + "epoch": 0.18367, + "grad_norm": 1.1804648048962672, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 18367 + }, + { + "epoch": 0.18368, + "grad_norm": 1.5449144085447852, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 18368 + }, + { + "epoch": 0.18369, + "grad_norm": 1.3209377674395755, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 18369 + }, + { + "epoch": 0.1837, + "grad_norm": 1.3595374939224947, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 18370 + }, + { + "epoch": 0.18371, + "grad_norm": 1.370017133050456, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 18371 + }, + { + "epoch": 0.18372, + "grad_norm": 1.2081595883407212, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 18372 + }, + { + "epoch": 0.18373, + "grad_norm": 1.3355722930673857, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 18373 + }, + { + "epoch": 0.18374, + "grad_norm": 1.1087151728729547, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 18374 + }, + { + "epoch": 0.18375, + "grad_norm": 1.3750018845575622, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 18375 + }, + { + "epoch": 0.18376, + "grad_norm": 1.1526653119120704, + "learning_rate": 0.003, + "loss": 4.027, + "step": 18376 + }, + { + "epoch": 0.18377, + "grad_norm": 1.3548582332658599, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 18377 + }, + { + "epoch": 0.18378, + "grad_norm": 1.2034073123719202, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 18378 + }, + { + "epoch": 0.18379, + "grad_norm": 1.2645937649587442, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 18379 + }, + { + "epoch": 0.1838, + "grad_norm": 1.2897176376196136, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 18380 + }, + { + "epoch": 0.18381, + "grad_norm": 1.1275837421245836, + "learning_rate": 0.003, + "loss": 4.029, + "step": 18381 + }, + { + "epoch": 0.18382, + "grad_norm": 1.2978066880951926, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 18382 + }, + { + "epoch": 0.18383, + "grad_norm": 1.128515129388203, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 18383 + }, + { + "epoch": 0.18384, + "grad_norm": 1.3590720577900184, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 18384 + }, + { + "epoch": 0.18385, + "grad_norm": 1.2900152539991172, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 18385 + }, + { + "epoch": 0.18386, + "grad_norm": 1.2926998050864664, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 18386 + }, + { + "epoch": 0.18387, + "grad_norm": 1.3083233474391918, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18387 + }, + { + "epoch": 0.18388, + "grad_norm": 1.2745656581962876, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18388 + }, + { + "epoch": 0.18389, + "grad_norm": 1.1794493601314162, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 18389 + }, + { + "epoch": 0.1839, + "grad_norm": 1.3066139417195313, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18390 + }, + { + "epoch": 0.18391, + "grad_norm": 1.4504619624419817, + "learning_rate": 0.003, + "loss": 3.996, + "step": 18391 + }, + { + "epoch": 0.18392, + "grad_norm": 1.3341319042643156, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 18392 + }, + { + "epoch": 0.18393, + "grad_norm": 1.3105393538818524, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 18393 + }, + { + "epoch": 0.18394, + "grad_norm": 1.4607388669237813, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 18394 + }, + { + "epoch": 0.18395, + "grad_norm": 0.9995550105643105, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 18395 + }, + { + "epoch": 0.18396, + "grad_norm": 1.384886578542871, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 18396 + }, + { + "epoch": 0.18397, + "grad_norm": 1.031830825123548, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 18397 + }, + { + "epoch": 0.18398, + "grad_norm": 1.4989640568496108, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18398 + }, + { + "epoch": 0.18399, + "grad_norm": 1.2320834742620708, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 18399 + }, + { + "epoch": 0.184, + "grad_norm": 1.1692123907725664, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 18400 + }, + { + "epoch": 0.18401, + "grad_norm": 1.3031665663242062, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 18401 + }, + { + "epoch": 0.18402, + "grad_norm": 1.1249323012008554, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 18402 + }, + { + "epoch": 0.18403, + "grad_norm": 1.198204293360691, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 18403 + }, + { + "epoch": 0.18404, + "grad_norm": 1.4103508443276054, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 18404 + }, + { + "epoch": 0.18405, + "grad_norm": 1.2898062167184823, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18405 + }, + { + "epoch": 0.18406, + "grad_norm": 1.3525117916984277, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 18406 + }, + { + "epoch": 0.18407, + "grad_norm": 1.3484237042989866, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 18407 + }, + { + "epoch": 0.18408, + "grad_norm": 1.4461309714025212, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 18408 + }, + { + "epoch": 0.18409, + "grad_norm": 1.2975573996791996, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 18409 + }, + { + "epoch": 0.1841, + "grad_norm": 1.2114947855094296, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 18410 + }, + { + "epoch": 0.18411, + "grad_norm": 1.2452581120611896, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 18411 + }, + { + "epoch": 0.18412, + "grad_norm": 1.470607565911789, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 18412 + }, + { + "epoch": 0.18413, + "grad_norm": 1.2262106788202565, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 18413 + }, + { + "epoch": 0.18414, + "grad_norm": 1.2580826518725825, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 18414 + }, + { + "epoch": 0.18415, + "grad_norm": 1.2762030597233442, + "learning_rate": 0.003, + "loss": 4.0, + "step": 18415 + }, + { + "epoch": 0.18416, + "grad_norm": 1.3006937132738488, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 18416 + }, + { + "epoch": 0.18417, + "grad_norm": 1.1144728301235787, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 18417 + }, + { + "epoch": 0.18418, + "grad_norm": 1.2882672348453161, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 18418 + }, + { + "epoch": 0.18419, + "grad_norm": 1.2156632443483395, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 18419 + }, + { + "epoch": 0.1842, + "grad_norm": 1.275090643963137, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 18420 + }, + { + "epoch": 0.18421, + "grad_norm": 1.1046831206828236, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 18421 + }, + { + "epoch": 0.18422, + "grad_norm": 1.458164902379381, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 18422 + }, + { + "epoch": 0.18423, + "grad_norm": 1.2523878984847892, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 18423 + }, + { + "epoch": 0.18424, + "grad_norm": 1.2214335287255744, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 18424 + }, + { + "epoch": 0.18425, + "grad_norm": 1.298262007903684, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 18425 + }, + { + "epoch": 0.18426, + "grad_norm": 1.297408437181403, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 18426 + }, + { + "epoch": 0.18427, + "grad_norm": 1.4194071833600868, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 18427 + }, + { + "epoch": 0.18428, + "grad_norm": 1.175283425110903, + "learning_rate": 0.003, + "loss": 4.011, + "step": 18428 + }, + { + "epoch": 0.18429, + "grad_norm": 1.1551904179812944, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 18429 + }, + { + "epoch": 0.1843, + "grad_norm": 1.2026732845787054, + "learning_rate": 0.003, + "loss": 3.983, + "step": 18430 + }, + { + "epoch": 0.18431, + "grad_norm": 1.2434555943262482, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 18431 + }, + { + "epoch": 0.18432, + "grad_norm": 1.1768527459774172, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 18432 + }, + { + "epoch": 0.18433, + "grad_norm": 1.240412598853343, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 18433 + }, + { + "epoch": 0.18434, + "grad_norm": 1.2692967701726239, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 18434 + }, + { + "epoch": 0.18435, + "grad_norm": 1.4220878500122178, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 18435 + }, + { + "epoch": 0.18436, + "grad_norm": 1.1348401403953556, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18436 + }, + { + "epoch": 0.18437, + "grad_norm": 1.2374013020325947, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 18437 + }, + { + "epoch": 0.18438, + "grad_norm": 1.2858501719262228, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18438 + }, + { + "epoch": 0.18439, + "grad_norm": 1.156999599660775, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 18439 + }, + { + "epoch": 0.1844, + "grad_norm": 1.3274603523403214, + "learning_rate": 0.003, + "loss": 3.999, + "step": 18440 + }, + { + "epoch": 0.18441, + "grad_norm": 1.2713213342559737, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 18441 + }, + { + "epoch": 0.18442, + "grad_norm": 1.3468046555432271, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 18442 + }, + { + "epoch": 0.18443, + "grad_norm": 1.3106291424707803, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 18443 + }, + { + "epoch": 0.18444, + "grad_norm": 1.250152339274203, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 18444 + }, + { + "epoch": 0.18445, + "grad_norm": 1.4384507501738175, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 18445 + }, + { + "epoch": 0.18446, + "grad_norm": 1.227753855892706, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 18446 + }, + { + "epoch": 0.18447, + "grad_norm": 1.3188968565202013, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 18447 + }, + { + "epoch": 0.18448, + "grad_norm": 1.3824282906882581, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 18448 + }, + { + "epoch": 0.18449, + "grad_norm": 1.1178579815144731, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 