| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.960544217687075, | |
| "eval_steps": 500, | |
| "global_step": 455, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010884353741496598, | |
| "grad_norm": 5.927079454035689, | |
| "learning_rate": 1.7391304347826088e-06, | |
| "loss": 0.872, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.021768707482993196, | |
| "grad_norm": 5.806903403372398, | |
| "learning_rate": 3.4782608695652175e-06, | |
| "loss": 0.851, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0326530612244898, | |
| "grad_norm": 5.7321763520118445, | |
| "learning_rate": 5.2173913043478265e-06, | |
| "loss": 0.8568, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04353741496598639, | |
| "grad_norm": 4.307089386101637, | |
| "learning_rate": 6.956521739130435e-06, | |
| "loss": 0.8166, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05442176870748299, | |
| "grad_norm": 2.319136655665802, | |
| "learning_rate": 8.695652173913044e-06, | |
| "loss": 0.7738, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0653061224489796, | |
| "grad_norm": 4.138608265566134, | |
| "learning_rate": 1.0434782608695653e-05, | |
| "loss": 0.778, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0761904761904762, | |
| "grad_norm": 4.3627207825365435, | |
| "learning_rate": 1.2173913043478263e-05, | |
| "loss": 0.7837, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08707482993197278, | |
| "grad_norm": 4.353495294571602, | |
| "learning_rate": 1.391304347826087e-05, | |
| "loss": 0.7412, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09795918367346938, | |
| "grad_norm": 3.776888211990866, | |
| "learning_rate": 1.565217391304348e-05, | |
| "loss": 0.7248, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.10884353741496598, | |
| "grad_norm": 2.2362554644133437, | |
| "learning_rate": 1.739130434782609e-05, | |
| "loss": 0.7153, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11972789115646258, | |
| "grad_norm": 2.346223941705739, | |
| "learning_rate": 1.9130434782608697e-05, | |
| "loss": 0.6763, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1306122448979592, | |
| "grad_norm": 1.8153540015951113, | |
| "learning_rate": 2.0869565217391306e-05, | |
| "loss": 0.6582, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1414965986394558, | |
| "grad_norm": 1.39370325247011, | |
| "learning_rate": 2.2608695652173914e-05, | |
| "loss": 0.6376, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1523809523809524, | |
| "grad_norm": 1.3214924895742388, | |
| "learning_rate": 2.4347826086956526e-05, | |
| "loss": 0.6266, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16326530612244897, | |
| "grad_norm": 1.2011292969457532, | |
| "learning_rate": 2.608695652173913e-05, | |
| "loss": 0.6091, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17414965986394557, | |
| "grad_norm": 0.9037834201567392, | |
| "learning_rate": 2.782608695652174e-05, | |
| "loss": 0.5957, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.18503401360544217, | |
| "grad_norm": 1.0485654993423914, | |
| "learning_rate": 2.956521739130435e-05, | |
| "loss": 0.6085, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.19591836734693877, | |
| "grad_norm": 1.0812897697040533, | |
| "learning_rate": 3.130434782608696e-05, | |
| "loss": 0.6002, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.20680272108843537, | |
| "grad_norm": 1.0775886304189544, | |
| "learning_rate": 3.304347826086957e-05, | |
| "loss": 0.585, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.21768707482993196, | |
| "grad_norm": 1.1157944244523579, | |
| "learning_rate": 3.478260869565218e-05, | |
| "loss": 0.573, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 1.4190920261301871, | |
| "learning_rate": 3.6521739130434786e-05, | |
| "loss": 0.5681, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.23945578231292516, | |
| "grad_norm": 0.9518352118682741, | |
| "learning_rate": 3.8260869565217395e-05, | |
| "loss": 0.5756, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2503401360544218, | |
| "grad_norm": 1.2554537506564052, | |
| "learning_rate": 4e-05, | |
| "loss": 0.5697, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2612244897959184, | |
| "grad_norm": 1.4638431122192201, | |
| "learning_rate": 4.173913043478261e-05, | |
| "loss": 0.5648, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.272108843537415, | |
| "grad_norm": 1.1275694213145084, | |
| "learning_rate": 4.347826086956522e-05, | |
| "loss": 0.563, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2829931972789116, | |
| "grad_norm": 1.5518574639376193, | |
| "learning_rate": 4.521739130434783e-05, | |
| "loss": 0.5537, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2938775510204082, | |
| "grad_norm": 1.1930318351154805, | |
| "learning_rate": 4.6956521739130444e-05, | |
| "loss": 0.5454, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3047619047619048, | |
| "grad_norm": 4.8505593988497235, | |
| "learning_rate": 4.869565217391305e-05, | |
| "loss": 0.5565, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.31564625850340133, | |
| "grad_norm": 1.7409964222339103, | |
| "learning_rate": 5.043478260869566e-05, | |
| "loss": 0.562, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.32653061224489793, | |
| "grad_norm": 1.0848266670999587, | |
| "learning_rate": 5.217391304347826e-05, | |
| "loss": 0.5407, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.33741496598639453, | |
| "grad_norm": 2.04059438376003, | |
| "learning_rate": 5.391304347826087e-05, | |
| "loss": 0.5428, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.34829931972789113, | |
| "grad_norm": 1.5450744349828915, | |
| "learning_rate": 5.565217391304348e-05, | |
| "loss": 0.5485, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.35918367346938773, | |
| "grad_norm": 1.920362365505494, | |
| "learning_rate": 5.739130434782609e-05, | |
| "loss": 0.525, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.37006802721088433, | |
| "grad_norm": 1.6037947711043383, | |
| "learning_rate": 5.91304347826087e-05, | |
| "loss": 0.5426, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 1.4766398465058153, | |
| "learning_rate": 6.086956521739131e-05, | |
| "loss": 0.5331, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.39183673469387753, | |
| "grad_norm": 1.551431177810188, | |
| "learning_rate": 6.260869565217392e-05, | |
| "loss": 0.5279, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.40272108843537413, | |
| "grad_norm": 1.3917411457191442, | |
| "learning_rate": 6.434782608695653e-05, | |
| "loss": 0.527, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.41360544217687073, | |
| "grad_norm": 2.182840299170343, | |
| "learning_rate": 6.608695652173914e-05, | |
| "loss": 0.5276, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.42448979591836733, | |
| "grad_norm": 1.1726558659517503, | |
| "learning_rate": 6.782608695652175e-05, | |
| "loss": 0.5232, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.43537414965986393, | |
| "grad_norm": 1.4603138366721788, | |
| "learning_rate": 6.956521739130436e-05, | |
| "loss": 0.529, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.44625850340136053, | |
| "grad_norm": 2.030276251063391, | |
| "learning_rate": 7.130434782608696e-05, | |
| "loss": 0.5286, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 1.2752316072355578, | |
| "learning_rate": 7.304347826086957e-05, | |
| "loss": 0.5302, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.46802721088435373, | |
| "grad_norm": 2.212305737759694, | |
| "learning_rate": 7.478260869565218e-05, | |
| "loss": 0.5239, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.47891156462585033, | |
| "grad_norm": 1.1961315830459824, | |
| "learning_rate": 7.652173913043479e-05, | |
| "loss": 0.5279, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4897959183673469, | |
| "grad_norm": 2.192201970865223, | |
| "learning_rate": 7.82608695652174e-05, | |
| "loss": 0.5241, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5006802721088436, | |
| "grad_norm": 1.8726468593248957, | |
| "learning_rate": 8e-05, | |
| "loss": 0.5249, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5115646258503401, | |
| "grad_norm": 1.6118424077468794, | |
| "learning_rate": 7.999882000276476e-05, | |
| "loss": 0.5154, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5224489795918368, | |
| "grad_norm": 1.7151790808569165, | |
| "learning_rate": 7.999528008067867e-05, | |
| "loss": 0.5152, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 1.5855695648608914, | |
| "learning_rate": 7.998938044259668e-05, | |
| "loss": 0.5085, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.54421768707483, | |
| "grad_norm": 2.2200684809852604, | |
| "learning_rate": 7.99811214365966e-05, | |
| "loss": 0.5121, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5551020408163265, | |
| "grad_norm": 1.318341921962779, | |
| "learning_rate": 7.997050354995864e-05, | |
| "loss": 0.5095, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5659863945578232, | |
| "grad_norm": 1.1834973619394002, | |
| "learning_rate": 7.995752740913666e-05, | |
| "loss": 0.5207, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5768707482993197, | |
| "grad_norm": 2.889598937166816, | |
| "learning_rate": 7.994219377972118e-05, | |
| "loss": 0.5321, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5877551020408164, | |
| "grad_norm": 1.5105048502502088, | |
| "learning_rate": 7.992450356639418e-05, | |
| "loss": 0.522, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5986394557823129, | |
| "grad_norm": 1.8379347638671795, | |
| "learning_rate": 7.990445781287584e-05, | |
| "loss": 0.5303, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6095238095238096, | |
| "grad_norm": 1.7843645911160906, | |
| "learning_rate": 7.988205770186282e-05, | |
| "loss": 0.5228, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6204081632653061, | |
| "grad_norm": 1.5019247336992891, | |
| "learning_rate": 7.985730455495858e-05, | |
| "loss": 0.5102, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6312925170068027, | |