18449 + }, + { + "epoch": 0.1845, + "grad_norm": 1.4560887414596302, + "learning_rate": 0.003, + "loss": 3.999, + "step": 18450 + }, + { + "epoch": 0.18451, + "grad_norm": 0.9358257931900706, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18451 + }, + { + "epoch": 0.18452, + "grad_norm": 1.4848874911550862, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 18452 + }, + { + "epoch": 0.18453, + "grad_norm": 1.338522971259659, + "learning_rate": 0.003, + "loss": 4.057, + "step": 18453 + }, + { + "epoch": 0.18454, + "grad_norm": 1.4725522007476801, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 18454 + }, + { + "epoch": 0.18455, + "grad_norm": 1.1681160129084804, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 18455 + }, + { + "epoch": 0.18456, + "grad_norm": 1.327267783294372, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18456 + }, + { + "epoch": 0.18457, + "grad_norm": 1.2578832106370543, + "learning_rate": 0.003, + "loss": 4.035, + "step": 18457 + }, + { + "epoch": 0.18458, + "grad_norm": 1.348036752864334, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 18458 + }, + { + "epoch": 0.18459, + "grad_norm": 1.3769169085120323, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 18459 + }, + { + "epoch": 0.1846, + "grad_norm": 1.7663306920162887, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 18460 + }, + { + "epoch": 0.18461, + "grad_norm": 1.1346456245594003, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 18461 + }, + { + "epoch": 0.18462, + "grad_norm": 1.2327644341051074, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 18462 + }, + { + "epoch": 0.18463, + "grad_norm": 1.1281774861720366, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 18463 + }, + { + "epoch": 0.18464, + "grad_norm": 1.2859152465019796, + "learning_rate": 0.003, + "loss": 4.022, + "step": 18464 + }, + { + "epoch": 0.18465, + "grad_norm": 1.3017491811776885, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 18465 + }, + { + "epoch": 0.18466, + "grad_norm": 1.2218602647698296, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 18466 + }, + { + "epoch": 0.18467, + "grad_norm": 1.3017897342152858, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 18467 + }, + { + "epoch": 0.18468, + "grad_norm": 1.1866958547635929, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 18468 + }, + { + "epoch": 0.18469, + "grad_norm": 1.1458159820620728, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 18469 + }, + { + "epoch": 0.1847, + "grad_norm": 1.2302168602397137, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 18470 + }, + { + "epoch": 0.18471, + "grad_norm": 1.3917697686622776, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 18471 + }, + { + "epoch": 0.18472, + "grad_norm": 1.1051796677592909, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 18472 + }, + { + "epoch": 0.18473, + "grad_norm": 1.4759237634105464, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 18473 + }, + { + "epoch": 0.18474, + "grad_norm": 1.4656747300837214, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 18474 + }, + { + "epoch": 0.18475, + "grad_norm": 1.299117857979297, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 18475 + }, + { + "epoch": 0.18476, + "grad_norm": 1.238280623167555, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 18476 + }, + { + "epoch": 0.18477, + "grad_norm": 1.4048847403795164, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 18477 + }, + { + "epoch": 0.18478, + "grad_norm": 1.1090434443199049, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 18478 + }, + { + "epoch": 0.18479, + "grad_norm": 1.4213046723745388, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 18479 + }, + { + "epoch": 0.1848, + "grad_norm": 1.279818057568252, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 18480 + }, + { + "epoch": 0.18481, + "grad_norm": 1.3643906040670437, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 18481 + }, + { + "epoch": 0.18482, + "grad_norm": 1.331545735966065, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 18482 + }, + { + "epoch": 0.18483, + "grad_norm": 1.4582865434308194, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 18483 + }, + { + "epoch": 0.18484, + "grad_norm": 1.1033483229133811, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 18484 + }, + { + "epoch": 0.18485, + "grad_norm": 1.220863438035, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18485 + }, + { + "epoch": 0.18486, + "grad_norm": 1.1819377221490115, + "learning_rate": 0.003, + "loss": 3.97, + "step": 18486 + }, + { + "epoch": 0.18487, + "grad_norm": 1.2913761031851838, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 18487 + }, + { + "epoch": 0.18488, + "grad_norm": 1.3534671638086488, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18488 + }, + { + "epoch": 0.18489, + "grad_norm": 1.4161610107981784, + "learning_rate": 0.003, + "loss": 4.022, + "step": 18489 + }, + { + "epoch": 0.1849, + "grad_norm": 1.1101131904508508, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 18490 + }, + { + "epoch": 0.18491, + "grad_norm": 1.4841512493611402, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18491 + }, + { + "epoch": 0.18492, + "grad_norm": 1.075679689426423, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 18492 + }, + { + "epoch": 0.18493, + "grad_norm": 1.5403311468925576, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 18493 + }, + { + "epoch": 0.18494, + "grad_norm": 1.0948628348607796, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 18494 + }, + { + "epoch": 0.18495, + "grad_norm": 1.5168282201704515, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 18495 + }, + { + "epoch": 0.18496, + "grad_norm": 1.066571763307485, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 18496 + }, + { + "epoch": 0.18497, + "grad_norm": 1.4482722506631425, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18497 + }, + { + "epoch": 0.18498, + "grad_norm": 1.3925698544647553, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 18498 + }, + { + "epoch": 0.18499, + "grad_norm": 1.407952620323706, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 18499 + }, + { + "epoch": 0.185, + "grad_norm": 1.2013361484565197, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 18500 + }, + { + "epoch": 0.18501, + "grad_norm": 1.358993966037197, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 18501 + }, + { + "epoch": 0.18502, + "grad_norm": 1.251426452479606, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 18502 + }, + { + "epoch": 0.18503, + "grad_norm": 1.3535328239988025, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 18503 + }, + { + "epoch": 0.18504, + "grad_norm": 1.2846308012873837, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18504 + }, + { + "epoch": 0.18505, + "grad_norm": 1.3831662011923265, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 18505 + }, + { + "epoch": 0.18506, + "grad_norm": 1.3756100946984784, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 18506 + }, + { + "epoch": 0.18507, + "grad_norm": 1.2082064307048737, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 18507 + }, + { + "epoch": 0.18508, + "grad_norm": 1.2346584097097666, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 18508 + }, + { + "epoch": 0.18509, + "grad_norm": 1.1118011345861378, + "learning_rate": 0.003, + "loss": 3.984, + "step": 18509 + }, + { + "epoch": 0.1851, + "grad_norm": 1.4618268255864204, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 18510 + }, + { + "epoch": 0.18511, + "grad_norm": 1.330748768986842, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 18511 + }, + { + "epoch": 0.18512, + "grad_norm": 1.1974218340672207, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 18512 + }, + { + "epoch": 0.18513, + "grad_norm": 1.3318792603017142, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 18513 + }, + { + "epoch": 0.18514, + "grad_norm": 1.0947871098533344, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 18514 + }, + { + "epoch": 0.18515, + "grad_norm": 1.4433388959752576, + "learning_rate": 0.003, + "loss": 3.972, + "step": 18515 + }, + { + "epoch": 0.18516, + "grad_norm": 1.061057613490036, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 18516 + }, + { + "epoch": 0.18517, + "grad_norm": 1.4735050302149566, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 18517 + }, + { + "epoch": 0.18518, + "grad_norm": 1.1761057438168303, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 18518 + }, + { + "epoch": 0.18519, + "grad_norm": 1.3637153392472225, + "learning_rate": 0.003, + "loss": 3.997, + "step": 18519 + }, + { + "epoch": 0.1852, + "grad_norm": 0.9907075153743939, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 18520 + }, + { + "epoch": 0.18521, + "grad_norm": 1.4587027918044375, + "learning_rate": 0.003, + "loss": 4.014, + "step": 18521 + }, + { + "epoch": 0.18522, + "grad_norm": 1.065505001624872, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 18522 + }, + { + "epoch": 0.18523, + "grad_norm": 1.2618717307106824, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 18523 + }, + { + "epoch": 0.18524, + "grad_norm": 1.2549102381618809, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 18524 + }, + { + "epoch": 0.18525, + "grad_norm": 1.1756058585589122, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 18525 + }, + { + "epoch": 0.18526, + "grad_norm": 1.308331068971566, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 18526 + }, + { + "epoch": 0.18527, + "grad_norm": 1.2308889043734412, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 18527 + }, + { + "epoch": 0.18528, + "grad_norm": 1.341743389444327, + "learning_rate": 0.003, + "loss": 4.032, + "step": 18528 + }, + { + "epoch": 0.18529, + "grad_norm": 1.2788693846335635, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 18529 + }, + { + "epoch": 0.1853, + "grad_norm": 1.212905158240625, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18530 + }, + { + "epoch": 0.18531, + "grad_norm": 1.2585666115433096, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 18531 + }, + { + "epoch": 0.18532, + "grad_norm": 