| "grad_norm": 1.1320725209421107, | |
| "learning_rate": 7.983019983259538e-05, | |
| "loss": 0.5102, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.6421768707482993, | |
| "grad_norm": 1.7581849217011933, | |
| "learning_rate": 7.980074513394808e-05, | |
| "loss": 0.5295, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6530612244897959, | |
| "grad_norm": 1.8510775829530046, | |
| "learning_rate": 7.976894219683982e-05, | |
| "loss": 0.5171, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6639455782312925, | |
| "grad_norm": 1.5406885206700403, | |
| "learning_rate": 7.97347928976395e-05, | |
| "loss": 0.5125, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6748299319727891, | |
| "grad_norm": 1.8552250917689803, | |
| "learning_rate": 7.969829925115106e-05, | |
| "loss": 0.5046, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 1.3392814085799072, | |
| "learning_rate": 7.96594634104946e-05, | |
| "loss": 0.5135, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6965986394557823, | |
| "grad_norm": 1.3362402388629475, | |
| "learning_rate": 7.961828766697934e-05, | |
| "loss": 0.4996, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7074829931972789, | |
| "grad_norm": 1.2428849483779671, | |
| "learning_rate": 7.957477444996846e-05, | |
| "loss": 0.4991, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7183673469387755, | |
| "grad_norm": 1.2465637954977793, | |
| "learning_rate": 7.952892632673575e-05, | |
| "loss": 0.5142, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7292517006802721, | |
| "grad_norm": 1.9083296305012454, | |
| "learning_rate": 7.948074600231413e-05, | |
| "loss": 0.4975, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7401360544217687, | |
| "grad_norm": 1.020460731568505, | |
| "learning_rate": 7.943023631933609e-05, | |
| "loss": 0.5041, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7510204081632653, | |
| "grad_norm": 1.415619445739188, | |
| "learning_rate": 7.937740025786595e-05, | |
| "loss": 0.5027, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 1.0076405770877774, | |
| "learning_rate": 7.932224093522403e-05, | |
| "loss": 0.4919, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7727891156462585, | |
| "grad_norm": 1.3215495361872542, | |
| "learning_rate": 7.926476160580272e-05, | |
| "loss": 0.496, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7836734693877551, | |
| "grad_norm": 1.1493698738598546, | |
| "learning_rate": 7.920496566087454e-05, | |
| "loss": 0.4928, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7945578231292517, | |
| "grad_norm": 1.5318966998644297, | |
| "learning_rate": 7.914285662839197e-05, | |
| "loss": 0.4992, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8054421768707483, | |
| "grad_norm": 1.5046951974745066, | |
| "learning_rate": 7.907843817277931e-05, | |
| "loss": 0.4873, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 1.0505258604862309, | |
| "learning_rate": 7.901171409471658e-05, | |
| "loss": 0.4932, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8272108843537415, | |
| "grad_norm": 1.3786476967693353, | |
| "learning_rate": 7.894268833091514e-05, | |
| "loss": 0.4971, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8380952380952381, | |
| "grad_norm": 1.844081438002858, | |
| "learning_rate": 7.887136495388551e-05, | |
| "loss": 0.497, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8489795918367347, | |
| "grad_norm": 1.156330366068553, | |
| "learning_rate": 7.879774817169707e-05, | |
| "loss": 0.4911, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8598639455782313, | |
| "grad_norm": 1.9423299271570273, | |
| "learning_rate": 7.872184232772979e-05, | |
| "loss": 0.5038, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8707482993197279, | |
| "grad_norm": 1.4453495603978352, | |
| "learning_rate": 7.8643651900418e-05, | |
| "loss": 0.4996, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8816326530612245, | |
| "grad_norm": 1.478285837484742, | |
| "learning_rate": 7.856318150298608e-05, | |
| "loss": 0.4929, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8925170068027211, | |
| "grad_norm": 1.1763036783402252, | |
| "learning_rate": 7.848043588317636e-05, | |
| "loss": 0.4806, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.9034013605442177, | |
| "grad_norm": 1.5197620561189824, | |
| "learning_rate": 7.839541992296897e-05, | |
| "loss": 0.4913, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 1.0609355821635684, | |
| "learning_rate": 7.830813863829381e-05, | |
| "loss": 0.4874, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9251700680272109, | |
| "grad_norm": 1.9669301612012153, | |
| "learning_rate": 7.82185971787346e-05, | |
| "loss": 0.4952, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9360544217687075, | |
| "grad_norm": 1.5613138260325288, | |
| "learning_rate": 7.81268008272251e-05, | |
| "loss": 0.4925, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9469387755102041, | |
| "grad_norm": 1.5790694267763936, | |
| "learning_rate": 7.803275499973733e-05, | |
| "loss": 0.4878, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9578231292517007, | |
| "grad_norm": 1.3970443446424614, | |
| "learning_rate": 7.793646524496214e-05, | |
| "loss": 0.4907, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9687074829931973, | |
| "grad_norm": 1.1320202431292419, | |
| "learning_rate": 7.783793724398172e-05, | |
| "loss": 0.4875, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.9795918367346939, | |
| "grad_norm": 1.4689184389227818, | |
| "learning_rate": 7.773717680993454e-05, | |
| "loss": 0.4929, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9904761904761905, | |
| "grad_norm": 0.7295228879372475, | |
| "learning_rate": 7.763418988767225e-05, | |
| "loss": 0.4822, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.0034013605442176, | |
| "grad_norm": 1.5850833195566263, | |
| "learning_rate": 7.752898255340904e-05, | |
| "loss": 0.6339, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.0142857142857142, | |
| "grad_norm": 0.8809505538309929, | |
| "learning_rate": 7.74215610143631e-05, | |
| "loss": 0.4638, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.025170068027211, | |
| "grad_norm": 0.8138907351455011, | |
| "learning_rate": 7.731193160839039e-05, | |
| "loss": 0.4646, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0360544217687075, | |
| "grad_norm": 1.3421641089630965, | |
| "learning_rate": 7.720010080361066e-05, | |
| "loss": 0.4617, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.046938775510204, | |
| "grad_norm": 0.9699314431631867, | |
| "learning_rate": 7.708607519802597e-05, | |
| "loss": 0.4529, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0578231292517006, | |
| "grad_norm": 1.3830035493958794, | |
| "learning_rate": 7.696986151913129e-05, | |
| "loss": 0.4677, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.0687074829931973, | |
| "grad_norm": 1.3132389090374292, | |
| "learning_rate": 7.685146662351759e-05, | |
| "loss": 0.4531, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.079591836734694, | |
| "grad_norm": 1.4791897260832974, | |
| "learning_rate": 7.673089749646735e-05, | |
| "loss": 0.4608, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0904761904761904, | |
| "grad_norm": 1.1758573069801321, | |
| "learning_rate": 7.660816125154241e-05, | |
| "loss": 0.462, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.101360544217687, | |
| "grad_norm": 1.3078468404045238, | |
| "learning_rate": 7.648326513016425e-05, | |
| "loss": 0.4555, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.1122448979591837, | |
| "grad_norm": 0.9321604035491504, | |
| "learning_rate": 7.635621650118677e-05, | |
| "loss": 0.4616, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.1231292517006803, | |
| "grad_norm": 1.2219163619571358, | |
| "learning_rate": 7.622702286046149e-05, | |
| "loss": 0.4551, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1340136054421768, | |
| "grad_norm": 1.3163067619688875, | |
| "learning_rate": 7.609569183039539e-05, | |
| "loss": 0.4616, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1448979591836734, | |
| "grad_norm": 0.926896781931121, | |
| "learning_rate": 7.596223115950106e-05, | |
| "loss": 0.4532, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.15578231292517, | |
| "grad_norm": 1.4512737378338405, | |
| "learning_rate": 7.582664872193965e-05, | |
| "loss": 0.4517, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1666666666666667, | |
| "grad_norm": 0.7883527596734434, | |
| "learning_rate": 7.568895251705623e-05, | |
| "loss": 0.4493, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.1775510204081632, | |
| "grad_norm": 0.9041858110501911, | |
| "learning_rate": 7.554915066890786e-05, | |
| "loss": 0.4556, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1884353741496598, | |
| "grad_norm": 0.9022577877958278, | |
| "learning_rate": 7.540725142578425e-05, | |
| "loss": 0.4548, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1993197278911565, | |
| "grad_norm": 0.8757545344574125, | |
| "learning_rate": 7.526326315972111e-05, | |
| "loss": 0.4463, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2102040816326531, | |
| "grad_norm": 0.8092543270637563, | |
| "learning_rate": 7.511719436600626e-05, | |
| "loss": 0.4586, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.2210884353741496, | |
| "grad_norm": 1.1539210434525098, | |
| "learning_rate": 7.496905366267833e-05, | |
| "loss": 0.4551, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2319727891156462, | |
| "grad_norm": 1.276897526615545, | |
| "learning_rate": 7.481884979001832e-05, | |
| "loss": 0.4446, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2428571428571429, | |
| "grad_norm": 0.8248913940724176, | |
| "learning_rate": 7.466659161003397e-05, | |
| "loss": 0.4536, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.2537414965986395, | |
| "grad_norm": 1.246435153529166, | |
| "learning_rate": 7.451228810593686e-05, | |
| "loss": 0.4563, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2646258503401362, | |
| "grad_norm": 1.2887024185923162, | |