1.6025444100201056, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 18532 + }, + { + "epoch": 0.18533, + "grad_norm": 1.2214621764586233, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 18533 + }, + { + "epoch": 0.18534, + "grad_norm": 1.430485782386072, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 18534 + }, + { + "epoch": 0.18535, + "grad_norm": 1.3164521462885879, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 18535 + }, + { + "epoch": 0.18536, + "grad_norm": 1.276515990912035, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 18536 + }, + { + "epoch": 0.18537, + "grad_norm": 1.2701183695610685, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 18537 + }, + { + "epoch": 0.18538, + "grad_norm": 1.1189715005336927, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 18538 + }, + { + "epoch": 0.18539, + "grad_norm": 1.3619235261804277, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 18539 + }, + { + "epoch": 0.1854, + "grad_norm": 1.233576204494665, + "learning_rate": 0.003, + "loss": 4.018, + "step": 18540 + }, + { + "epoch": 0.18541, + "grad_norm": 1.3692736931002547, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 18541 + }, + { + "epoch": 0.18542, + "grad_norm": 1.0912947060898697, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 18542 + }, + { + "epoch": 0.18543, + "grad_norm": 1.4822550236144991, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 18543 + }, + { + "epoch": 0.18544, + "grad_norm": 1.0481702315363652, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 18544 + }, + { + "epoch": 0.18545, + "grad_norm": 1.3190514168279799, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 18545 + }, + { + "epoch": 0.18546, + "grad_norm": 1.1342856434951607, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 18546 + }, + { + "epoch": 0.18547, + "grad_norm": 1.4563981963795782, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 18547 + }, + { + "epoch": 0.18548, + "grad_norm": 1.1505326612441675, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 18548 + }, + { + "epoch": 0.18549, + "grad_norm": 1.348571565178306, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 18549 + }, + { + "epoch": 0.1855, + "grad_norm": 1.2459203506838015, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18550 + }, + { + "epoch": 0.18551, + "grad_norm": 1.3952337410832205, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 18551 + }, + { + "epoch": 0.18552, + "grad_norm": 1.2513571202488907, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 18552 + }, + { + "epoch": 0.18553, + "grad_norm": 1.308467551674208, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 18553 + }, + { + "epoch": 0.18554, + "grad_norm": 1.0101160805280176, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 18554 + }, + { + "epoch": 0.18555, + "grad_norm": 1.201354445868851, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 18555 + }, + { + "epoch": 0.18556, + "grad_norm": 1.3008756176695497, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 18556 + }, + { + "epoch": 0.18557, + "grad_norm": 1.1840527587871843, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 18557 + }, + { + "epoch": 0.18558, + "grad_norm": 1.5502327699695684, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 18558 + }, + { + "epoch": 0.18559, + "grad_norm": 1.1948999804123228, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 18559 + }, + { + "epoch": 0.1856, + "grad_norm": 1.2310717704591958, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 18560 + }, + { + "epoch": 0.18561, + "grad_norm": 1.2907980238799903, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 18561 + }, + { + "epoch": 0.18562, + "grad_norm": 1.3389978550625135, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 18562 + }, + { + "epoch": 0.18563, + "grad_norm": 1.2244157868322538, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 18563 + }, + { + "epoch": 0.18564, + "grad_norm": 1.1847773023779853, + "learning_rate": 0.003, + "loss": 3.989, + "step": 18564 + }, + { + "epoch": 0.18565, + "grad_norm": 1.5214512122252601, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 18565 + }, + { + "epoch": 0.18566, + "grad_norm": 1.0766529054176561, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 18566 + }, + { + "epoch": 0.18567, + "grad_norm": 1.4010076254541792, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 18567 + }, + { + "epoch": 0.18568, + "grad_norm": 0.991739314032095, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 18568 + }, + { + "epoch": 0.18569, + "grad_norm": 1.2838558466306362, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 18569 + }, + { + "epoch": 0.1857, + "grad_norm": 1.220292414323158, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 18570 + }, + { + "epoch": 0.18571, + "grad_norm": 1.322845412629368, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 18571 + }, + { + "epoch": 0.18572, + "grad_norm": 1.3582577169073362, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 18572 + }, + { + "epoch": 0.18573, + "grad_norm": 1.200184426982225, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 18573 + }, + { + "epoch": 0.18574, + "grad_norm": 1.5527522075179154, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 18574 + }, + { + "epoch": 0.18575, + "grad_norm": 1.1582205921040074, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 18575 + }, + { + "epoch": 0.18576, + "grad_norm": 1.2330870377840064, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 18576 + }, + { + "epoch": 0.18577, + "grad_norm": 1.2252907480898436, + "learning_rate": 0.003, + "loss": 3.99, + "step": 18577 + }, + { + "epoch": 0.18578, + "grad_norm": 1.023257381028075, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 18578 + }, + { + "epoch": 0.18579, + "grad_norm": 1.4218033839458162, + "learning_rate": 0.003, + "loss": 4.041, + "step": 18579 + }, + { + "epoch": 0.1858, + "grad_norm": 1.3118197249151076, + "learning_rate": 0.003, + "loss": 3.985, + "step": 18580 + }, + { + "epoch": 0.18581, + "grad_norm": 1.3664557565587128, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 18581 + }, + { + "epoch": 0.18582, + "grad_norm": 1.4781456426824633, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18582 + }, + { + "epoch": 0.18583, + "grad_norm": 1.128398572392948, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 18583 + }, + { + "epoch": 0.18584, + "grad_norm": 1.6347654062622279, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18584 + }, + { + "epoch": 0.18585, + "grad_norm": 1.031502279419643, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 18585 + }, + { + "epoch": 0.18586, + "grad_norm": 1.433874910092499, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 18586 + }, + { + "epoch": 0.18587, + "grad_norm": 1.0733447413741035, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 18587 + }, + { + "epoch": 0.18588, + "grad_norm": 1.5475095622967026, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 18588 + }, + { + "epoch": 0.18589, + "grad_norm": 1.0194451496728019, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18589 + }, + { + "epoch": 0.1859, + "grad_norm": 1.5307823062281203, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18590 + }, + { + "epoch": 0.18591, + "grad_norm": 1.1275286995807612, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 18591 + }, + { + "epoch": 0.18592, + "grad_norm": 1.2933826175634457, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 18592 + }, + { + "epoch": 0.18593, + "grad_norm": 1.3601993739587592, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 18593 + }, + { + "epoch": 0.18594, + "grad_norm": 1.2130492766252166, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 18594 + }, + { + "epoch": 0.18595, + "grad_norm": 1.2934433237292364, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 18595 + }, + { + "epoch": 0.18596, + "grad_norm": 1.2421427960715283, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 18596 + }, + { + "epoch": 0.18597, + "grad_norm": 1.490141776937668, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 18597 + }, + { + "epoch": 0.18598, + "grad_norm": 1.3658652684554957, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 18598 + }, + { + "epoch": 0.18599, + "grad_norm": 1.286534104846973, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 18599 + }, + { + "epoch": 0.186, + "grad_norm": 1.2191421245523897, + "learning_rate": 0.003, + "loss": 4.0, + "step": 18600 + }, + { + "epoch": 0.18601, + "grad_norm": 1.3072430551357463, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 18601 + }, + { + "epoch": 0.18602, + "grad_norm": 1.242549894399749, + "learning_rate": 0.003, + "loss": 3.978, + "step": 18602 + }, + { + "epoch": 0.18603, + "grad_norm": 1.453242763100252, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 18603 + }, + { + "epoch": 0.18604, + "grad_norm": 1.2777779597545433, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 18604 + }, + { + "epoch": 0.18605, + "grad_norm": 1.2684022449348649, + "learning_rate": 0.003, + "loss": 4.012, + "step": 18605 + }, + { + "epoch": 0.18606, + "grad_norm": 1.1734346703901422, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 18606 + }, + { + "epoch": 0.18607, + "grad_norm": 1.4009858584521866, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 18607 + }, + { + "epoch": 0.18608, + "grad_norm": 0.9551117231543502, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 18608 + }, + { + "epoch": 0.18609, + "grad_norm": 1.3261185996717406, + "learning_rate": 0.003, + "loss": 4.024, + "step": 18609 + }, + { + "epoch": 0.1861, + "grad_norm": 1.0500612330987849, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 18610 + }, + { + "epoch": 0.18611, + "grad_norm": 1.792206931986703, + "learning_rate": 0.003, + "loss": 4.021, + "step": 18611 + }, + { + "epoch": 0.18612, + "grad_norm": 1.0644116367360816, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 18612 + }, + { + "epoch": 0.18613, + "grad_norm": 1.6239202401149886, + "learning_rate": 0.003, + "loss": 4.061, + "step": 18613 + }, + { + "epoch": 0.18614, + "grad_norm": 1.1228693236587073, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 18614 + }, + { + "epoch": 0.18615, + "grad_norm": 1.3021924314613067, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 18615 + }, + { + "epoch": 0.18616, + "grad_norm": 1.3454363448652122, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 18616 + }, + { + "epoch": 0.18617, + "grad_norm": 1.1555190229942982, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18617 + }, + { + "epoch": 0.18618, + "grad_norm": 1.3870926743097056, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 18618 + }, + { + "epoch": 0.18619, + "grad_norm": 1.4728282282089995, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 18619 + }, + { + "epoch": 0.1862, + "grad_norm": 1.3300007177718682, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 18620 + }, + { + "epoch": 0.18621, + "grad_norm": 1.2998396089776643, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 18621 + }, + { + "epoch": 0.18622, + "grad_norm": 1.663830580793456, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 18622 + }, + { + "epoch": 0.18623, + "grad_norm": 1.1297681880161663, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 18623 + }, + { + "epoch": 0.18624, + "grad_norm": 1.3008520977812261, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 18624 + }, + { + "epoch": 0.18625, + "grad_norm": 1.2806620435849159, + "learning_rate": 0.003, + "loss": 4.015, + "step": 18625 + }, + { + "epoch": 0.18626, + "grad_norm": 1.3016566169943902, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 18626 + }, + { + "epoch": 0.18627, + "grad_norm": 1.2693203511028142, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 18627 + }, + { + "epoch": 0.18628, + "grad_norm": 1.2351005941004773, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18628 + }, + { + "epoch": 0.18629, + "grad_norm": 1.296226250236291, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 18629 + }, + { + "epoch": 0.1863, + "grad_norm": 1.0927295505933619, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 18630 + }, + { + "epoch": 0.18631, + "grad_norm": 1.289378498409426, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 18631 + }, + { + "epoch": 0.18632, + "grad_norm": 1.3590661123448855, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 18632 + }, + { + "epoch": 0.18633, + "grad_norm": 1.1964437431765969, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 18633 + }, + { + "epoch": 0.18634, + "grad_norm": 1.3144287374617258, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 18634 + }, + { + "epoch": 0.18635, + "grad_norm": 1.0505604453924078, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 18635 + }, + { + "epoch": 0.18636, + "grad_norm": 1.470577508208389, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 18636 + }, + { + "epoch": 0.18637, + "grad_norm": 1.1033046640383908, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 18637 + }, + { + "epoch": 0.18638, + "grad_norm": 1.522697024614939, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 18638 + }, + { + "epoch": 0.18639, + "grad_norm": 1.2251818029759538, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 18639 + }, + { + "epoch": 0.1864, + "grad_norm": 1.3033259395285126, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18640 + }, + { + "epoch": 0.18641, + "grad_norm": 1.2300175335879666, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 18641 + }, + { + "epoch": 0.18642, + "grad_norm": 1.1259367697089997, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 18642 + }, + { + "epoch": 0.18643, + "grad_norm": 1.444943076921912, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 18643 + }, + { + "epoch": 0.18644, + "grad_norm": 1.0268303432894565, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 18644 + }, + { + "epoch": 0.18645, + "grad_norm": 1.5852151526494689, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 18645 + }, + { + "epoch": 0.18646, + "grad_norm": 1.0043957113578743, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 18646 + }, + { + "epoch": 0.18647, + "grad_norm": 1.785965563009103, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 18647 + }, + { + "epoch": 0.18648, + "grad_norm": 1.0897731949512062, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18648 + }, + { + "epoch": 0.18649, + "grad_norm": 1.1966497791830422, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 18649 + }, + { + "epoch": 0.1865, + "grad_norm": 1.1030508775016512, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 18650 + }, + { + "epoch": 0.18651, + "grad_norm": 1.4359024778859895, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 18651 + }, + { + "epoch": 0.18652, + "grad_norm": 1.3060847385318657, + "learning_rate": 0.003, + "loss": 4.044, + "step": 18652 + }, + { + "epoch": 0.18653, + "grad_norm": 1.2815378552222991, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18653 + }, + { + "epoch": 0.18654, + "grad_norm": 1.3081912359863301, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 18654 + }, + { + "epoch": 0.18655, + "grad_norm": 1.1841226312470723, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 18655 + }, + { + "epoch": 0.18656, + "grad_norm": 1.0474684392139904, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 18656 + }, + { + "epoch": 0.18657, + "grad_norm": 1.2089374947444256, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 18657 + }, + { + "epoch": 0.18658, + "grad_norm": 1.2119073765559805, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 18658 + }, + { + "epoch": 0.18659, + "grad_norm": 1.2712102129241536, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 18659 + }, + { + "epoch": 0.1866, + "grad_norm": 1.3573528545499387, + "learning_rate": 0.003, + "loss": 3.995, + "step": 18660 + }, + { + "epoch": 0.18661, + "grad_norm": 1.4130170197773813, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 18661 + }, + { + "epoch": 0.18662, + "grad_norm": 1.1507026696311244, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 18662 + }, + { + "epoch": 0.18663, + "grad_norm": 1.5848939484272642, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 18663 + }, + { + "epoch": 0.18664, + "grad_norm": 1.3501497587585538, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 18664 + }, + { + "epoch": 0.18665, + "grad_norm": 1.3863001903757999, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 18665 + }, + { + "epoch": 0.18666, + "grad_norm": 1.2816916346464242, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 18666 + }, + { + "epoch": 0.18667, + "grad_norm": 1.0273145525259801, + "learning_rate": 0.003, + "loss": 4.015, + "step": 18667 + }, + { + "epoch": 0.18668, + "grad_norm": 1.4102810177452896, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 18668 + }, + { + "epoch": 0.18669, + "grad_norm": 1.0623139759039228, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 18669 + }, + { + "epoch": 0.1867, + "grad_norm": 1.4461637831636758, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 18670 + }, + { + "epoch": 0.18671, + "grad_norm": 1.0309532446981018, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 18671 + }, + { + "epoch": 0.18672, + "grad_norm": 1.5276803462303443, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 18672 + }, + { + "epoch": 0.18673, + "grad_norm": 0.9511090963111622, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 18673 + }, + { + "epoch": 0.18674, + "grad_norm": 1.301143672970116, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 18674 + }, + { + "epoch": 0.18675, + "grad_norm": 1.2147389803487254, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 18675 + }, + { + "epoch": 0.18676, + "grad_norm": 1.110001119192037, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 18676 + }, + { + "epoch": 0.18677, + "grad_norm": 1.5169873876232451, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 18677 + }, + { + "epoch": 0.18678, + "grad_norm": 1.30388104159054, + "learning_rate": 0.003, + "loss": 4.012, + "step": 18678 + }, + { + "epoch": 0.18679, + "grad_norm": 1.4226466874028336, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 18679 + }, + { + "epoch": 0.1868, + "grad_norm": 1.2521023613556785, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18680 + }, + { + "epoch": 0.18681, + "grad_norm": 1.342154011332591, + "learning_rate": 0.003, + "loss": 4.011, + "step": 18681 + }, + { + "epoch": 0.18682, + "grad_norm": 1.151351223071168, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 18682 + }, + { + "epoch": 0.18683, + "grad_norm": 1.3783113488733716, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 18683 + }, + { + "epoch": 0.18684, + "grad_norm": 1.0337304440380508, + "learning_rate": 0.003, + "loss": 4.024, + "step": 18684 + }, + { + "epoch": 0.18685, + "grad_norm": 1.5841944883036514, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 18685 + }, + { + "epoch": 0.18686, + "grad_norm": 0.9641070777729437, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 18686 + }, + { + "epoch": 0.18687, + "grad_norm": 1.6626898746580234, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 18687 + }, + { + "epoch": 0.18688, + "grad_norm": 0.9977541504269526, + "learning_rate": 0.003, + "loss": 4.009, + "step": 18688 + }, + { + "epoch": 0.18689, + "grad_norm": 1.3963131002959657, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 18689 + }, + { + "epoch": 0.1869, + "grad_norm": 1.2682488828591068, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 18690 + }, + { + "epoch": 0.18691, + "grad_norm": 1.21170146686143, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 18691 + }, + { + "epoch": 0.18692, + "grad_norm": 1.4365591225461198, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 18692 + }, + { + "epoch": 0.18693, + "grad_norm": 1.1154666207606887, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 18693 + }, + { + "epoch": 0.18694, + "grad_norm": 1.4480910223178798, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 18694 + }, + { + "epoch": 0.18695, + "grad_norm": 1.5202125009114869, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 18695 + }, + { + "epoch": 0.18696, + "grad_norm": 1.0368577367197942, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 18696 + }, + { + "epoch": 0.18697, + "grad_norm": 1.528670472485098, + "learning_rate": 