| "learning_rate": 7.435594838161237e-05, | |
| "loss": 0.4484, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.2755102040816326, | |
| "grad_norm": 0.9162489444999514, | |
| "learning_rate": 7.419758166108266e-05, | |
| "loss": 0.4406, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.2863945578231293, | |
| "grad_norm": 1.040658506235111, | |
| "learning_rate": 7.403719728796233e-05, | |
| "loss": 0.4537, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.297278911564626, | |
| "grad_norm": 0.8750938601699227, | |
| "learning_rate": 7.387480472490721e-05, | |
| "loss": 0.4404, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.3081632653061224, | |
| "grad_norm": 1.201728810077736, | |
| "learning_rate": 7.371041355305611e-05, | |
| "loss": 0.4475, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.319047619047619, | |
| "grad_norm": 1.4845906854873157, | |
| "learning_rate": 7.35440334714654e-05, | |
| "loss": 0.46, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.3299319727891157, | |
| "grad_norm": 0.9281427605043925, | |
| "learning_rate": 7.337567429653692e-05, | |
| "loss": 0.4509, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.3408163265306121, | |
| "grad_norm": 1.001102568409995, | |
| "learning_rate": 7.320534596143872e-05, | |
| "loss": 0.4488, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.3517006802721088, | |
| "grad_norm": 0.801780617998991, | |
| "learning_rate": 7.303305851551901e-05, | |
| "loss": 0.4488, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.3625850340136054, | |
| "grad_norm": 1.356706812518101, | |
| "learning_rate": 7.285882212371329e-05, | |
| "loss": 0.4586, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.373469387755102, | |
| "grad_norm": 1.0764564449246725, | |
| "learning_rate": 7.268264706594459e-05, | |
| "loss": 0.445, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.3843537414965987, | |
| "grad_norm": 0.6837548439098308, | |
| "learning_rate": 7.250454373651696e-05, | |
| "loss": 0.4588, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.3952380952380952, | |
| "grad_norm": 1.0602494402340488, | |
| "learning_rate": 7.232452264350223e-05, | |
| "loss": 0.4441, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.4061224489795918, | |
| "grad_norm": 1.348508059230177, | |
| "learning_rate": 7.214259440811999e-05, | |
| "loss": 0.4482, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.4170068027210885, | |
| "grad_norm": 0.7796798360257114, | |
| "learning_rate": 7.195876976411098e-05, | |
| "loss": 0.4419, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.427891156462585, | |
| "grad_norm": 1.1962320960568655, | |
| "learning_rate": 7.177305955710377e-05, | |
| "loss": 0.4595, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.4387755102040816, | |
| "grad_norm": 0.8790010135597905, | |
| "learning_rate": 7.158547474397495e-05, | |
| "loss": 0.4485, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.4496598639455782, | |
| "grad_norm": 1.2190711883432448, | |
| "learning_rate": 7.13960263922025e-05, | |
| "loss": 0.4499, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.4605442176870749, | |
| "grad_norm": 1.1191994343915495, | |
| "learning_rate": 7.120472567921303e-05, | |
| "loss": 0.4482, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4714285714285715, | |
| "grad_norm": 1.4555899933237715, | |
| "learning_rate": 7.101158389172215e-05, | |
| "loss": 0.4495, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.482312925170068, | |
| "grad_norm": 0.5648633573893016, | |
| "learning_rate": 7.08166124250686e-05, | |
| "loss": 0.4382, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.4931972789115646, | |
| "grad_norm": 1.4980585705796372, | |
| "learning_rate": 7.061982278254199e-05, | |
| "loss": 0.4518, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.5040816326530613, | |
| "grad_norm": 1.013726284837878, | |
| "learning_rate": 7.042122657470404e-05, | |
| "loss": 0.4515, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.5149659863945577, | |
| "grad_norm": 1.0657416056937796, | |
| "learning_rate": 7.02208355187035e-05, | |
| "loss": 0.4446, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5258503401360546, | |
| "grad_norm": 1.0651244759360778, | |
| "learning_rate": 7.001866143758502e-05, | |
| "loss": 0.4558, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.536734693877551, | |
| "grad_norm": 1.1753962967671663, | |
| "learning_rate": 6.981471625959144e-05, | |
| "loss": 0.4431, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.5476190476190477, | |
| "grad_norm": 1.2431833223672966, | |
| "learning_rate": 6.960901201746003e-05, | |
| "loss": 0.4511, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.5585034013605443, | |
| "grad_norm": 0.5885831162637244, | |
| "learning_rate": 6.940156084771267e-05, | |
| "loss": 0.4445, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.5693877551020408, | |
| "grad_norm": 0.9269977660478792, | |
| "learning_rate": 6.919237498993969e-05, | |
| "loss": 0.4553, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5802721088435374, | |
| "grad_norm": 1.3368384976249577, | |
| "learning_rate": 6.898146678607779e-05, | |
| "loss": 0.4564, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.591156462585034, | |
| "grad_norm": 0.8296616725485729, | |
| "learning_rate": 6.876884867968183e-05, | |
| "loss": 0.4513, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.6020408163265305, | |
| "grad_norm": 0.8722297340162901, | |
| "learning_rate": 6.85545332151907e-05, | |
| "loss": 0.4412, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.6129251700680272, | |
| "grad_norm": 0.7465168895009395, | |
| "learning_rate": 6.833853303718719e-05, | |
| "loss": 0.4411, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.6238095238095238, | |
| "grad_norm": 0.8288671891155779, | |
| "learning_rate": 6.812086088965192e-05, | |
| "loss": 0.4399, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.6346938775510202, | |
| "grad_norm": 0.9422384071031074, | |
| "learning_rate": 6.790152961521152e-05, | |
| "loss": 0.4513, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6455782312925171, | |
| "grad_norm": 0.8195855535647977, | |
| "learning_rate": 6.768055215438085e-05, | |
| "loss": 0.4486, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.6564625850340136, | |
| "grad_norm": 0.541315993639702, | |
| "learning_rate": 6.745794154479958e-05, | |
| "loss": 0.4413, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.6673469387755102, | |
| "grad_norm": 0.592580026109299, | |
| "learning_rate": 6.723371092046286e-05, | |
| "loss": 0.4423, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.6782312925170069, | |
| "grad_norm": 0.6392717194389805, | |
| "learning_rate": 6.700787351094656e-05, | |
| "loss": 0.4432, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.6891156462585033, | |
| "grad_norm": 0.8718762593821492, | |
| "learning_rate": 6.678044264062661e-05, | |
| "loss": 0.4444, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.2290064597395718, | |
| "learning_rate": 6.655143172789293e-05, | |
| "loss": 0.4426, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.7108843537414966, | |
| "grad_norm": 0.8757170575022087, | |
| "learning_rate": 6.63208542843577e-05, | |
| "loss": 0.4538, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.721768707482993, | |
| "grad_norm": 0.5896551012018776, | |
| "learning_rate": 6.608872391405822e-05, | |
| "loss": 0.4397, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.73265306122449, | |
| "grad_norm": 0.6625130166018679, | |
| "learning_rate": 6.585505431265425e-05, | |
| "loss": 0.4388, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.7435374149659864, | |
| "grad_norm": 0.8253999786654783, | |
| "learning_rate": 6.561985926661997e-05, | |
| "loss": 0.4454, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.754421768707483, | |
| "grad_norm": 1.0381989151586537, | |
| "learning_rate": 6.538315265243058e-05, | |
| "loss": 0.4479, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.7653061224489797, | |
| "grad_norm": 1.4403158752179994, | |
| "learning_rate": 6.51449484357436e-05, | |
| "loss": 0.4393, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.776190476190476, | |
| "grad_norm": 0.56782093063078, | |
| "learning_rate": 6.490526067057489e-05, | |
| "loss": 0.4362, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.7870748299319728, | |
| "grad_norm": 0.7487753783339337, | |
| "learning_rate": 6.466410349846944e-05, | |
| "loss": 0.4461, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.7979591836734694, | |
| "grad_norm": 1.435945446689746, | |
| "learning_rate": 6.44214911476671e-05, | |
| "loss": 0.4382, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.8088435374149658, | |
| "grad_norm": 0.642928569923028, | |
| "learning_rate": 6.4177437932263e-05, | |
| "loss": 0.4424, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.8197278911564627, | |
| "grad_norm": 1.1150172892638044, | |
| "learning_rate": 6.393195825136314e-05, | |
| "loss": 0.4498, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.8306122448979592, | |
| "grad_norm": 1.162525745704566, | |
| "learning_rate": 6.368506658823473e-05, | |
| "loss": 0.4459, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.8414965986394558, | |
| "grad_norm": 0.921502777493419, | |
| "learning_rate": 6.343677750945177e-05, | |
| "loss": 0.4448, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.8523809523809525, | |
| "grad_norm": 1.0298572207780852, | |
| "learning_rate": 6.318710566403559e-05, | |
| "loss": 0.4443, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.863265306122449, | |
| "grad_norm": 0.6276886479215986, | |
| "learning_rate": 6.293606578259057e-05, | |
| "loss": 0.4445, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.8741496598639455, | |
| "grad_norm": 0.8605722904904601, | |
| "learning_rate": 6.2683672676435e-05, | |
| "loss": 0.4458, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.8850340136054422, | |
| "grad_norm": 0.9934365285628636, | |
| "learning_rate": 6.242994123672725e-05, | |