0.003, + "loss": 4.003, + "step": 18697 + }, + { + "epoch": 0.18698, + "grad_norm": 1.0822675430061508, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 18698 + }, + { + "epoch": 0.18699, + "grad_norm": 1.6733763431881195, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 18699 + }, + { + "epoch": 0.187, + "grad_norm": 1.0336614282538712, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 18700 + }, + { + "epoch": 0.18701, + "grad_norm": 1.306762222550216, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 18701 + }, + { + "epoch": 0.18702, + "grad_norm": 1.453126556274193, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 18702 + }, + { + "epoch": 0.18703, + "grad_norm": 1.0154430609976544, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18703 + }, + { + "epoch": 0.18704, + "grad_norm": 1.5981658504580856, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 18704 + }, + { + "epoch": 0.18705, + "grad_norm": 1.0265576550733797, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 18705 + }, + { + "epoch": 0.18706, + "grad_norm": 1.3013077718481592, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 18706 + }, + { + "epoch": 0.18707, + "grad_norm": 1.265742895619668, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 18707 + }, + { + "epoch": 0.18708, + "grad_norm": 1.1374193006402677, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 18708 + }, + { + "epoch": 0.18709, + "grad_norm": 1.3431087880743602, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 18709 + }, + { + "epoch": 0.1871, + "grad_norm": 1.314450011320492, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 18710 + }, + { + "epoch": 0.18711, + "grad_norm": 1.3019829829934781, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 18711 + }, + { + "epoch": 0.18712, + "grad_norm": 1.1132955579832229, + "learning_rate": 0.003, + "loss": 4.026, + "step": 18712 + }, + { + "epoch": 0.18713, + "grad_norm": 1.2236759716132808, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 18713 + }, + { + "epoch": 0.18714, + "grad_norm": 1.270980229472738, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18714 + }, + { + "epoch": 0.18715, + "grad_norm": 1.322438126171909, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 18715 + }, + { + "epoch": 0.18716, + "grad_norm": 1.0512819280225125, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 18716 + }, + { + "epoch": 0.18717, + "grad_norm": 1.458252119498716, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 18717 + }, + { + "epoch": 0.18718, + "grad_norm": 1.0607680755776179, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 18718 + }, + { + "epoch": 0.18719, + "grad_norm": 1.3862670015696035, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 18719 + }, + { + "epoch": 0.1872, + "grad_norm": 1.2237777419380014, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 18720 + }, + { + "epoch": 0.18721, + "grad_norm": 1.376918904288084, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 18721 + }, + { + "epoch": 0.18722, + "grad_norm": 1.170486727504894, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 18722 + }, + { + "epoch": 0.18723, + "grad_norm": 1.3146797562996735, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 18723 + }, + { + "epoch": 0.18724, + "grad_norm": 1.3750013180924125, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 18724 + }, + { + "epoch": 0.18725, + "grad_norm": 1.1278171591869586, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 18725 + }, + { + "epoch": 0.18726, + "grad_norm": 1.3376324892975, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 18726 + }, + { + "epoch": 0.18727, + "grad_norm": 1.057740063162026, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 18727 + }, + { + "epoch": 0.18728, + "grad_norm": 1.3043756435412168, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 18728 + }, + { + "epoch": 0.18729, + "grad_norm": 1.0378976127669879, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 18729 + }, + { + "epoch": 0.1873, + "grad_norm": 1.7193087087203027, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 18730 + }, + { + "epoch": 0.18731, + "grad_norm": 1.116419553335292, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 18731 + }, + { + "epoch": 0.18732, + "grad_norm": 1.3772300591436983, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 18732 + }, + { + "epoch": 0.18733, + "grad_norm": 1.4544935001798796, + "learning_rate": 0.003, + "loss": 4.012, + "step": 18733 + }, + { + "epoch": 0.18734, + "grad_norm": 1.3162098252868233, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 18734 + }, + { + "epoch": 0.18735, + "grad_norm": 1.1633225815507455, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 18735 + }, + { + "epoch": 0.18736, + "grad_norm": 1.5858181808219405, + "learning_rate": 0.003, + "loss": 4.027, + "step": 18736 + }, + { + "epoch": 0.18737, + "grad_norm": 1.149873887174443, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 18737 + }, + { + "epoch": 0.18738, + "grad_norm": 1.358716025247225, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 18738 + }, + { + "epoch": 0.18739, + "grad_norm": 1.091467221483659, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 18739 + }, + { + "epoch": 0.1874, + "grad_norm": 1.4703193300528377, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 18740 + }, + { + "epoch": 0.18741, + "grad_norm": 1.098064255985667, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 18741 + }, + { + "epoch": 0.18742, + "grad_norm": 1.409983007857012, + "learning_rate": 0.003, + "loss": 4.006, + "step": 18742 + }, + { + "epoch": 0.18743, + "grad_norm": 1.2106177297652705, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 18743 + }, + { + "epoch": 0.18744, + "grad_norm": 1.2815108498530823, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 18744 + }, + { + "epoch": 0.18745, + "grad_norm": 1.1932920186130076, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 18745 + }, + { + "epoch": 0.18746, + "grad_norm": 1.2752691820517976, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 18746 + }, + { + "epoch": 0.18747, + "grad_norm": 1.417029736211795, + "learning_rate": 0.003, + "loss": 4.04, + "step": 18747 + }, + { + "epoch": 0.18748, + "grad_norm": 1.1874280280107192, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 18748 + }, + { + "epoch": 0.18749, + "grad_norm": 1.4466801190568057, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 18749 + }, + { + "epoch": 0.1875, + "grad_norm": 1.128243469158432, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 18750 + }, + { + "epoch": 0.18751, + "grad_norm": 1.3755903593701546, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18751 + }, + { + "epoch": 0.18752, + "grad_norm": 1.286503093271871, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18752 + }, + { + "epoch": 0.18753, + "grad_norm": 1.2690163012949174, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 18753 + }, + { + "epoch": 0.18754, + "grad_norm": 1.3603536497722082, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 18754 + }, + { + "epoch": 0.18755, + "grad_norm": 1.4681285132760695, + "learning_rate": 0.003, + "loss": 3.992, + "step": 18755 + }, + { + "epoch": 0.18756, + "grad_norm": 1.1572490629926064, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 18756 + }, + { + "epoch": 0.18757, + "grad_norm": 1.2509963895410787, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18757 + }, + { + "epoch": 0.18758, + "grad_norm": 1.0698688906875218, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 18758 + }, + { + "epoch": 0.18759, + "grad_norm": 1.2782355106290009, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 18759 + }, + { + "epoch": 0.1876, + "grad_norm": 1.0630946077760033, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 18760 + }, + { + "epoch": 0.18761, + "grad_norm": 1.5442058962361622, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18761 + }, + { + "epoch": 0.18762, + "grad_norm": 1.1588513781519307, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 18762 + }, + { + "epoch": 0.18763, + "grad_norm": 1.39154666361133, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18763 + }, + { + "epoch": 0.18764, + "grad_norm": 1.2438414106731435, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 18764 + }, + { + "epoch": 0.18765, + "grad_norm": 1.233326266286673, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 18765 + }, + { + "epoch": 0.18766, + "grad_norm": 1.5246131084850232, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 18766 + }, + { + "epoch": 0.18767, + "grad_norm": 1.1164940103492973, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18767 + }, + { + "epoch": 0.18768, + "grad_norm": 1.2732871741442202, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 18768 + }, + { + "epoch": 0.18769, + "grad_norm": 1.3576031831596136, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 18769 + }, + { + "epoch": 0.1877, + "grad_norm": 1.4855913600576, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 18770 + }, + { + "epoch": 0.18771, + "grad_norm": 1.0985197522285455, + "learning_rate": 0.003, + "loss": 4.023, + "step": 18771 + }, + { + "epoch": 0.18772, + "grad_norm": 1.292770457214615, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 18772 + }, + { + "epoch": 0.18773, + "grad_norm": 1.1431768628823065, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 18773 + }, + { + "epoch": 0.18774, + "grad_norm": 1.3401269635220132, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 18774 + }, + { + "epoch": 0.18775, + "grad_norm": 1.4440907767201874, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 18775 + }, + { + "epoch": 0.18776, + "grad_norm": 1.4016436996853219, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 18776 + }, + { + "epoch": 0.18777, + "grad_norm": 1.1107532056560023, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 18777 + }, + { + "epoch": 0.18778, + "grad_norm": 1.4163198512437702, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 18778 + }, + { + "epoch": 0.18779, + "grad_norm": 1.1721215333974473, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 18779 + }, + { + "epoch": 0.1878, + "grad_norm": 