| "loss": 0.4388, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.8959183673469386, | |
| "grad_norm": 0.8730130209669299, | |
| "learning_rate": 6.21748864335872e-05, | |
| "loss": 0.4323, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.9068027210884355, | |
| "grad_norm": 0.6540737360682325, | |
| "learning_rate": 6.191852331521294e-05, | |
| "loss": 0.436, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.917687074829932, | |
| "grad_norm": 0.7654958043455478, | |
| "learning_rate": 6.166086700699306e-05, | |
| "loss": 0.442, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 0.8887001609804748, | |
| "learning_rate": 6.140193271061411e-05, | |
| "loss": 0.4276, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.9394557823129253, | |
| "grad_norm": 0.8204714163689445, | |
| "learning_rate": 6.114173570316378e-05, | |
| "loss": 0.4388, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.9503401360544217, | |
| "grad_norm": 0.6675554963136549, | |
| "learning_rate": 6.088029133622955e-05, | |
| "loss": 0.4426, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.9612244897959183, | |
| "grad_norm": 0.6210957022083031, | |
| "learning_rate": 6.061761503499291e-05, | |
| "loss": 0.4449, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.972108843537415, | |
| "grad_norm": 0.5985091475788507, | |
| "learning_rate": 6.0353722297319344e-05, | |
| "loss": 0.4435, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.9829931972789114, | |
| "grad_norm": 0.561189475626402, | |
| "learning_rate": 6.0088628692843877e-05, | |
| "loss": 0.4355, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.9938775510204083, | |
| "grad_norm": 0.5447821415803845, | |
| "learning_rate": 5.982234986205252e-05, | |
| "loss": 0.5047, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.006802721088435, | |
| "grad_norm": 0.6277749488977088, | |
| "learning_rate": 5.955490151535951e-05, | |
| "loss": 0.4933, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.017687074829932, | |
| "grad_norm": 0.6187666709394345, | |
| "learning_rate": 5.928629943218029e-05, | |
| "loss": 0.4045, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.0285714285714285, | |
| "grad_norm": 0.7808989155818872, | |
| "learning_rate": 5.901655946000068e-05, | |
| "loss": 0.4102, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.0394557823129253, | |
| "grad_norm": 0.661050182290931, | |
| "learning_rate": 5.8745697513441734e-05, | |
| "loss": 0.3939, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.050340136054422, | |
| "grad_norm": 0.6585476168351413, | |
| "learning_rate": 5.8473729573320845e-05, | |
| "loss": 0.4113, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.061224489795918, | |
| "grad_norm": 0.911313578766385, | |
| "learning_rate": 5.820067168570888e-05, | |
| "loss": 0.4082, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.072108843537415, | |
| "grad_norm": 1.0469880706603127, | |
| "learning_rate": 5.7926539960983485e-05, | |
| "loss": 0.4121, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.0829931972789115, | |
| "grad_norm": 0.7191049383152186, | |
| "learning_rate": 5.76513505728785e-05, | |
| "loss": 0.4051, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.093877551020408, | |
| "grad_norm": 0.45051085645785816, | |
| "learning_rate": 5.7375119757529806e-05, | |
| "loss": 0.4008, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.104761904761905, | |
| "grad_norm": 0.6773584179523391, | |
| "learning_rate": 5.70978638125173e-05, | |
| "loss": 0.3988, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.1156462585034013, | |
| "grad_norm": 0.7896934843059149, | |
| "learning_rate": 5.681959909590342e-05, | |
| "loss": 0.399, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.126530612244898, | |
| "grad_norm": 0.443780434741391, | |
| "learning_rate": 5.654034202526799e-05, | |
| "loss": 0.3987, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.1374149659863946, | |
| "grad_norm": 0.5023289394189521, | |
| "learning_rate": 5.626010907673955e-05, | |
| "loss": 0.4064, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.148299319727891, | |
| "grad_norm": 0.6002819530169419, | |
| "learning_rate": 5.5978916784023356e-05, | |
| "loss": 0.404, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.159183673469388, | |
| "grad_norm": 0.45036480622265124, | |
| "learning_rate": 5.569678173742577e-05, | |
| "loss": 0.4029, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.1700680272108843, | |
| "grad_norm": 0.4234983416414813, | |
| "learning_rate": 5.541372058287557e-05, | |
| "loss": 0.4004, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.1809523809523808, | |
| "grad_norm": 0.5843235500892456, | |
| "learning_rate": 5.512975002094172e-05, | |
| "loss": 0.4073, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.1918367346938776, | |
| "grad_norm": 0.47027866282925174, | |
| "learning_rate": 5.484488680584815e-05, | |
| "loss": 0.4108, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.202721088435374, | |
| "grad_norm": 0.5705510783184461, | |
| "learning_rate": 5.455914774448515e-05, | |
| "loss": 0.4092, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.213605442176871, | |
| "grad_norm": 0.6358413293322973, | |
| "learning_rate": 5.427254969541784e-05, | |
| "loss": 0.4068, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.2244897959183674, | |
| "grad_norm": 0.5159793783086416, | |
| "learning_rate": 5.3985109567891495e-05, | |
| "loss": 0.3981, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.235374149659864, | |
| "grad_norm": 0.37977672367359394, | |
| "learning_rate": 5.369684432083391e-05, | |
| "loss": 0.3969, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.2462585034013607, | |
| "grad_norm": 0.5102363422380745, | |
| "learning_rate": 5.340777096185482e-05, | |
| "loss": 0.405, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.257142857142857, | |
| "grad_norm": 0.4235067529697151, | |
| "learning_rate": 5.311790654624243e-05, | |
| "loss": 0.3978, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.2680272108843536, | |
| "grad_norm": 0.3869416986593155, | |
| "learning_rate": 5.2827268175957204e-05, | |
| "loss": 0.4037, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.2789115646258504, | |
| "grad_norm": 0.3459197499629783, | |
| "learning_rate": 5.2535872998622805e-05, | |
| "loss": 0.3991, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.289795918367347, | |
| "grad_norm": 0.4061776937952424, | |
| "learning_rate": 5.224373820651441e-05, | |
| "loss": 0.4008, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.3006802721088437, | |
| "grad_norm": 0.2828373312543394, | |
| "learning_rate": 5.1950881035544375e-05, | |
| "loss": 0.3964, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.31156462585034, | |
| "grad_norm": 0.4037016748215062, | |
| "learning_rate": 5.165731876424531e-05, | |
| "loss": 0.3908, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.3224489795918366, | |
| "grad_norm": 0.2668077682766602, | |
| "learning_rate": 5.136306871275063e-05, | |
| "loss": 0.4017, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.32985921196728035, | |
| "learning_rate": 5.106814824177269e-05, | |
| "loss": 0.4079, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.34421768707483, | |
| "grad_norm": 0.3983009624763948, | |
| "learning_rate": 5.077257475157852e-05, | |
| "loss": 0.3961, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.3551020408163263, | |
| "grad_norm": 0.28997246390855486, | |
| "learning_rate": 5.047636568096318e-05, | |
| "loss": 0.396, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.3659863945578232, | |
| "grad_norm": 0.353479423426273, | |
| "learning_rate": 5.01795385062209e-05, | |
| "loss": 0.3977, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.3768707482993197, | |
| "grad_norm": 0.3149728478919283, | |
| "learning_rate": 4.988211074011394e-05, | |
| "loss": 0.3981, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.387755102040816, | |
| "grad_norm": 0.29447084389008193, | |
| "learning_rate": 4.958409993083938e-05, | |
| "loss": 0.4018, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.398639455782313, | |
| "grad_norm": 0.30340006360819105, | |
| "learning_rate": 4.9285523660993786e-05, | |
| "loss": 0.399, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.4095238095238094, | |
| "grad_norm": 0.26690833941342385, | |
| "learning_rate": 4.8986399546535805e-05, | |
| "loss": 0.3975, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.4204081632653063, | |
| "grad_norm": 0.3360364345210551, | |
| "learning_rate": 4.868674523574684e-05, | |
| "loss": 0.3992, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.4312925170068027, | |
| "grad_norm": 0.39834754044674225, | |
| "learning_rate": 4.838657840818979e-05, | |
| "loss": 0.4002, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.442176870748299, | |
| "grad_norm": 0.35848487200390305, | |
| "learning_rate": 4.8085916773666006e-05, | |
| "loss": 0.3894, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.453061224489796, | |
| "grad_norm": 0.26561850084387234, | |
| "learning_rate": 4.7784778071170334e-05, | |
| "loss": 0.3969, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.4639455782312925, | |
| "grad_norm": 0.25933266150714, | |
| "learning_rate": 4.748318006784463e-05, | |
| "loss": 0.3991, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.4748299319727893, | |
| "grad_norm": 0.22763876530025542, | |
| "learning_rate": 4.7181140557929365e-05, | |
| "loss": 0.4046, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.4857142857142858, | |
| "grad_norm": 0.28285284569921715, | |
| "learning_rate": 4.6878677361713886e-05, | |
| "loss": 0.403, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.496598639455782, | |
| "grad_norm": 0.23120988448000654, | |
| "learning_rate": 4.657580832448496e-05, | |
| "loss": 0.4044, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.507482993197279, | |
| "grad_norm": 0.2491717348489704, | |
| "learning_rate": 4.627255131547391e-05, | |
| "loss": 0.399, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.5183673469387755, | |
| "grad_norm": 0.25850052315456495, | |
| "learning_rate": 4.596892422680237e-05, | |
| "loss": 0.4027, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.5292517006802724, | |
| "grad_norm": 0.25098684145251837, | |
| "learning_rate": 4.566494497242655e-05, | |
| "loss": 0.3935, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.540136054421769, | |
| "grad_norm": 0.28359043341703694, | |
| "learning_rate": 4.536063148708047e-05, | |
| "loss": 0.4092, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.5510204081632653, | |
| "grad_norm": 0.2691550823951915, | |
| "learning_rate": 4.505600172521769e-05, | |
| "loss": 0.3965, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.5619047619047617, | |
| "grad_norm": 0.22718444292959802, | |
| "learning_rate": 4.475107365995206e-05, | |
| "loss": 0.395, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.5727891156462586, | |
| "grad_norm": 0.2533624032162829, | |
| "learning_rate": 4.444586528199727e-05, | |
| "loss": 0.3911, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.583673469387755, | |
| "grad_norm": 0.26258878584903417, | |
| "learning_rate": 4.414039459860542e-05, | |
| "loss": 0.397, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.594557823129252, | |
| "grad_norm": 0.21708562366582754, | |
| "learning_rate": 4.3834679632504606e-05, | |
| "loss": 0.3983, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.6054421768707483, | |
| "grad_norm": 0.20456400104309047, | |
| "learning_rate": 4.352873842083558e-05, | |
| "loss": 0.3985, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.6163265306122447, | |
| "grad_norm": 0.20807703125099125, | |
| "learning_rate": 4.322258901408752e-05, | |
| "loss": 0.4044, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.6272108843537416, | |
| "grad_norm": 0.1757775396046016, | |
| "learning_rate": 4.2916249475033114e-05, | |
| "loss": 0.3936, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.638095238095238, | |
| "grad_norm": 0.2081974793590528, | |
| "learning_rate": 4.260973787766282e-05, | |
| "loss": 0.3957, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.648979591836735, | |
| "grad_norm": 0.23777319469622146, | |
| "learning_rate": 4.230307230611851e-05, | |
| "loss": 0.4002, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.6598639455782314, | |
| "grad_norm": 0.23523713066143154, | |
| "learning_rate": 4.199627085362649e-05, | |
| "loss": 0.3981, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.670748299319728, | |
| "grad_norm": 0.27766508540483775, | |
| "learning_rate": 4.168935162143009e-05, | |
| "loss": 0.3959, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.6816326530612242, | |
| "grad_norm": 0.2875135405402182, | |
| "learning_rate": 4.138233271772155e-05, | |
| "loss": 0.4052, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.692517006802721, | |
| "grad_norm": 0.21692037182410404, | |
| "learning_rate": 4.107523225657375e-05, | |
| "loss": 0.3879, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.7034013605442175, | |
| "grad_norm": 0.22904494850511364, | |
| "learning_rate": 4.076806835687145e-05, | |
| "loss": 0.4017, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.7142857142857144, | |
| "grad_norm": 0.23262515207207654, | |
| "learning_rate": 4.046085914124227e-05, | |
| "loss": 0.4001, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.725170068027211, | |
| "grad_norm": 0.20573148436607291, | |
| "learning_rate": 4.015362273498747e-05, | |
| "loss": 0.3968, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.7360544217687073, | |
| "grad_norm": 0.26382175021100807, | |
| "learning_rate": 3.9846377265012535e-05, | |
| "loss": 0.4028, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.746938775510204, | |
| "grad_norm": 0.23542305446748288, | |
| "learning_rate": 3.953914085875774e-05, | |
| "loss": 0.4011, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.7578231292517006, | |
| "grad_norm": 0.18599068389111756, | |
| "learning_rate": 3.923193164312857e-05, | |
| "loss": 0.4079, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.7687074829931975, | |
| "grad_norm": 0.22813282345099503, | |
| "learning_rate": 3.892476774342626e-05, | |
| "loss": 0.3951, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.779591836734694, | |
| "grad_norm": 0.17780855634412887, | |
| "learning_rate": 3.861766728227847e-05, | |
| "loss": 0.3952, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.7904761904761903, | |
| "grad_norm": 0.21195953442928203, | |
| "learning_rate": 3.8310648378569914e-05, | |
| "loss": 0.4105, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.801360544217687, | |
| "grad_norm": 0.22748748201389732, | |
| "learning_rate": 3.800372914637352e-05, | |
| "loss": 0.3983, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.8122448979591836, | |
| "grad_norm": 0.20056553430427193, | |
| "learning_rate": 3.769692769388151e-05, | |
| "loss": 0.4054, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.8231292517006805, | |
| "grad_norm": 0.24634483325768172, | |
| "learning_rate": 3.7390262122337195e-05, | |
| "loss": 0.3989, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.834013605442177, | |
| "grad_norm": 0.19455006181368653, | |
| "learning_rate": 3.708375052496689e-05, | |
| "loss": 0.3953, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.8448979591836734, | |
| "grad_norm": 0.2037769624898762, | |
| "learning_rate": 3.677741098591249e-05, | |
| "loss": 0.4038, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.85578231292517, | |
| "grad_norm": 0.1889292033164084, | |
| "learning_rate": 3.647126157916443e-05, | |
| "loss": 0.4001, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.8666666666666667, | |
| "grad_norm": 0.18651256474304687, | |
| "learning_rate": 3.616532036749541e-05, | |
| "loss": 0.3913, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.877551020408163, | |
| "grad_norm": 0.2222641833561834, | |
| "learning_rate": 3.5859605401394594e-05, | |
| "loss": 0.4006, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.88843537414966, | |
| "grad_norm": 0.15833108742876167, | |
| "learning_rate": 3.555413471800275e-05, | |
| "loss": 0.402, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.8993197278911564, | |
| "grad_norm": 0.18698241486043574, | |
| "learning_rate": 3.524892634004795e-05, | |
| "loss": 0.4032, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.910204081632653, | |
| "grad_norm": 0.1566143167225701, | |
| "learning_rate": 3.4943998274782324e-05, | |
| "loss": 0.3976, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.9210884353741497, | |
| "grad_norm": 0.18372520204803328, | |
| "learning_rate": 3.4639368512919546e-05, | |
| "loss": 0.3857, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.931972789115646, | |
| "grad_norm": 0.16856978707068154, | |
| "learning_rate": 3.433505502757346e-05, | |
| "loss": 0.3916, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.942857142857143, | |
| "grad_norm": 0.18813153276502972, | |
| "learning_rate": 3.403107577319766e-05, | |
| "loss": 0.404, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.9537414965986395, | |
| "grad_norm": 0.17468778423516987, | |
| "learning_rate": 3.3727448684526095e-05, | |
| "loss": 0.3976, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.964625850340136, | |
| "grad_norm": 0.18419955817162595, | |
| "learning_rate": 3.342419167551505e-05, | |
| "loss": 0.4004, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.975510204081633, | |
| "grad_norm": 0.18271982709236256, | |
| "learning_rate": 3.312132263828613e-05, | |
| "loss": 0.4055, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.9863945578231292, | |
| "grad_norm": 0.1651743278895426, | |
| "learning_rate": 3.281885944207065e-05, | |
| "loss": 0.3933, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.997278911564626, | |
| "grad_norm": 0.2456121705657945, | |
| "learning_rate": 3.2516819932155386e-05, | |
| "loss": 0.5208, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.010204081632653, | |
| "grad_norm": 0.23960570793480176, | |
| "learning_rate": 3.221522192882967e-05, | |
| "loss": 0.3767, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.02108843537415, | |
| "grad_norm": 0.2203433260910606, | |
| "learning_rate": 3.1914083226334014e-05, | |
| "loss": 0.3719, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.0319727891156463, | |
| "grad_norm": 0.20784386772069446, | |
| "learning_rate": 3.1613421591810226e-05, | |
| "loss": 0.3653, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.0428571428571427, | |
| "grad_norm": 0.2523046090952252, | |
| "learning_rate": 3.1313254764253174e-05, | |
| "loss": 0.3654, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.0537414965986396, | |
| "grad_norm": 0.22271729085507425, | |
| "learning_rate": 3.101360045346421e-05, | |
| "loss": 0.3721, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.064625850340136, | |
| "grad_norm": 0.2065491727367014, | |
| "learning_rate": 3.0714476339006214e-05, | |
| "loss": 0.369, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.0755102040816324, | |
| "grad_norm": 0.2652821591584382, | |
| "learning_rate": 3.0415900069160634e-05, | |
| "loss": 0.3698, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.0863945578231293, | |
| "grad_norm": 0.2172159761840688, | |
| "learning_rate": 3.0117889259886088e-05, | |
| "loss": 0.3671, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.0972789115646258, | |
| "grad_norm": 0.22451418814997312, | |
| "learning_rate": 2.9820461493779117e-05, | |
| "loss": 0.3647, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.1081632653061226, | |
| "grad_norm": 0.21683351034608486, | |
| "learning_rate": 2.952363431903683e-05, | |
| "loss": 0.3623, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.119047619047619, | |
| "grad_norm": 0.2158697560884106, | |
| "learning_rate": 2.922742524842148e-05, | |
| "loss": 0.3676, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.1299319727891155, | |
| "grad_norm": 0.19602289035159243, | |
| "learning_rate": 2.8931851758227325e-05, | |
| "loss": 0.3668, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.1408163265306124, | |