1.278257339205773, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 18780 + }, + { + "epoch": 0.18781, + "grad_norm": 1.3053228746542866, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 18781 + }, + { + "epoch": 0.18782, + "grad_norm": 1.1375049537145128, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 18782 + }, + { + "epoch": 0.18783, + "grad_norm": 1.2513112953606718, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 18783 + }, + { + "epoch": 0.18784, + "grad_norm": 1.1693247682321866, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 18784 + }, + { + "epoch": 0.18785, + "grad_norm": 1.3205673873811723, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 18785 + }, + { + "epoch": 0.18786, + "grad_norm": 1.1292968914157666, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 18786 + }, + { + "epoch": 0.18787, + "grad_norm": 1.4473043584109355, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 18787 + }, + { + "epoch": 0.18788, + "grad_norm": 1.0521059265770933, + "learning_rate": 0.003, + "loss": 4.008, + "step": 18788 + }, + { + "epoch": 0.18789, + "grad_norm": 1.4731019996158254, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 18789 + }, + { + "epoch": 0.1879, + "grad_norm": 1.1114846782522791, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 18790 + }, + { + "epoch": 0.18791, + "grad_norm": 1.3651276065078815, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 18791 + }, + { + "epoch": 0.18792, + "grad_norm": 1.093525883067943, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18792 + }, + { + "epoch": 0.18793, + "grad_norm": 1.2688393406196965, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 18793 + }, + { + "epoch": 0.18794, + "grad_norm": 1.3340696193587918, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 18794 + }, + { + "epoch": 0.18795, + "grad_norm": 1.3630128950756113, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 18795 + }, + { + "epoch": 0.18796, + "grad_norm": 1.212412852205088, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 18796 + }, + { + "epoch": 0.18797, + "grad_norm": 1.4512811090790925, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18797 + }, + { + "epoch": 0.18798, + "grad_norm": 1.0094139765105765, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 18798 + }, + { + "epoch": 0.18799, + "grad_norm": 1.6586776547444237, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 18799 + }, + { + "epoch": 0.188, + "grad_norm": 1.0297074310474934, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 18800 + }, + { + "epoch": 0.18801, + "grad_norm": 1.734932585171164, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18801 + }, + { + "epoch": 0.18802, + "grad_norm": 1.0348704782328961, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 18802 + }, + { + "epoch": 0.18803, + "grad_norm": 1.1504415390030405, + "learning_rate": 0.003, + "loss": 4.012, + "step": 18803 + }, + { + "epoch": 0.18804, + "grad_norm": 1.389416151087945, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 18804 + }, + { + "epoch": 0.18805, + "grad_norm": 1.3290382566790426, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 18805 + }, + { + "epoch": 0.18806, + "grad_norm": 1.3859814992056898, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 18806 + }, + { + "epoch": 0.18807, + "grad_norm": 1.3702533713002063, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 18807 + }, + { + "epoch": 0.18808, + "grad_norm": 1.2715168233177794, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 18808 + }, + { + "epoch": 0.18809, + "grad_norm": 1.2704671484468457, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 18809 + }, + { + "epoch": 0.1881, + "grad_norm": 1.3509670439146595, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 18810 + }, + { + "epoch": 0.18811, + "grad_norm": 1.2373546477645962, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 18811 + }, + { + "epoch": 0.18812, + "grad_norm": 1.4410704719575569, + "learning_rate": 0.003, + "loss": 3.987, + "step": 18812 + }, + { + "epoch": 0.18813, + "grad_norm": 1.229364311713738, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 18813 + }, + { + "epoch": 0.18814, + "grad_norm": 1.4114113896071405, + "learning_rate": 0.003, + "loss": 4.002, + "step": 18814 + }, + { + "epoch": 0.18815, + "grad_norm": 1.3822494345951508, + "learning_rate": 0.003, + "loss": 4.018, + "step": 18815 + }, + { + "epoch": 0.18816, + "grad_norm": 1.3748248890074133, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 18816 + }, + { + "epoch": 0.18817, + "grad_norm": 1.1973746701862775, + "learning_rate": 0.003, + "loss": 4.047, + "step": 18817 + }, + { + "epoch": 0.18818, + "grad_norm": 1.362257965096609, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 18818 + }, + { + "epoch": 0.18819, + "grad_norm": 1.0156086218524036, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 18819 + }, + { + "epoch": 0.1882, + "grad_norm": 1.2566168851036394, + "learning_rate": 0.003, + "loss": 3.999, + "step": 18820 + }, + { + "epoch": 0.18821, + "grad_norm": 1.5324163598981786, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 18821 + }, + { + "epoch": 0.18822, + "grad_norm": 1.2512580689858643, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 18822 + }, + { + "epoch": 0.18823, + "grad_norm": 1.2935741105961847, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 18823 + }, + { + "epoch": 0.18824, + "grad_norm": 1.1644334936516052, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 18824 + }, + { + "epoch": 0.18825, + "grad_norm": 1.3761851337505675, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 18825 + }, + { + "epoch": 0.18826, + "grad_norm": 1.0995130171261787, + "learning_rate": 0.003, + "loss": 4.003, + "step": 18826 + }, + { + "epoch": 0.18827, + "grad_norm": 1.443020168232223, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 18827 + }, + { + "epoch": 0.18828, + "grad_norm": 1.1282469620944757, + "learning_rate": 0.003, + "loss": 4.0, + "step": 18828 + }, + { + "epoch": 0.18829, + "grad_norm": 1.323305598272999, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18829 + }, + { + "epoch": 0.1883, + "grad_norm": 1.1828572268105, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 18830 + }, + { + "epoch": 0.18831, + "grad_norm": 1.4276740531309458, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 18831 + }, + { + "epoch": 0.18832, + "grad_norm": 1.184544184111074, + "learning_rate": 0.003, + "loss": 4.026, + "step": 18832 + }, + { + "epoch": 0.18833, + "grad_norm": 1.4822322926422464, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18833 + }, + { + "epoch": 0.18834, + "grad_norm": 1.3143441193176422, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 18834 + }, + { + "epoch": 0.18835, + "grad_norm": 1.254993659017894, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 18835 + }, + { + "epoch": 0.18836, + "grad_norm": 1.1184698641428459, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 18836 + }, + { + "epoch": 0.18837, + "grad_norm": 1.2404210698345246, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 18837 + }, + { + "epoch": 0.18838, + "grad_norm": 1.1070446605113913, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 18838 + }, + { + "epoch": 0.18839, + "grad_norm": 1.4704983128492852, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 18839 + }, + { + "epoch": 0.1884, + "grad_norm": 1.3008673418840326, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 18840 + }, + { + "epoch": 0.18841, + "grad_norm": 1.2595207232349506, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 18841 + }, + { + "epoch": 0.18842, + "grad_norm": 1.342808067808002, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 18842 + }, + { + "epoch": 0.18843, + "grad_norm": 0.9922458732256416, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18843 + }, + { + "epoch": 0.18844, + "grad_norm": 1.445638244499319, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 18844 + }, + { + "epoch": 0.18845, + "grad_norm": 0.9789811370228675, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18845 + }, + { + "epoch": 0.18846, + "grad_norm": 1.5034543902524964, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 18846 + }, + { + "epoch": 0.18847, + "grad_norm": 0.927876524239201, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 18847 + }, + { + "epoch": 0.18848, + "grad_norm": 1.3296711416190021, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 18848 + }, + { + "epoch": 0.18849, + "grad_norm": 1.4641733211136532, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 18849 + }, + { + "epoch": 0.1885, + "grad_norm": 1.2706728987678715, + "learning_rate": 0.003, + "loss": 4.014, + "step": 18850 + }, + { + "epoch": 0.18851, + "grad_norm": 1.37942646887191, + "learning_rate": 0.003, + "loss": 4.016, + "step": 18851 + }, + { + "epoch": 0.18852, + "grad_norm": 1.4416990193136976, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 18852 + }, + { + "epoch": 0.18853, + "grad_norm": 1.2494013061938976, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 18853 + }, + { + "epoch": 0.18854, + "grad_norm": 1.4870346591138417, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 18854 + }, + { + "epoch": 0.18855, + "grad_norm": 1.2022339227850636, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 18855 + }, + { + "epoch": 0.18856, + "grad_norm": 1.3724267696002113, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 18856 + }, + { + "epoch": 0.18857, + "grad_norm": 1.1434147235476109, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 18857 + }, + { + "epoch": 0.18858, + "grad_norm": 1.3253368873665614, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 18858 + }, + { + "epoch": 0.18859, + "grad_norm": 1.3516363221601224, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 18859 + }, + { + "epoch": 0.1886, + "grad_norm": 1.21235742171777, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 18860 + }, + { + "epoch": 0.18861, + "grad_norm": 1.3597598477739947, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 18861 + }, + { + "epoch": 0.18862, + "grad_norm": 1.2549641906031592, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 18862 + }, + { + "epoch": 0.18863, + "grad_norm": 1.429341267930262, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 18863 + }, + { + "epoch": 0.18864, + "grad_norm": 1.0721288573102947, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 18864 + }, + { + "epoch": 0.18865, + "grad_norm": 1.4264240127264443, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 18865 + }, + { + "epoch": 0.18866, + "grad_norm": 1.1983018274161035, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18866 + }, + { + "epoch": 0.18867, + "grad_norm": 1.350228467376567, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 18867 + }, + { + "epoch": 0.18868, + "grad_norm": 1.2367787437442568, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 18868 + }, + { + "epoch": 0.18869, + "grad_norm": 1.4888523853636724, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 18869 + }, + { + "epoch": 0.1887, + "grad_norm": 1.1358909114521822, + "learning_rate": 0.003, + "loss": 4.001, + "step": 18870 + }, + { + "epoch": 0.18871, + "grad_norm": 1.2907629612968026, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 18871 + }, + { + "epoch": 0.18872, + "grad_norm": 1.1005953395431796, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 18872 + }, + { + "epoch": 0.18873, + "grad_norm": 1.6345853186178532, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 18873 + }, + { + "epoch": 0.18874, + "grad_norm": 1.1746353686890867, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 18874 + }, + { + "epoch": 0.18875, + "grad_norm": 1.2573364312123834, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 18875 + }, + { + "epoch": 0.18876, + "grad_norm": 1.364399823656439, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 18876 + }, + { + "epoch": 0.18877, + "grad_norm": 1.211411391260069, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 18877 + }, + { + "epoch": 0.18878, + "grad_norm": 1.3657723068251055, + "learning_rate": 0.003, + "loss": 3.989, + "step": 18878 + }, + { + "epoch": 0.18879, + "grad_norm": 1.2149671592494593, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 18879 + }, + { + "epoch": 0.1888, + "grad_norm": 1.3051352574687405, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18880 + }, + { + "epoch": 0.18881, + "grad_norm": 1.1329075107385167, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 18881 + }, + { + "epoch": 0.18882, + "grad_norm": 1.304321722250987, + "learning_rate": 0.003, + "loss": 4.036, + "step": 18882 + }, + { + "epoch": 0.18883, + "grad_norm": 1.2529829747626335, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 18883 + }, + { + "epoch": 0.18884, + "grad_norm": 1.201355611042984, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 18884 + }, + { + "epoch": 0.18885, + "grad_norm": 1.3400792509794779, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 18885 + }, + { + "epoch": 0.18886, + "grad_norm": 1.3073238246591101, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 18886 + }, + { + "epoch": 0.18887, + "grad_norm": 1.1548865607783483, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 18887 + }, + { + "epoch": 0.18888, + "grad_norm": 1.2996683356200782, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 18888 + }, + { + "epoch": 0.18889, + "grad_norm": 1.2910113853304779, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 18889 + }, + { + "epoch": 0.1889, + "grad_norm": 1.392740912920975, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 18890 + }, + { + "epoch": 0.18891, + "grad_norm": 1.1319046905564893, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 18891 + }, + { + "epoch": 0.18892, + "grad_norm": 1.297455272784623, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 18892 + }, + { + "epoch": 0.18893, + "grad_norm": 1.1822205544925295, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 18893 + }, + { + "epoch": 0.18894, + "grad_norm": 1.3799401432733638, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 18894 + }, + { + "epoch": 0.18895, + "grad_norm": 1.3113136082152823, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 18895 + }, + { + "epoch": 0.18896, + "grad_norm": 1.114155835934784, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 18896 + }, + { + "epoch": 0.18897, + "grad_norm": 1.3355867118053277, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 18897 + }, + { + "epoch": 0.18898, + "grad_norm": 1.2254191046973921, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 18898 + }, + { + "epoch": 0.18899, + "grad_norm": 1.5779703946160442, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 18899 + }, + { + "epoch": 0.189, + "grad_norm": 1.2173062699404291, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 18900 + }, + { + "epoch": 0.18901, + "grad_norm": 1.2936130368839325, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 18901 + }, + { + "epoch": 0.18902, + "grad_norm": 1.2887485223488224, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 18902 + }, + { + "epoch": 0.18903, + "grad_norm": 1.2893185921850154, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 18903 + }, + { + "epoch": 0.18904, + "grad_norm": 1.3204077224346007, + "learning_rate": 0.003, + "loss": 4.036, + "step": 18904 + }, + { + "epoch": 0.18905, + "grad_norm": 1.376599778433685, + "learning_rate": 0.003, + "loss": 4.02, + "step": 18905 + }, + { + "epoch": 0.18906, + "grad_norm": 1.1842652767120574, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 18906 + }, + { + "epoch": 0.18907, + "grad_norm": 1.8752157539520744, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 18907 + }, + { + "epoch": 0.18908, + "grad_norm": 1.0844141954120532, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 18908 + }, + { + "epoch": 0.18909, + "grad_norm": 1.4494358850508267, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 18909 + }, + { + "epoch": 0.1891, + "grad_norm": 1.3815892662527047, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 18910 + }, + { + "epoch": 0.18911, + "grad_norm": 1.22776110818462, + "learning_rate": 0.003, + "loss": 4.046, + "step": 18911 + }, + { + "epoch": 0.18912, + "grad_norm": 1.2546354098864025, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 18912 + }, + { + "epoch": 0.18913, + "grad_norm": 1.4322587831332994, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18913 + }, + { + "epoch": 0.18914, + "grad_norm": 1.4357672228376845, + "learning_rate": 0.003, + "loss": 4.028, + "step": 18914 + }, + { + "epoch": 0.18915, + "grad_norm": 1.381660269032073, + "learning_rate": 0.003, + "loss": 3.996, + "step": 18915 + }, + { + "epoch": 0.18916, + "grad_norm": 1.2120151654235733, + "learning_rate": 0.003, + "loss": 4.019, + "step": 18916 + }, + { + "epoch": 0.18917, + "grad_norm": 1.155511029221399, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 18917 + }, + { + "epoch": 0.18918, + "grad_norm": 1.3798380706777866, + "learning_rate": 0.003, + "loss": 3.996, + "step": 18918 + }, + { + "epoch": 0.18919, + "grad_norm": 1.2430163913873487, + "learning_rate": 0.003, + "loss": 4.035, + "step": 18919 + }, + { + "epoch": 0.1892, + "grad_norm": 1.2632121393382478, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 18920 + }, + { + "epoch": 0.18921, + "grad_norm": 1.2240506042274173, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 18921 + }, + { + "epoch": 0.18922, + "grad_norm": 1.5078155308240528, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 18922 + }, + { + "epoch": 0.18923, + "grad_norm": 1.0321468505232072, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 18923 + }, + { + "epoch": 0.18924, + "grad_norm": 1.3759820406243037, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 18924 + }, + { + "epoch": 0.18925, + "grad_norm": 1.1578600968645194, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 18925 + }, + { + "epoch": 0.18926, + "grad_norm": 1.4386004900604346, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 18926 + }, + { + "epoch": 0.18927, + "grad_norm": 1.1975391679472591, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 18927 + }, + { + "epoch": 0.18928, + "grad_norm": 1.4233703755004665, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 18928 + }, + { + "epoch": 0.18929, + "grad_norm": 1.3298361094853883, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 18929 + }, + { + "epoch": 0.1893, + "grad_norm": 1.1668141803513414, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18930 + }, + { + "epoch": 0.18931, + "grad_norm": 1.1927574278879163, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 18931 + }, + { + "epoch": 0.18932, + "grad_norm": 1.2278395725435298, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 18932 + }, + { + "epoch": 0.18933, + "grad_norm": 1.2719217916469485, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 18933 + }, + { + "epoch": 0.18934, + "grad_norm": 1.3688014376645494, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 18934 + }, + { + "epoch": 0.18935, + "grad_norm": 1.2959450429209367, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 18935 + }, + { + "epoch": 0.18936, + "grad_norm": 1.1930725451251745, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 18936 + }, + { + "epoch": 0.18937, + "grad_norm": 1.2407030124021434, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 18937 + }, + { + "epoch": 0.18938, + "grad_norm": 1.3330548919923813, + "learning_rate": 0.003, + "loss": 4.02, + "step": 18938 + }, + { + "epoch": 0.18939, + "grad_norm": 1.3082167721320919, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 18939 + }, + { + "epoch": 0.1894, + "grad_norm": 1.4748190566899932, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 18940 + }, + { + "epoch": 0.18941, + "grad_norm": 1.116748397977893, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 18941 + }, + { + "epoch": 0.18942, + "grad_norm": 1.4291352947153417, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 18942 + }, + { + "epoch": 0.18943, + "grad_norm": 1.1199615328961368, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 18943 + }, + { + "epoch": 0.18944, + "grad_norm": 1.5294642175166582, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 18944 + }, + { + "epoch": 0.18945, + "grad_norm": 1.0882092069192857, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 18945 + }, + { + "epoch": 0.18946, + "grad_norm": 1.6066215519903786, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 18946 + }, + { + "epoch": 0.18947, + "grad_norm": 1.2083091516202875, + "learning_rate": 0.003, + "loss": 4.001, + "step": 18947 + }, + { + "epoch": 0.18948, + "grad_norm": 1.1985481857292923, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 18948 + }, + { + "epoch": 0.18949, + "grad_norm": 1.2775459767805657, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 18949 + }, + { + "epoch": 0.1895, + "grad_norm": 1.298008618901343, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 18950 + }, + { + "epoch": 0.18951, + "grad_norm": 1.3213237601028365, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 18951 + }, + { + "epoch": 0.18952, + "grad_norm": 1.2355302765967384, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 18952 + }, + { + "epoch": 0.18953, + "grad_norm": 1.298428704219246, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18953 + }, + { + "epoch": 0.18954, + "grad_norm": 1.2595444817645047, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18954 + }, + { + "epoch": 0.18955, + "grad_norm": 1.4295919433866582, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 18955 + }, + { + "epoch": 0.18956, + "grad_norm": 1.050915980475837, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 18956 + }, + { + "epoch": 0.18957, + "grad_norm": 1.3569253053260681, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 18957 + }, + { + "epoch": 0.18958, + "grad_norm": 1.323218007787156, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 18958 + }, + { + "epoch": 0.18959, + "grad_norm": 1.1827405133434696, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 18959 + }, + { + "epoch": 0.1896, + "grad_norm": 1.249099258846444, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 18960 + }, + { + "epoch": 0.18961, + "grad_norm": 1.199208344986785, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 18961 + }, + { + "epoch": 0.18962, + "grad_norm": 1.2709152360429374, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 18962 + }, + { + "epoch": 0.18963, + "grad_norm": 1.1441255867778828, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 18963 + }, + { + "epoch": 0.18964, + "grad_norm": 1.2224256617719833, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 18964 + }, + { + "epoch": 0.18965, + "grad_norm": 1.2316616781902274, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 18965 + }, + { + "epoch": 0.18966, + "grad_norm": 1.3247909543506073, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 18966 + }, + { + "epoch": 0.18967, + "grad_norm": 1.0669828300995452, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 18967 + }, + { + "epoch": 0.18968, + "grad_norm": 1.6222343013067921, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 18968 + }, + { + "epoch": 0.18969, + "grad_norm": 1.179308640345885, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 18969 + }, + { + "epoch": 0.1897, + "grad_norm": 1.3939614509485425, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 18970 + }, + { + "epoch": 0.18971, + "grad_norm": 1.0954415561570288, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 18971 + }, + { + "epoch": 0.18972, + "grad_norm": 1.3422469479213295, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 18972 + }, + { + "epoch": 0.18973, + "grad_norm": 1.1209971142146853, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 18973 + }, + { + "epoch": 0.18974, + "grad_norm": 1.216262616098274, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 18974 + }, + { + "epoch": 0.18975, + "grad_norm": 1.2521601171024372, + "learning_rate": 0.003, + "loss": 4.005, + "step": 18975 + }, + { + "epoch": 0.18976, + "grad_norm": 1.0936838393827564, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 18976 + }, + { + "epoch": 0.18977, + "grad_norm": 1.4222352337005453, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 18977 + }, + { + "epoch": 0.18978, + "grad_norm": 1.1516237804968388, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 18978 + }, + { + "epoch": 0.18979, + "grad_norm": 1.4556857373688754, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 18979 + }, + { + "epoch": 0.1898, + "grad_norm": 1.0075365323494918, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18980 + }, + { + "epoch": 0.18981, + "grad_norm": 1.4602101366952436, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18981 + }, + { + "epoch": 0.18982, + "grad_norm": 1.0574097890207579, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 18982 + }, + { + "epoch": 0.18983, + "grad_norm": 1.7161545080248717, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 18983 + }, + { + "epoch": 0.18984, + "grad_norm": 0.9621329220924494, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 18984 + }, + { + "epoch": 0.18985, + "grad_norm": 1.5805646455801179, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 18985 + }, + { + "epoch": 0.18986, + "grad_norm": 1.2399912292567599, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 18986 + }, + { + "epoch": 0.18987, + "grad_norm": 1.3783649312622117, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 18987 + }, + { + "epoch": 0.18988, + "grad_norm": 1.498681352981351, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 18988 + }, + { + "epoch": 0.18989, + "grad_norm": 1.3807216236031543, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 18989 + }, + { + "epoch": 0.1899, + "grad_norm": 1.375538414014236, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 18990 + }, + { + "epoch": 0.18991, + "grad_norm": 1.2420049603098005, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 18991 + }, + { + "epoch": 0.18992, + "grad_norm": 1.2223551716105183, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 18992 + }, + { + "epoch": 0.18993, + "grad_norm": 1.2570012470085465, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 18993 + }, + { + "epoch": 0.18994, + "grad_norm": 1.4628515256064998, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 18994 + }, + { + "epoch": 0.18995, + "grad_norm": 0.9789874467067208, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 18995 + }, + { + "epoch": 0.18996, + "grad_norm": 1.1670502812183778, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 18996 + }, + { + "epoch": 0.18997, + "grad_norm": 1.3794493322023509, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 18997 + }, + { + "epoch": 0.18998, + "grad_norm": 1.2742959001740786, + "learning_rate": 0.003, + "loss": 4.026, + "step": 18998 + }, + { + "epoch": 0.18999, + "grad_norm": 1.149930266244774, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 18999 + }, + { + "epoch": 0.19, + "grad_norm": 1.3926689602735243, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 19000 + }, + { + "epoch": 0.19001, + "grad_norm": 1.3281844354153476, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 19001 + }, + { + "epoch": 0.19002, + "grad_norm": 1.3399837271544746, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 19002 + }, + { + "epoch": 0.19003, + "grad_norm": 1.151917376655648, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 19003 + }, + { + "epoch": 0.19004, + "grad_norm": 1.402780168163471, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 19004 + }, + { + "epoch": 0.19005, + "grad_norm": 1.1601986147353196, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 19005 + }, + { + "epoch": 0.19006, + "grad_norm": 1.4341768719680854, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 19006 + }, + { + "epoch": 0.19007, + "grad_norm": 1.2738095860949428, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 19007 + }, + { + "epoch": 0.19008, + "grad_norm": 1.4703448321356738, + "learning_rate": 0.003, + "loss": 4.005, + "step": 19008 + }, + { + "epoch": 0.19009, + "grad_norm": 1.2759772286311846, + "learning_rate": 0.003, + "loss": 4.03, + "step": 19009 + }, + { + "epoch": 0.1901, + "grad_norm": 1.4164090562929703, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 19010 + }, + { + "epoch": 0.19011, + "grad_norm": 1.2514843916867464, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 19011 + }, + { + "epoch": 0.19012, + "grad_norm": 1.1829876887961355, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 19012 + }, + { + "epoch": 0.19013, + "grad_norm": 1.365541613847154, + "learning_rate": 0.003, + "loss": 3.99, + "step": 19013 + }, + { + "epoch": 0.19014, + "grad_norm": 1.3682420465503087, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 19014 + }, + { + "epoch": 0.19015, + "grad_norm": 1.2157757363541761, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 19015 + }, + { + "epoch": 0.19016, + "grad_norm": 1.30835776770585, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 19016 + }, + { + "epoch": 0.19017, + "grad_norm": 1.2049003231229503, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 19017 + }, + { + "epoch": 0.19018, + "grad_norm": 1.1741936561386186, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 19018 + }, + { + "epoch": 0.19019, + "grad_norm": 1.0853739123496602, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 19019 + }, + { + "epoch": 0.1902, + "grad_norm": 1.2708312439016338, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 19020 + }, + { + "epoch": 0.19021, + "grad_norm": 1.278559727595023, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 19021 + }, + { + "epoch": 0.19022, + "grad_norm": 1.4935703188671352, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 19022 + }, + { + "epoch": 0.19023, + "grad_norm": 1.3448299722365684, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 19023 + }, + { + "epoch": 0.19024, + "grad_norm": 0.9982529244950282, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 19024 + }, + { + "epoch": 0.19025, + "grad_norm": 1.1850540718595843, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 19025 + } + ], + "logging_steps": 1, + "max_steps": 100000, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 25, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 7.542203053244416e+17, + "train_batch_size": 256, + "trial_name": null, + "trial_params": null +}