| "grad_norm": 0.20952047645898908, | |
| "learning_rate": 2.8636931287249393e-05, | |
| "loss": 0.3618, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.151700680272109, | |
| "grad_norm": 0.22727644896398902, | |
| "learning_rate": 2.8342681235754697e-05, | |
| "loss": 0.3614, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.1625850340136052, | |
| "grad_norm": 0.18973207315735008, | |
| "learning_rate": 2.8049118964455635e-05, | |
| "loss": 0.3638, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.173469387755102, | |
| "grad_norm": 0.2059879834709156, | |
| "learning_rate": 2.775626179348559e-05, | |
| "loss": 0.3681, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.1843537414965986, | |
| "grad_norm": 0.20974273512672587, | |
| "learning_rate": 2.746412700137721e-05, | |
| "loss": 0.3735, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.1952380952380954, | |
| "grad_norm": 0.1624512775469249, | |
| "learning_rate": 2.71727318240428e-05, | |
| "loss": 0.3605, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.206122448979592, | |
| "grad_norm": 0.1841826692146313, | |
| "learning_rate": 2.688209345375757e-05, | |
| "loss": 0.3605, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.2170068027210883, | |
| "grad_norm": 0.15775552615251343, | |
| "learning_rate": 2.659222903814519e-05, | |
| "loss": 0.3637, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.227891156462585, | |
| "grad_norm": 0.17956234909271063, | |
| "learning_rate": 2.6303155679166087e-05, | |
| "loss": 0.3598, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.2387755102040816, | |
| "grad_norm": 0.15107059770904596, | |
| "learning_rate": 2.6014890432108515e-05, | |
| "loss": 0.362, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.249659863945578, | |
| "grad_norm": 0.1729470704096737, | |
| "learning_rate": 2.5727450304582183e-05, | |
| "loss": 0.3653, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.260544217687075, | |
| "grad_norm": 0.14695616813488518, | |
| "learning_rate": 2.5440852255514862e-05, | |
| "loss": 0.3619, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.2714285714285714, | |
| "grad_norm": 0.15153239550233033, | |
| "learning_rate": 2.5155113194151866e-05, | |
| "loss": 0.3647, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.282312925170068, | |
| "grad_norm": 0.14386901617429848, | |
| "learning_rate": 2.4870249979058277e-05, | |
| "loss": 0.3683, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.2931972789115647, | |
| "grad_norm": 0.16505695312061414, | |
| "learning_rate": 2.4586279417124442e-05, | |
| "loss": 0.3642, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.304081632653061, | |
| "grad_norm": 0.13136291309750409, | |
| "learning_rate": 2.4303218262574248e-05, | |
| "loss": 0.3589, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.314965986394558, | |
| "grad_norm": 0.14648334198611954, | |
| "learning_rate": 2.402108321597666e-05, | |
| "loss": 0.3689, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.3258503401360544, | |
| "grad_norm": 0.13706322637356855, | |
| "learning_rate": 2.373989092326046e-05, | |
| "loss": 0.3686, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.336734693877551, | |
| "grad_norm": 0.1461705155320018, | |
| "learning_rate": 2.3459657974732018e-05, | |
| "loss": 0.3594, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.3476190476190477, | |
| "grad_norm": 0.13113890606662817, | |
| "learning_rate": 2.318040090409658e-05, | |
| "loss": 0.3717, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.358503401360544, | |
| "grad_norm": 0.13070704625373702, | |
| "learning_rate": 2.290213618748271e-05, | |
| "loss": 0.3691, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.369387755102041, | |
| "grad_norm": 0.13756838029516322, | |
| "learning_rate": 2.26248802424702e-05, | |
| "loss": 0.3676, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.3802721088435375, | |
| "grad_norm": 0.1300726141584972, | |
| "learning_rate": 2.2348649427121502e-05, | |
| "loss": 0.3678, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.391156462585034, | |
| "grad_norm": 0.13698477885929453, | |
| "learning_rate": 2.2073460039016518e-05, | |
| "loss": 0.3579, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.4020408163265308, | |
| "grad_norm": 0.127952587088721, | |
| "learning_rate": 2.1799328314291124e-05, | |
| "loss": 0.3688, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 3.412925170068027, | |
| "grad_norm": 0.13599378638348314, | |
| "learning_rate": 2.1526270426679178e-05, | |
| "loss": 0.3648, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 3.4238095238095236, | |
| "grad_norm": 0.13302334026973409, | |
| "learning_rate": 2.1254302486558272e-05, | |
| "loss": 0.37, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 3.4346938775510205, | |
| "grad_norm": 0.1350124126585442, | |
| "learning_rate": 2.0983440539999326e-05, | |
| "loss": 0.3635, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 3.445578231292517, | |
| "grad_norm": 0.12014814686561753, | |
| "learning_rate": 2.0713700567819705e-05, | |
| "loss": 0.364, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 3.4564625850340134, | |
| "grad_norm": 0.1309081194484556, | |
| "learning_rate": 2.044509848464051e-05, | |
| "loss": 0.3632, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 3.4673469387755103, | |
| "grad_norm": 0.11665106119511277, | |
| "learning_rate": 2.0177650137947493e-05, | |
| "loss": 0.3639, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 3.4782312925170067, | |
| "grad_norm": 0.1330904699075859, | |
| "learning_rate": 1.9911371307156137e-05, | |
| "loss": 0.3739, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 3.4891156462585036, | |
| "grad_norm": 0.12075945257165684, | |
| "learning_rate": 1.9646277702680673e-05, | |
| "loss": 0.3616, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.13303895659252055, | |
| "learning_rate": 1.9382384965007092e-05, | |
| "loss": 0.3654, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 3.5108843537414964, | |
| "grad_norm": 0.1278694845983754, | |
| "learning_rate": 1.9119708663770465e-05, | |
| "loss": 0.3779, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 3.5217687074829933, | |
| "grad_norm": 0.1253521665884257, | |
| "learning_rate": 1.8858264296836232e-05, | |
| "loss": 0.365, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 3.5326530612244897, | |
| "grad_norm": 0.12311807593450247, | |
| "learning_rate": 1.8598067289385896e-05, | |
| "loss": 0.3708, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 3.5435374149659866, | |
| "grad_norm": 0.12757634153807798, | |
| "learning_rate": 1.8339132993006952e-05, | |
| "loss": 0.3646, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.554421768707483, | |
| "grad_norm": 0.11611444956464281, | |
| "learning_rate": 1.8081476684787063e-05, | |
| "loss": 0.3613, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 3.5653061224489795, | |
| "grad_norm": 0.1373379713477394, | |
| "learning_rate": 1.7825113566412828e-05, | |
| "loss": 0.3594, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 3.576190476190476, | |
| "grad_norm": 0.1238139171501049, | |
| "learning_rate": 1.7570058763272774e-05, | |
| "loss": 0.3594, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 3.587074829931973, | |
| "grad_norm": 0.13091245071909532, | |
| "learning_rate": 1.7316327323565013e-05, | |
| "loss": 0.358, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 3.5979591836734692, | |
| "grad_norm": 0.12875284272011517, | |
| "learning_rate": 1.7063934217409445e-05, | |
| "loss": 0.3667, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.608843537414966, | |
| "grad_norm": 0.12552085776059319, | |
| "learning_rate": 1.681289433596441e-05, | |
| "loss": 0.3635, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 3.6197278911564625, | |
| "grad_norm": 0.13293083890337762, | |
| "learning_rate": 1.6563222490548246e-05, | |
| "loss": 0.3691, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 3.630612244897959, | |
| "grad_norm": 0.12231951238569645, | |
| "learning_rate": 1.631493341176529e-05, | |
| "loss": 0.3654, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 3.641496598639456, | |
| "grad_norm": 0.1178045268839638, | |
| "learning_rate": 1.606804174863688e-05, | |
| "loss": 0.3686, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 3.6523809523809523, | |
| "grad_norm": 0.11149430822353905, | |
| "learning_rate": 1.5822562067737e-05, | |
| "loss": 0.3588, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.663265306122449, | |
| "grad_norm": 0.1303840473675821, | |
| "learning_rate": 1.557850885233291e-05, | |
| "loss": 0.3672, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.6741496598639456, | |
| "grad_norm": 0.10551516116987211, | |
| "learning_rate": 1.5335896501530555e-05, | |
| "loss": 0.3573, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 3.685034013605442, | |
| "grad_norm": 0.12470092805120879, | |
| "learning_rate": 1.5094739329425121e-05, | |
| "loss": 0.3694, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.695918367346939, | |
| "grad_norm": 0.12323569653399578, | |
| "learning_rate": 1.4855051564256399e-05, | |
| "loss": 0.3669, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.7068027210884353, | |
| "grad_norm": 0.11500208573388462, | |
| "learning_rate": 1.4616847347569425e-05, | |
| "loss": 0.3605, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.717687074829932, | |
| "grad_norm": 0.12121616369263918, | |
| "learning_rate": 1.4380140733380027e-05, | |
| "loss": 0.3612, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.7285714285714286, | |
| "grad_norm": 0.11946548753604758, | |
| "learning_rate": 1.4144945687345755e-05, | |
| "loss": 0.3622, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.739455782312925, | |
| "grad_norm": 0.1266069711549675, | |
| "learning_rate": 1.3911276085941792e-05, | |
| "loss": 0.3624, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.7503401360544215, | |
| "grad_norm": 0.1123073682607083, | |
| "learning_rate": 1.3679145715642305e-05, | |
| "loss": 0.3624, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.7612244897959184, | |
| "grad_norm": 0.11784151935161251, | |
| "learning_rate": 1.344856827210708e-05, | |
| "loss": 0.3615, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.772108843537415, | |
| "grad_norm": 0.12945592509583792, | |
| "learning_rate": 1.3219557359373388e-05, | |
| "loss": 0.3625, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.7829931972789117, | |
| "grad_norm": 0.10670513405678321, | |
| "learning_rate": 1.2992126489053446e-05, | |
| "loss": 0.3641, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.793877551020408, | |
| "grad_norm": 0.12659423146947774, | |
| "learning_rate": 1.2766289079537151e-05, | |
| "loss": 0.3664, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.8047619047619046, | |
| "grad_norm": 0.11130274376327387, | |
| "learning_rate": 1.2542058455200432e-05, | |
| "loss": 0.3705, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.8156462585034014, | |
| "grad_norm": 0.11826983032480667, | |
| "learning_rate": 1.2319447845619154e-05, | |
| "loss": 0.3685, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.826530612244898, | |
| "grad_norm": 0.11041065098921128, | |
| "learning_rate": 1.2098470384788485e-05, | |
| "loss": 0.3616, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.8374149659863948, | |
| "grad_norm": 0.10717227222325487, | |
| "learning_rate": 1.187913911034809e-05, | |
| "loss": 0.3605, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.848299319727891, | |
| "grad_norm": 0.10666671914874028, | |
| "learning_rate": 1.166146696281283e-05, | |
| "loss": 0.3602, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.8591836734693876, | |
| "grad_norm": 0.10095960559508507, | |
| "learning_rate": 1.1445466784809303e-05, | |
| "loss": 0.3601, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.8700680272108845, | |
| "grad_norm": 0.10136507845344848, | |
| "learning_rate": 1.1231151320318178e-05, | |
| "loss": 0.3652, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.880952380952381, | |
| "grad_norm": 0.11594019435680152, | |
| "learning_rate": 1.1018533213922211e-05, | |
| "loss": 0.3663, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.891836734693878, | |
| "grad_norm": 0.10236332089943526, | |
| "learning_rate": 1.0807625010060313e-05, | |
| "loss": 0.37, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.9027210884353742, | |
| "grad_norm": 0.10696438976142728, | |
| "learning_rate": 1.0598439152287346e-05, | |
| "loss": 0.3702, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.9136054421768707, | |
| "grad_norm": 0.10415916356446016, | |
| "learning_rate": 1.0390987982539983e-05, | |
| "loss": 0.3632, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.924489795918367, | |
| "grad_norm": 0.10220339688296681, | |
| "learning_rate": 1.0185283740408583e-05, | |
| "loss": 0.359, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.935374149659864, | |
| "grad_norm": 0.10301989669223478, | |
| "learning_rate": 9.981338562414984e-06, | |
| "loss": 0.3595, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.9462585034013604, | |
| "grad_norm": 0.10931528658602535, | |
| "learning_rate": 9.779164481296509e-06, | |
| "loss": 0.3584, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.9571428571428573, | |
| "grad_norm": 0.12050352831212997, | |
| "learning_rate": 9.578773425295988e-06, | |
| "loss": 0.3682, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.9680272108843537, | |
| "grad_norm": 0.10028402542481818, | |
| "learning_rate": 9.380177217458014e-06, | |
| "loss": 0.366, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.97891156462585, | |
| "grad_norm": 0.10659990087833227, | |
| "learning_rate": 9.183387574931411e-06, | |
| "loss": 0.3596, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.989795918367347, | |
| "grad_norm": 0.10930883629577616, | |
| "learning_rate": 8.988416108277867e-06, | |
| "loss": 0.3572, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.002721088435374, | |
| "grad_norm": 0.13574097898710413, | |
| "learning_rate": 8.795274320786985e-06, | |
| "loss": 0.4663, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.01360544217687, | |
| "grad_norm": 0.14783584782906442, | |
| "learning_rate": 8.603973607797504e-06, | |
| "loss": 0.3454, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.024489795918368, | |
| "grad_norm": 0.13285325719918925, | |
| "learning_rate": 8.414525256025069e-06, | |
| "loss": 0.3383, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.035374149659864, | |
| "grad_norm": 0.11044378505123642, | |
| "learning_rate": 8.226940442896225e-06, | |
| "loss": 0.3358, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.0462585034013605, | |
| "grad_norm": 0.1164873429433178, | |
| "learning_rate": 8.04123023588903e-06, | |
| "loss": 0.344, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.057142857142857, | |
| "grad_norm": 0.131074129183923, | |
| "learning_rate": 7.857405591880014e-06, | |
| "loss": 0.3447, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.068027210884353, | |
| "grad_norm": 0.128752600063885, | |
| "learning_rate": 7.675477356497775e-06, | |
| "loss": 0.3471, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.078911564625851, | |
| "grad_norm": 0.11702773712400304, | |
| "learning_rate": 7.495456263483034e-06, | |
| "loss": 0.3401, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.089795918367347, | |
| "grad_norm": 0.12447256123820202, | |
| "learning_rate": 7.317352934055413e-06, | |
| "loss": 0.3516, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.100680272108844, | |
| "grad_norm": 0.11159427948052647, | |
| "learning_rate": 7.141177876286706e-06, | |
| "loss": 0.3461, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.11156462585034, | |
| "grad_norm": 0.11574387273740891, | |
| "learning_rate": 6.966941484480991e-06, | |
| "loss": 0.3399, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.122448979591836, | |
| "grad_norm": 0.11705876868290095, | |
| "learning_rate": 6.794654038561285e-06, | |
| "loss": 0.3478, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.133333333333334, | |
| "grad_norm": 0.11205521919284007, | |
| "learning_rate": 6.624325703463075e-06, | |
| "loss": 0.3474, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.14421768707483, | |
| "grad_norm": 0.10916103777536715, | |
| "learning_rate": 6.455966528534605e-06, | |
| "loss": 0.3471, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.155102040816327, | |
| "grad_norm": 0.10850947298262742, | |
| "learning_rate": 6.289586446943898e-06, | |
| "loss": 0.3368, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.165986394557823, | |
| "grad_norm": 0.11514113628015869, | |
| "learning_rate": 6.1251952750927875e-06, | |
| "loss": 0.344, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.1768707482993195, | |
| "grad_norm": 0.10385908273367876, | |
| "learning_rate": 5.962802712037685e-06, | |
| "loss": 0.3396, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.187755102040816, | |
| "grad_norm": 0.09512302830262585, | |
| "learning_rate": 5.802418338917344e-06, | |
| "loss": 0.3374, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.198639455782313, | |
| "grad_norm": 0.10274721856365138, | |
| "learning_rate": 5.644051618387636e-06, | |
| "loss": 0.338, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.20952380952381, | |
| "grad_norm": 0.09886214923951267, | |
| "learning_rate": 5.48771189406315e-06, | |
| "loss": 0.3466, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.220408163265306, | |
| "grad_norm": 0.10504332956184576, | |
| "learning_rate": 5.333408389966037e-06, | |
| "loss": 0.3495, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.2312925170068025, | |
| "grad_norm": 0.09875168815734763, | |
| "learning_rate": 5.181150209981693e-06, | |
| "loss": 0.342, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.242176870748299, | |
| "grad_norm": 0.09802223656901986, | |
| "learning_rate": 5.030946337321685e-06, | |
| "loss": 0.3457, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.253061224489796, | |
| "grad_norm": 0.09410603329511927, | |
| "learning_rate": 4.882805633993752e-06, | |
| "loss": 0.3484, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.263945578231293, | |
| "grad_norm": 0.09422246682254971, | |
| "learning_rate": 4.736736840278901e-06, | |
| "loss": 0.3494, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 4.274829931972789, | |
| "grad_norm": 0.09861037593912006, | |
| "learning_rate": 4.592748574215766e-06, | |
| "loss": 0.3509, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.0935311998173292, | |
| "learning_rate": 4.450849331092153e-06, | |
| "loss": 0.3412, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 4.296598639455782, | |
| "grad_norm": 0.0944360546022518, | |
| "learning_rate": 4.311047482943771e-06, | |
| "loss": 0.3448, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 4.307482993197279, | |
| "grad_norm": 0.0932248976438753, | |
| "learning_rate": 4.1733512780603556e-06, | |
| "loss": 0.3375, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 4.318367346938776, | |
| "grad_norm": 0.09584158230979174, | |
| "learning_rate": 4.037768840498948e-06, | |
| "loss": 0.3449, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 4.329251700680272, | |
| "grad_norm": 0.09364314245427798, | |
| "learning_rate": 3.904308169604623e-06, | |
| "loss": 0.346, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 4.340136054421769, | |
| "grad_norm": 0.08822462366906103, | |
| "learning_rate": 3.7729771395385163e-06, | |
| "loss": 0.3471, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 4.351020408163265, | |
| "grad_norm": 0.08900635779184796, | |
| "learning_rate": 3.6437834988132425e-06, | |
| "loss": 0.344, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 4.3619047619047615, | |
| "grad_norm": 0.09076664134467055, | |
| "learning_rate": 3.5167348698357605e-06, | |
| "loss": 0.3463, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.372789115646259, | |
| "grad_norm": 0.09776364783589368, | |
| "learning_rate": 3.391838748457592e-06, | |
| "loss": 0.3436, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 4.383673469387755, | |
| "grad_norm": 0.08753421394783582, | |
| "learning_rate": 3.269102503532655e-06, | |
| "loss": 0.3519, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 4.394557823129252, | |
| "grad_norm": 0.08339034539428186, | |
| "learning_rate": 3.1485333764824212e-06, | |
| "loss": 0.3391, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 4.405442176870748, | |
| "grad_norm": 0.08797062668746661, | |
| "learning_rate": 3.0301384808687274e-06, | |
| "loss": 0.3473, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 4.416326530612245, | |
| "grad_norm": 0.08350010520624601, | |
| "learning_rate": 2.913924801974033e-06, | |
| "loss": 0.3398, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 4.427210884353742, | |
| "grad_norm": 0.08710062739995046, | |
| "learning_rate": 2.799899196389353e-06, | |
| "loss": 0.343, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 4.438095238095238, | |
| "grad_norm": 0.08534861791323987, | |
| "learning_rate": 2.688068391609626e-06, | |
| "loss": 0.3363, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 4.448979591836735, | |
| "grad_norm": 0.09014749439901862, | |
| "learning_rate": 2.5784389856368998e-06, | |
| "loss": 0.342, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 4.459863945578231, | |
| "grad_norm": 0.08443909503900539, | |
| "learning_rate": 2.471017446590951e-06, | |
| "loss": 0.3418, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 4.470748299319728, | |
| "grad_norm": 0.08303643427540018, | |
| "learning_rate": 2.365810112327753e-06, | |
| "loss": 0.3488, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 4.481632653061224, | |
| "grad_norm": 0.08389796519011447, | |
| "learning_rate": 2.262823190065473e-06, | |
| "loss": 0.3399, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 4.492517006802721, | |
| "grad_norm": 0.08360010802242693, | |
| "learning_rate": 2.1620627560182815e-06, | |
| "loss": 0.3406, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 4.503401360544218, | |
| "grad_norm": 0.08547740433430215, | |
| "learning_rate": 2.063534755037875e-06, | |
| "loss": 0.3376, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 4.514285714285714, | |
| "grad_norm": 0.08619879618806697, | |
| "learning_rate": 1.967245000262672e-06, | |
| "loss": 0.3399, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 4.525170068027211, | |
| "grad_norm": 0.09482348477356095, | |
| "learning_rate": 1.8731991727749133e-06, | |
| "loss": 0.3399, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 4.536054421768707, | |
| "grad_norm": 0.0866735178858248, | |
| "learning_rate": 1.7814028212654078e-06, | |
| "loss": 0.3482, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 4.546938775510204, | |
| "grad_norm": 0.0883733178691092, | |
| "learning_rate": 1.6918613617061997e-06, | |
| "loss": 0.3489, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 4.557823129251701, | |
| "grad_norm": 0.08018314461437841, | |
| "learning_rate": 1.604580077031037e-06, | |
| "loss": 0.3422, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 4.568707482993197, | |
| "grad_norm": 0.07859676520534044, | |
| "learning_rate": 1.5195641168236441e-06, | |
| "loss": 0.3336, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 4.579591836734694, | |
| "grad_norm": 0.08484327867196195, | |
| "learning_rate": 1.4368184970139276e-06, | |
| "loss": 0.3386, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.59047619047619, | |
| "grad_norm": 0.09231691136667937, | |
| "learning_rate": 1.35634809958201e-06, | |
| "loss": 0.3416, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 4.6013605442176875, | |
| "grad_norm": 0.08648460698182472, | |
| "learning_rate": 1.2781576722702106e-06, | |
| "loss": 0.3464, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 4.612244897959184, | |
| "grad_norm": 0.0823755612130137, | |
| "learning_rate": 1.2022518283029495e-06, | |
| "loss": 0.3451, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 4.62312925170068, | |
| "grad_norm": 0.08454325641651687, | |
| "learning_rate": 1.128635046114508e-06, | |
| "loss": 0.3427, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 4.634013605442177, | |
| "grad_norm": 0.08135746673539428, | |
| "learning_rate": 1.0573116690848794e-06, | |
| "loss": 0.3485, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 4.644897959183673, | |
| "grad_norm": 0.08262176693120167, | |
| "learning_rate": 9.882859052834326e-07, | |
| "loss": 0.348, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 4.6557823129251705, | |
| "grad_norm": 0.12042028064513204, | |
| "learning_rate": 9.215618272206917e-07, | |
| "loss": 0.3476, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 4.666666666666667, | |
| "grad_norm": 0.0845679088852923, | |
| "learning_rate": 8.571433716080446e-07, | |
| "loss": 0.3398, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 4.677551020408163, | |
| "grad_norm": 0.083021063525801, | |
| "learning_rate": 7.950343391254623e-07, | |
| "loss": 0.3364, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 4.68843537414966, | |
| "grad_norm": 0.07996219621656833, | |
| "learning_rate": 7.352383941972818e-07, | |
| "loss": 0.3429, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 4.699319727891156, | |
| "grad_norm": 0.07824400301840871, | |
| "learning_rate": 6.777590647759846e-07, | |
| "loss": 0.3446, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 4.710204081632653, | |
| "grad_norm": 0.07741287163171348, | |
| "learning_rate": 6.225997421340558e-07, | |
| "loss": 0.3445, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 4.72108843537415, | |
| "grad_norm": 0.07981480666126996, | |
| "learning_rate": 5.697636806639173e-07, | |
| "loss": 0.3445, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 4.7319727891156464, | |
| "grad_norm": 0.08279099258937904, | |
| "learning_rate": 5.192539976858779e-07, | |
| "loss": 0.3425, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 4.742857142857143, | |
| "grad_norm": 0.0831753409761627, | |
| "learning_rate": 4.7107367326426176e-07, | |
| "loss": 0.3432, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 4.753741496598639, | |
| "grad_norm": 0.08354270000701715, | |
| "learning_rate": 4.25225550031545e-07, | |
| "loss": 0.3405, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 4.764625850340136, | |
| "grad_norm": 0.07783160549260915, | |
| "learning_rate": 3.8171233302066293e-07, | |
| "loss": 0.3336, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 4.775510204081632, | |
| "grad_norm": 0.07981736990104178, | |
| "learning_rate": 3.405365895054002e-07, | |
| "loss": 0.3377, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 4.7863945578231295, | |
| "grad_norm": 0.08062382660451918, | |
| "learning_rate": 3.0170074884893874e-07, | |
| "loss": 0.3472, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 4.797278911564626, | |
| "grad_norm": 0.07779156567585847, | |
| "learning_rate": 2.652071023605007e-07, | |
| "loss": 0.3393, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 4.808163265306122, | |
| "grad_norm": 0.08264795496109793, | |
| "learning_rate": 2.3105780316018622e-07, | |
| "loss": 0.3489, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 4.819047619047619, | |
| "grad_norm": 0.07821832548085135, | |
| "learning_rate": 1.9925486605193224e-07, | |
| "loss": 0.3506, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 4.829931972789115, | |
| "grad_norm": 0.08266595280750075, | |
| "learning_rate": 1.6980016740463012e-07, | |
| "loss": 0.349, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 4.840816326530613, | |
| "grad_norm": 0.08083949073024113, | |
| "learning_rate": 1.4269544504142308e-07, | |
| "loss": 0.3479, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 4.851700680272109, | |
| "grad_norm": 0.07762251357306305, | |
| "learning_rate": 1.1794229813718804e-07, | |
| "loss": 0.3442, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 4.862585034013605, | |
| "grad_norm": 0.07806464783206725, | |
| "learning_rate": 9.55421871241713e-08, | |
| "loss": 0.3426, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 4.873469387755102, | |
| "grad_norm": 0.07951754007871666, | |
| "learning_rate": 7.549643360582615e-08, | |
| "loss": 0.3451, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 4.884353741496598, | |
| "grad_norm": 0.0815159225210058, | |
| "learning_rate": 5.7806220278835423e-08, | |
| "loss": 0.3448, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 4.895238095238096, | |
| "grad_norm": 0.07810722165035679, | |
| "learning_rate": 4.247259086334055e-08, | |
| "loss": 0.344, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 4.906122448979592, | |
| "grad_norm": 0.07734161649061848, | |
| "learning_rate": 2.9496450041359702e-08, | |
| "loss": 0.3393, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 4.9170068027210885, | |
| "grad_norm": 0.0771974639166234, | |
| "learning_rate": 1.8878563403408325e-08, | |
| "loss": 0.3481, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 4.927891156462585, | |
| "grad_norm": 0.07825445118858486, | |
| "learning_rate": 1.061955740332632e-08, | |
| "loss": 0.3409, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 4.938775510204081, | |
| "grad_norm": 0.07938774625918343, | |
| "learning_rate": 4.71991932133431e-09, | |
| "loss": 0.348, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 4.949659863945579, | |
| "grad_norm": 0.0903762745926248, | |
| "learning_rate": 1.1799972352521948e-09, | |
| "loss": 0.3391, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 4.960544217687075, | |
| "grad_norm": 0.08057468790850034, | |
| "learning_rate": 0.0, | |
| "loss": 0.3414, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 4.960544217687075, | |
| "step": 455, | |
| "total_flos": 1.1708504273615585e+19, | |
| "train_loss": 0.42467863664522276, | |
| "train_runtime": 101078.8365, | |
| "train_samples_per_second": 2.327, | |
| "train_steps_per_second": 0.005 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 455, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1708504273615585e+19, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |