| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.850632911392405, | |
| "eval_steps": 500, | |
| "global_step": 224, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.030379746835443037, | |
| "grad_norm": 6.4974170666129405, | |
| "learning_rate": 8.695652173913044e-07, | |
| "loss": 0.8286, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.060759493670886074, | |
| "grad_norm": 6.289827405515917, | |
| "learning_rate": 1.7391304347826088e-06, | |
| "loss": 0.8151, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09113924050632911, | |
| "grad_norm": 6.4036402262813406, | |
| "learning_rate": 2.6086956521739132e-06, | |
| "loss": 0.8193, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12151898734177215, | |
| "grad_norm": 5.94576312643178, | |
| "learning_rate": 3.4782608695652175e-06, | |
| "loss": 0.8001, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 4.332743590295384, | |
| "learning_rate": 4.347826086956522e-06, | |
| "loss": 0.7664, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.18227848101265823, | |
| "grad_norm": 2.3167722633137267, | |
| "learning_rate": 5.2173913043478265e-06, | |
| "loss": 0.7308, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21265822784810126, | |
| "grad_norm": 2.0549879974300556, | |
| "learning_rate": 6.086956521739132e-06, | |
| "loss": 0.7232, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.2430379746835443, | |
| "grad_norm": 3.9216268287534466, | |
| "learning_rate": 6.956521739130435e-06, | |
| "loss": 0.7073, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.27341772151898736, | |
| "grad_norm": 4.400648238604796, | |
| "learning_rate": 7.82608695652174e-06, | |
| "loss": 0.7347, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 4.27563822678231, | |
| "learning_rate": 8.695652173913044e-06, | |
| "loss": 0.7053, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3341772151898734, | |
| "grad_norm": 4.025198490705621, | |
| "learning_rate": 9.565217391304349e-06, | |
| "loss": 0.6675, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.36455696202531646, | |
| "grad_norm": 3.420456932222551, | |
| "learning_rate": 1.0434782608695653e-05, | |
| "loss": 0.6849, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3949367088607595, | |
| "grad_norm": 2.1030806427648177, | |
| "learning_rate": 1.1304347826086957e-05, | |
| "loss": 0.6451, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.4253164556962025, | |
| "grad_norm": 1.8697807789878425, | |
| "learning_rate": 1.2173913043478263e-05, | |
| "loss": 0.6094, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 2.26056148981361, | |
| "learning_rate": 1.3043478260869566e-05, | |
| "loss": 0.6282, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4860759493670886, | |
| "grad_norm": 1.5886195054713648, | |
| "learning_rate": 1.391304347826087e-05, | |
| "loss": 0.6003, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5164556962025316, | |
| "grad_norm": 1.1004639599591655, | |
| "learning_rate": 1.4782608695652174e-05, | |
| "loss": 0.5957, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5468354430379747, | |
| "grad_norm": 1.097476031820012, | |
| "learning_rate": 1.565217391304348e-05, | |
| "loss": 0.5666, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5772151898734177, | |
| "grad_norm": 1.0052979984531278, | |
| "learning_rate": 1.6521739130434785e-05, | |
| "loss": 0.5719, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 0.9894560127963354, | |
| "learning_rate": 1.739130434782609e-05, | |
| "loss": 0.5652, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6379746835443038, | |
| "grad_norm": 0.9132016901726216, | |
| "learning_rate": 1.8260869565217393e-05, | |
| "loss": 0.5456, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6683544303797468, | |
| "grad_norm": 0.8028655023056698, | |
| "learning_rate": 1.9130434782608697e-05, | |
| "loss": 0.5328, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.6987341772151898, | |
| "grad_norm": 0.9425641229106339, | |
| "learning_rate": 2e-05, | |
| "loss": 0.5481, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7291139240506329, | |
| "grad_norm": 0.7611348569787119, | |
| "learning_rate": 1.999877856940653e-05, | |
| "loss": 0.5592, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 0.7564654959636459, | |
| "learning_rate": 1.999511457600466e-05, | |
| "loss": 0.5466, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.789873417721519, | |
| "grad_norm": 0.7930008987865393, | |
| "learning_rate": 1.9989008914857115e-05, | |
| "loss": 0.535, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.8202531645569621, | |
| "grad_norm": 0.7493626663234454, | |
| "learning_rate": 1.998046307749216e-05, | |
| "loss": 0.528, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.850632911392405, | |
| "grad_norm": 0.6101946445971717, | |
| "learning_rate": 1.9969479151539238e-05, | |
| "loss": 0.5212, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8810126582278481, | |
| "grad_norm": 0.5769806586436026, | |
| "learning_rate": 1.9956059820218982e-05, | |
| "loss": 0.5391, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 0.7157670331581075, | |
| "learning_rate": 1.9940208361687762e-05, | |
| "loss": 0.5242, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9417721518987342, | |
| "grad_norm": 0.4762911173255839, | |
| "learning_rate": 1.9921928648236855e-05, | |
| "loss": 0.5079, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.9721518987341772, | |
| "grad_norm": 0.5944527252102053, | |
| "learning_rate": 1.990122514534651e-05, | |
| "loss": 0.4989, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.010126582278481, | |
| "grad_norm": 0.7298002749171054, | |
| "learning_rate": 1.9878102910595097e-05, | |
| "loss": 0.5064, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.040506329113924, | |
| "grad_norm": 0.5414950063745023, | |
| "learning_rate": 1.985256759242359e-05, | |
| "loss": 0.4756, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.070886075949367, | |
| "grad_norm": 0.6441360723708793, | |
| "learning_rate": 1.982462542875576e-05, | |
| "loss": 0.4914, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.1012658227848102, | |
| "grad_norm": 0.4946596745212841, | |
| "learning_rate": 1.979428324547432e-05, | |
| "loss": 0.4845, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.1316455696202532, | |
| "grad_norm": 0.5093476336294606, | |
| "learning_rate": 1.9761548454753455e-05, | |
| "loss": 0.4774, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.1620253164556962, | |
| "grad_norm": 0.5294711825014295, | |
| "learning_rate": 1.972642905324813e-05, | |
| "loss": 0.465, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.1924050632911392, | |
| "grad_norm": 0.42433857748865206, | |
| "learning_rate": 1.9688933620140638e-05, | |
| "loss": 0.4655, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.2227848101265824, | |
| "grad_norm": 0.5103486723929203, | |
| "learning_rate": 1.96490713150448e-05, | |
| "loss": 0.4768, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 0.49791097543662277, | |
| "learning_rate": 1.9606851875768404e-05, | |
| "loss": 0.4766, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.2835443037974683, | |
| "grad_norm": 0.5283318712004925, | |
| "learning_rate": 1.956228561593441e-05, | |
| "loss": 0.4574, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.3139240506329113, | |
| "grad_norm": 0.47403972426656127, | |
| "learning_rate": 1.9515383422461457e-05, | |
| "loss": 0.4776, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.3443037974683545, | |
| "grad_norm": 0.5069935307311383, | |
| "learning_rate": 1.9466156752904344e-05, | |
| "loss": 0.4764, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.3746835443037975, | |
| "grad_norm": 0.47875932287647166, | |
| "learning_rate": 1.9414617632655114e-05, | |
| "loss": 0.4548, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.4050632911392404, | |
| "grad_norm": 0.48389828010345176, | |
| "learning_rate": 1.9360778652005416e-05, | |
| "loss": 0.472, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.4354430379746836, | |
| "grad_norm": 0.47182918877963814, | |
| "learning_rate": 1.9304652963070868e-05, | |
| "loss": 0.4609, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.4658227848101266, | |
| "grad_norm": 0.5277737630304099, | |
| "learning_rate": 1.9246254276578175e-05, | |
| "loss": 0.4616, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.4962025316455696, | |
| "grad_norm": 0.49037387751432276, | |
| "learning_rate": 1.9185596858515797e-05, | |
| "loss": 0.465, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.5265822784810128, | |
| "grad_norm": 0.4683776025075759, | |
| "learning_rate": 1.9122695526648968e-05, | |
| "loss": 0.4607, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 0.5657198678079827, | |
| "learning_rate": 1.905756564689991e-05, | |
| "loss": 0.475, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.5873417721518988, | |
| "grad_norm": 0.36540383133355375, | |
| "learning_rate": 1.8990223129594146e-05, | |
| "loss": 0.4554, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.6177215189873417, | |
| "grad_norm": 0.44871637377849793, | |
| "learning_rate": 1.8920684425573865e-05, | |
| "loss": 0.4427, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.6481012658227847, | |
| "grad_norm": 0.41443169088823717, | |
| "learning_rate": 1.884896652217917e-05, | |
| "loss": 0.4709, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.678481012658228, | |
| "grad_norm": 0.44779153436928243, | |
| "learning_rate": 1.877508693909831e-05, | |
| "loss": 0.4749, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 0.4247551126128865, | |
| "learning_rate": 1.8699063724087905e-05, | |
| "loss": 0.4765, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.7392405063291139, | |
| "grad_norm": 0.6036395212394368, | |
| "learning_rate": 1.862091544856407e-05, | |
| "loss": 0.4363, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.769620253164557, | |
| "grad_norm": 0.4260481151358374, | |
| "learning_rate": 1.854066120306571e-05, | |
| "loss": 0.4441, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.7549989851715053, | |
| "learning_rate": 1.8458320592590976e-05, | |
| "loss": 0.4621, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.830379746835443, | |
| "grad_norm": 0.5058105533094102, | |
| "learning_rate": 1.837391373180801e-05, | |
| "loss": 0.4423, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 0.5370621851699969, | |
| "learning_rate": 1.8287461240141217e-05, | |
| "loss": 0.4466, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.891139240506329, | |
| "grad_norm": 0.5610764225893673, | |
| "learning_rate": 1.8198984236734246e-05, | |
| "loss": 0.4378, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.9215189873417722, | |
| "grad_norm": 0.5922052596113978, | |
| "learning_rate": 1.8108504335290852e-05, | |
| "loss": 0.4559, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.9518987341772152, | |
| "grad_norm": 0.4739092523369615, | |
| "learning_rate": 1.8016043638794975e-05, | |
| "loss": 0.4448, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.9822784810126581, | |
| "grad_norm": 0.6028554849088902, | |
| "learning_rate": 1.7921624734111292e-05, | |
| "loss": 0.4338, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.020253164556962, | |
| "grad_norm": 0.48528388263771194, | |
| "learning_rate": 1.7825270686467567e-05, | |
| "loss": 0.4168, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.050632911392405, | |
| "grad_norm": 0.4435067765000118, | |
| "learning_rate": 1.7727005033820117e-05, | |
| "loss": 0.4079, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.081012658227848, | |
| "grad_norm": 0.49895054830192304, | |
| "learning_rate": 1.762685178110382e-05, | |
| "loss": 0.3987, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.1113924050632913, | |
| "grad_norm": 0.4840645731023052, | |
| "learning_rate": 1.752483539436807e-05, | |
| "loss": 0.4141, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.141772151898734, | |
| "grad_norm": 0.49161000311716674, | |
| "learning_rate": 1.7420980794800013e-05, | |
| "loss": 0.3821, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.1721518987341772, | |
| "grad_norm": 0.40635171868742836, | |
| "learning_rate": 1.731531335263669e-05, | |
| "loss": 0.41, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.4767019681490631, | |
| "learning_rate": 1.720785888096743e-05, | |
| "loss": 0.4169, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.232911392405063, | |
| "grad_norm": 0.5356414359619475, | |
| "learning_rate": 1.7098643629428035e-05, | |
| "loss": 0.3974, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.2632911392405064, | |
| "grad_norm": 0.44342208441500364, | |
| "learning_rate": 1.698769427778842e-05, | |
| "loss": 0.3927, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.293670886075949, | |
| "grad_norm": 0.4470607689391908, | |
| "learning_rate": 1.687503792943506e-05, | |
| "loss": 0.3973, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.3240506329113924, | |
| "grad_norm": 0.5652174939048651, | |
| "learning_rate": 1.6760702104750046e-05, | |
| "loss": 0.4007, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.3544303797468356, | |
| "grad_norm": 0.480791916654712, | |
| "learning_rate": 1.664471473438822e-05, | |
| "loss": 0.3912, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.3848101265822783, | |
| "grad_norm": 0.49626658108912697, | |
| "learning_rate": 1.6527104152454096e-05, | |
| "loss": 0.3843, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.4151898734177215, | |
| "grad_norm": 0.5356897609748498, | |
| "learning_rate": 1.6407899089580263e-05, | |
| "loss": 0.3882, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.4455696202531647, | |
| "grad_norm": 0.49760734807745505, | |
| "learning_rate": 1.628712866590885e-05, | |
| "loss": 0.3827, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.4759493670886075, | |
| "grad_norm": 0.5231652008328098, | |
| "learning_rate": 1.6164822383977912e-05, | |
| "loss": 0.3951, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.5159729083805341, | |
| "learning_rate": 1.604101012151436e-05, | |
| "loss": 0.3873, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.536708860759494, | |
| "grad_norm": 0.5272589632436286, | |
| "learning_rate": 1.5915722124135227e-05, | |
| "loss": 0.3871, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.5670886075949366, | |
| "grad_norm": 0.4764387228380176, | |
| "learning_rate": 1.5788988997959115e-05, | |
| "loss": 0.3861, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.59746835443038, | |
| "grad_norm": 0.4669252181715756, | |
| "learning_rate": 1.5660841702129533e-05, | |
| "loss": 0.3795, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.6278481012658226, | |
| "grad_norm": 0.4012441103446406, | |
| "learning_rate": 1.5531311541251995e-05, | |
| "loss": 0.3853, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 0.4435465833041234, | |
| "learning_rate": 1.540043015774676e-05, | |
| "loss": 0.3873, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.688607594936709, | |
| "grad_norm": 0.3843356668548089, | |
| "learning_rate": 1.5268229524119007e-05, | |
| "loss": 0.3907, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.7189873417721517, | |
| "grad_norm": 0.4441580848844239, | |
| "learning_rate": 1.513474193514842e-05, | |
| "loss": 0.41, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.749367088607595, | |
| "grad_norm": 0.34006736698576706, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.3832, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.779746835443038, | |
| "grad_norm": 0.39379243235273864, | |
| "learning_rate": 1.4864036634258112e-05, | |
| "loss": 0.3848, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.3523668648983496, | |
| "learning_rate": 1.4726885051885654e-05, | |
| "loss": 0.3863, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.840506329113924, | |
| "grad_norm": 0.3983109819934856, | |
| "learning_rate": 1.4588578757110359e-05, | |
| "loss": 0.3863, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.8708860759493673, | |
| "grad_norm": 0.380540221826223, | |
| "learning_rate": 1.4449151536240167e-05, | |
| "loss": 0.3779, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.90126582278481, | |
| "grad_norm": 0.41734016320094736, | |
| "learning_rate": 1.4308637449409705e-05, | |
| "loss": 0.3867, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.9316455696202532, | |
| "grad_norm": 0.38016674499407604, | |
| "learning_rate": 1.4167070822259868e-05, | |
| "loss": 0.3889, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "grad_norm": 0.4045997822103201, | |
| "learning_rate": 1.402448623755254e-05, | |
| "loss": 0.3877, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.992405063291139, | |
| "grad_norm": 0.3698254473141723, | |
| "learning_rate": 1.3880918526722497e-05, | |
| "loss": 0.3825, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.030379746835443, | |
| "grad_norm": 0.522327389194682, | |
| "learning_rate": 1.3736402761368597e-05, | |
| "loss": 0.3446, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.060759493670886, | |
| "grad_norm": 0.48797768270381986, | |
| "learning_rate": 1.3590974244686248e-05, | |
| "loss": 0.3323, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.091139240506329, | |
| "grad_norm": 0.5317349524299804, | |
| "learning_rate": 1.344466850284333e-05, | |
| "loss": 0.3434, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.1215189873417724, | |
| "grad_norm": 0.5825524962365572, | |
| "learning_rate": 1.3297521276301666e-05, | |
| "loss": 0.3381, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.4775943015873331, | |
| "learning_rate": 1.3149568511086104e-05, | |
| "loss": 0.3211, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.1822784810126583, | |
| "grad_norm": 0.37088721176541106, | |
| "learning_rate": 1.300084635000341e-05, | |
| "loss": 0.3243, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.212658227848101, | |
| "grad_norm": 0.524410882067887, | |
| "learning_rate": 1.2851391123813075e-05, | |
| "loss": 0.3433, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.2430379746835443, | |
| "grad_norm": 0.42287255582884636, | |
| "learning_rate": 1.2701239342352223e-05, | |
| "loss": 0.3391, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.2734177215189875, | |
| "grad_norm": 0.4554593428965173, | |
| "learning_rate": 1.2550427685616767e-05, | |
| "loss": 0.3403, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.3037974683544302, | |
| "grad_norm": 0.39210904323253903, | |
| "learning_rate": 1.239899299480098e-05, | |
| "loss": 0.3509, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.3341772151898734, | |
| "grad_norm": 0.43734491288996724, | |
| "learning_rate": 1.2246972263297718e-05, | |
| "loss": 0.3363, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.3645569620253166, | |
| "grad_norm": 0.4047367761425651, | |
| "learning_rate": 1.2094402627661447e-05, | |
| "loss": 0.3402, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.3949367088607594, | |
| "grad_norm": 0.3937209895604764, | |
| "learning_rate": 1.1941321358536278e-05, | |
| "loss": 0.3433, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.4253164556962026, | |
| "grad_norm": 0.38313925509714003, | |
| "learning_rate": 1.1787765851551296e-05, | |
| "loss": 0.323, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.4113012036096517, | |
| "learning_rate": 1.1633773618185302e-05, | |
| "loss": 0.3313, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.4860759493670885, | |
| "grad_norm": 0.35108239034348615, | |
| "learning_rate": 1.14793822766033e-05, | |
| "loss": 0.3366, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.5164556962025317, | |
| "grad_norm": 0.35242823013256014, | |
| "learning_rate": 1.132462954246688e-05, | |
| "loss": 0.3212, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.546835443037975, | |
| "grad_norm": 0.35770624011974284, | |
| "learning_rate": 1.1169553219720828e-05, | |
| "loss": 0.3126, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 3.5772151898734177, | |
| "grad_norm": 0.3131897689300393, | |
| "learning_rate": 1.1014191191358118e-05, | |
| "loss": 0.3263, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 3.607594936708861, | |
| "grad_norm": 0.31506231194363454, | |
| "learning_rate": 1.085858141016566e-05, | |
| "loss": 0.3194, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.6379746835443036, | |
| "grad_norm": 0.3300433520032172, | |
| "learning_rate": 1.070276188945293e-05, | |
| "loss": 0.3277, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.668354430379747, | |
| "grad_norm": 0.35857960912106696, | |
| "learning_rate": 1.0546770693765859e-05, | |
| "loss": 0.339, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.69873417721519, | |
| "grad_norm": 0.3472575557331752, | |
| "learning_rate": 1.0390645929588197e-05, | |
| "loss": 0.3265, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.729113924050633, | |
| "grad_norm": 0.3593221910905371, | |
| "learning_rate": 1.0234425736032607e-05, | |
| "loss": 0.3361, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.35323858073043973, | |
| "learning_rate": 1.007814827552384e-05, | |
| "loss": 0.3197, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.7898734177215188, | |
| "grad_norm": 0.3358027219446924, | |
| "learning_rate": 9.92185172447616e-06, | |
| "loss": 0.3371, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.820253164556962, | |
| "grad_norm": 0.33462898097721216, | |
| "learning_rate": 9.765574263967397e-06, | |
| "loss": 0.3444, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.850632911392405, | |
| "grad_norm": 0.3146369213691247, | |
| "learning_rate": 9.609354070411807e-06, | |
| "loss": 0.329, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.8810126582278484, | |
| "grad_norm": 0.3424174883046021, | |
| "learning_rate": 9.453229306234143e-06, | |
| "loss": 0.3372, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.911392405063291, | |
| "grad_norm": 0.2943707886567511, | |
| "learning_rate": 9.297238110547075e-06, | |
| "loss": 0.3402, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.9417721518987343, | |
| "grad_norm": 0.35308049321203056, | |
| "learning_rate": 9.14141858983434e-06, | |
| "loss": 0.3327, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.972151898734177, | |
| "grad_norm": 0.2794126776921179, | |
| "learning_rate": 8.985808808641883e-06, | |
| "loss": 0.3294, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.010126582278481, | |
| "grad_norm": 0.34707540544255155, | |
| "learning_rate": 8.830446780279175e-06, | |
| "loss": 0.3366, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.040506329113924, | |
| "grad_norm": 0.4310342590093157, | |
| "learning_rate": 8.675370457533122e-06, | |
| "loss": 0.2885, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.0708860759493675, | |
| "grad_norm": 0.39223005192351007, | |
| "learning_rate": 8.520617723396702e-06, | |
| "loss": 0.3018, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.5008029874507762, | |
| "learning_rate": 8.366226381814698e-06, | |
| "loss": 0.2917, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.131645569620253, | |
| "grad_norm": 0.3717883851042529, | |
| "learning_rate": 8.212234148448708e-06, | |
| "loss": 0.2948, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.162025316455696, | |
| "grad_norm": 0.48772070261352396, | |
| "learning_rate": 8.058678641463724e-06, | |
| "loss": 0.2856, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.192405063291139, | |
| "grad_norm": 0.29990235187784275, | |
| "learning_rate": 7.905597372338558e-06, | |
| "loss": 0.299, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.222784810126583, | |
| "grad_norm": 0.3646267826485481, | |
| "learning_rate": 7.753027736702283e-06, | |
| "loss": 0.283, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.253164556962025, | |
| "grad_norm": 0.3749448960383453, | |
| "learning_rate": 7.601007005199022e-06, | |
| "loss": 0.2815, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.283544303797468, | |
| "grad_norm": 0.29510288215437896, | |
| "learning_rate": 7.449572314383237e-06, | |
| "loss": 0.2778, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.313924050632911, | |
| "grad_norm": 0.3421691653275206, | |
| "learning_rate": 7.298760657647779e-06, | |
| "loss": 0.2998, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.3443037974683545, | |
| "grad_norm": 0.34243454860761985, | |
| "learning_rate": 7.148608876186931e-06, | |
| "loss": 0.2765, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.374683544303798, | |
| "grad_norm": 0.3114696535262121, | |
| "learning_rate": 6.999153649996595e-06, | |
| "loss": 0.2821, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.3663624838947982, | |
| "learning_rate": 6.8504314889138956e-06, | |
| "loss": 0.2919, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 4.435443037974683, | |
| "grad_norm": 0.3094633405060417, | |
| "learning_rate": 6.702478723698336e-06, | |
| "loss": 0.2692, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.465822784810126, | |
| "grad_norm": 0.30967837068990584, | |
| "learning_rate": 6.555331497156671e-06, | |
| "loss": 0.2862, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 4.49620253164557, | |
| "grad_norm": 0.36674001741323564, | |
| "learning_rate": 6.4090257553137566e-06, | |
| "loss": 0.2867, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 4.526582278481013, | |
| "grad_norm": 0.300696126518697, | |
| "learning_rate": 6.263597238631405e-06, | |
| "loss": 0.2931, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 4.556962025316456, | |
| "grad_norm": 0.2969652346854883, | |
| "learning_rate": 6.119081473277502e-06, | |
| "loss": 0.2894, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 4.587341772151898, | |
| "grad_norm": 0.31128084271919876, | |
| "learning_rate": 5.975513762447465e-06, | |
| "loss": 0.2891, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.6177215189873415, | |
| "grad_norm": 0.29739493327330657, | |
| "learning_rate": 5.832929177740134e-06, | |
| "loss": 0.2857, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 4.648101265822785, | |
| "grad_norm": 0.2867071618954023, | |
| "learning_rate": 5.6913625505902966e-06, | |
| "loss": 0.2928, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 4.678481012658228, | |
| "grad_norm": 0.2886542058623698, | |
| "learning_rate": 5.550848463759835e-06, | |
| "loss": 0.2762, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.2909709652558278, | |
| "learning_rate": 5.411421242889643e-06, | |
| "loss": 0.2836, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 4.739240506329114, | |
| "grad_norm": 0.28437330504528613, | |
| "learning_rate": 5.273114948114346e-06, | |
| "loss": 0.273, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.769620253164557, | |
| "grad_norm": 0.2639747513769549, | |
| "learning_rate": 5.135963365741892e-06, | |
| "loss": 0.2667, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.29796444837888897, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.2814, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 4.830379746835443, | |
| "grad_norm": 0.304854313462495, | |
| "learning_rate": 4.865258064851579e-06, | |
| "loss": 0.2823, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.860759493670886, | |
| "grad_norm": 0.30257719321842363, | |
| "learning_rate": 4.731770475880995e-06, | |
| "loss": 0.2819, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.891139240506329, | |
| "grad_norm": 0.2996911970589863, | |
| "learning_rate": 4.599569842253244e-06, | |
| "loss": 0.2869, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.921518987341772, | |
| "grad_norm": 0.29799634956925103, | |
| "learning_rate": 4.468688458748006e-06, | |
| "loss": 0.2866, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 4.951898734177215, | |
| "grad_norm": 0.29617907360825707, | |
| "learning_rate": 4.339158297870469e-06, | |
| "loss": 0.2728, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 4.982278481012658, | |
| "grad_norm": 0.29056554075558516, | |
| "learning_rate": 4.211011002040885e-06, | |
| "loss": 0.2997, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 5.020253164556962, | |
| "grad_norm": 0.3357348394161449, | |
| "learning_rate": 4.084277875864776e-06, | |
| "loss": 0.2409, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.3261604322472783, | |
| "learning_rate": 3.958989878485644e-06, | |
| "loss": 0.2536, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.0810126582278485, | |
| "grad_norm": 0.2808339100023715, | |
| "learning_rate": 3.83517761602209e-06, | |
| "loss": 0.2539, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 5.111392405063291, | |
| "grad_norm": 0.40913027095598886, | |
| "learning_rate": 3.712871334091154e-06, | |
| "loss": 0.2617, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 5.141772151898734, | |
| "grad_norm": 0.3918467633543267, | |
| "learning_rate": 3.592100910419738e-06, | |
| "loss": 0.2438, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 5.172151898734177, | |
| "grad_norm": 0.3107419405776118, | |
| "learning_rate": 3.4728958475459052e-06, | |
| "loss": 0.2545, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 5.2025316455696204, | |
| "grad_norm": 0.34048246278427013, | |
| "learning_rate": 3.355285265611784e-06, | |
| "loss": 0.2483, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 5.232911392405064, | |
| "grad_norm": 0.3639824877961244, | |
| "learning_rate": 3.2392978952499553e-06, | |
| "loss": 0.2427, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 5.263291139240506, | |
| "grad_norm": 0.2715170371452864, | |
| "learning_rate": 3.1249620705649417e-06, | |
| "loss": 0.249, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 5.293670886075949, | |
| "grad_norm": 0.284255183952654, | |
| "learning_rate": 3.0123057222115835e-06, | |
| "loss": 0.2598, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 5.324050632911392, | |
| "grad_norm": 0.33462942711759586, | |
| "learning_rate": 2.9013563705719673e-06, | |
| "loss": 0.2496, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.31106129770475943, | |
| "learning_rate": 2.7921411190325753e-06, | |
| "loss": 0.2541, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.384810126582279, | |
| "grad_norm": 0.2803838122944453, | |
| "learning_rate": 2.6846866473633126e-06, | |
| "loss": 0.2406, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 5.415189873417722, | |
| "grad_norm": 0.276815621879905, | |
| "learning_rate": 2.579019205199992e-06, | |
| "loss": 0.2406, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 5.445569620253164, | |
| "grad_norm": 0.29344147429774503, | |
| "learning_rate": 2.4751646056319334e-06, | |
| "loss": 0.2547, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 5.4759493670886075, | |
| "grad_norm": 0.2659294541117492, | |
| "learning_rate": 2.373148218896182e-06, | |
| "loss": 0.2519, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 5.506329113924051, | |
| "grad_norm": 0.2842720165003485, | |
| "learning_rate": 2.2729949661798876e-06, | |
| "loss": 0.2492, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.536708860759494, | |
| "grad_norm": 0.2588045078733543, | |
| "learning_rate": 2.174729313532433e-06, | |
| "loss": 0.2398, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 5.567088607594937, | |
| "grad_norm": 0.28715393265858163, | |
| "learning_rate": 2.078375265888707e-06, | |
| "loss": 0.2553, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 5.597468354430379, | |
| "grad_norm": 0.28745554828704595, | |
| "learning_rate": 1.9839563612050273e-06, | |
| "loss": 0.2471, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 5.627848101265823, | |
| "grad_norm": 0.2899226041664699, | |
| "learning_rate": 1.8914956647091497e-06, | |
| "loss": 0.2481, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.2590067155318091, | |
| "learning_rate": 1.8010157632657544e-06, | |
| "loss": 0.2675, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 5.688607594936709, | |
| "grad_norm": 0.25687756224345437, | |
| "learning_rate": 1.7125387598587862e-06, | |
| "loss": 0.2535, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 5.718987341772152, | |
| "grad_norm": 0.2595077909838717, | |
| "learning_rate": 1.6260862681919965e-06, | |
| "loss": 0.2547, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 5.749367088607595, | |
| "grad_norm": 0.2689692147004909, | |
| "learning_rate": 1.5416794074090258e-06, | |
| "loss": 0.2508, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 5.779746835443038, | |
| "grad_norm": 0.26248679544653875, | |
| "learning_rate": 1.459338796934293e-06, | |
| "loss": 0.2527, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 5.810126582278481, | |
| "grad_norm": 0.2531892479672906, | |
| "learning_rate": 1.3790845514359363e-06, | |
| "loss": 0.2558, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.840506329113924, | |
| "grad_norm": 0.24298820143174016, | |
| "learning_rate": 1.300936275912098e-06, | |
| "loss": 0.2535, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 5.870886075949367, | |
| "grad_norm": 0.23720099457299162, | |
| "learning_rate": 1.224913060901688e-06, | |
| "loss": 0.2516, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 5.9012658227848105, | |
| "grad_norm": 0.2478735432227225, | |
| "learning_rate": 1.1510334778208332e-06, | |
| "loss": 0.2643, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 5.931645569620253, | |
| "grad_norm": 0.2512383888339122, | |
| "learning_rate": 1.0793155744261352e-06, | |
| "loss": 0.2604, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.24730027387913064, | |
| "learning_rate": 1.0097768704058542e-06, | |
| "loss": 0.2662, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 5.992405063291139, | |
| "grad_norm": 0.24325883427668096, | |
| "learning_rate": 9.424343531000968e-07, | |
| "loss": 0.2403, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 6.030379746835443, | |
| "grad_norm": 0.31211263509749093, | |
| "learning_rate": 8.773044733510338e-07, | |
| "loss": 0.2362, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 6.060759493670886, | |
| "grad_norm": 0.2737512885730665, | |
| "learning_rate": 8.144031414842012e-07, | |
| "loss": 0.2429, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 6.091139240506329, | |
| "grad_norm": 0.25911659980255664, | |
| "learning_rate": 7.537457234218271e-07, | |
| "loss": 0.2328, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 6.121518987341772, | |
| "grad_norm": 0.23760901390554406, | |
| "learning_rate": 6.953470369291349e-07, | |
| "loss": 0.2306, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.151898734177215, | |
| "grad_norm": 0.23753911038915718, | |
| "learning_rate": 6.392213479945852e-07, | |
| "loss": 0.2389, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 6.182278481012658, | |
| "grad_norm": 0.23552145755488546, | |
| "learning_rate": 5.853823673448877e-07, | |
| "loss": 0.2388, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 6.2126582278481015, | |
| "grad_norm": 0.2442014087456443, | |
| "learning_rate": 5.33843247095659e-07, | |
| "loss": 0.221, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 6.243037974683545, | |
| "grad_norm": 0.2696456418903012, | |
| "learning_rate": 4.846165775385459e-07, | |
| "loss": 0.2452, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 6.273417721518987, | |
| "grad_norm": 0.2811169402953374, | |
| "learning_rate": 4.3771438406559173e-07, | |
| "loss": 0.2407, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.2484087773457379, | |
| "learning_rate": 3.931481242315993e-07, | |
| "loss": 0.2414, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 6.334177215189873, | |
| "grad_norm": 0.26911811496038707, | |
| "learning_rate": 3.5092868495520294e-07, | |
| "loss": 0.2494, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 6.364556962025317, | |
| "grad_norm": 0.23156443461752707, | |
| "learning_rate": 3.110663798593616e-07, | |
| "loss": 0.2438, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 6.39493670886076, | |
| "grad_norm": 0.22517950714907936, | |
| "learning_rate": 2.735709467518699e-07, | |
| "loss": 0.2355, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 6.425316455696202, | |
| "grad_norm": 0.23805252833253757, | |
| "learning_rate": 2.384515452465475e-07, | |
| "loss": 0.2342, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.455696202531645, | |
| "grad_norm": 0.22792456827183868, | |
| "learning_rate": 2.0571675452567997e-07, | |
| "loss": 0.2454, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 6.4860759493670885, | |
| "grad_norm": 0.2251059123835004, | |
| "learning_rate": 1.7537457124423896e-07, | |
| "loss": 0.2404, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 6.516455696202532, | |
| "grad_norm": 0.2310710916460582, | |
| "learning_rate": 1.474324075764111e-07, | |
| "loss": 0.2467, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 6.546835443037975, | |
| "grad_norm": 0.2356748861565607, | |
| "learning_rate": 1.2189708940490653e-07, | |
| "loss": 0.2456, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 6.577215189873417, | |
| "grad_norm": 0.24523700563100917, | |
| "learning_rate": 9.877485465349057e-08, | |
| "loss": 0.2296, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.22877897634830702, | |
| "learning_rate": 7.807135176314707e-08, | |
| "loss": 0.2424, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 6.637974683544304, | |
| "grad_norm": 0.22596205092905755, | |
| "learning_rate": 5.979163831223988e-08, | |
| "loss": 0.2435, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 6.668354430379747, | |
| "grad_norm": 0.22901270273269697, | |
| "learning_rate": 4.394017978101905e-08, | |
| "loss": 0.231, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 6.69873417721519, | |
| "grad_norm": 0.23549751073449784, | |
| "learning_rate": 3.0520848460765525e-08, | |
| "loss": 0.226, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 6.729113924050633, | |
| "grad_norm": 0.2167704625820124, | |
| "learning_rate": 1.9536922507841227e-08, | |
| "loss": 0.2313, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 6.759493670886076, | |
| "grad_norm": 0.21833223668838464, | |
| "learning_rate": 1.099108514288627e-08, | |
| "loss": 0.2441, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 6.789873417721519, | |
| "grad_norm": 0.233277070307917, | |
| "learning_rate": 4.885423995341088e-09, | |
| "loss": 0.2409, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 6.820253164556962, | |
| "grad_norm": 0.23319931748778974, | |
| "learning_rate": 1.2214305934699078e-09, | |
| "loss": 0.2305, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 6.850632911392405, | |
| "grad_norm": 0.22865911802144936, | |
| "learning_rate": 0.0, | |
| "loss": 0.2344, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 6.850632911392405, | |
| "step": 224, | |
| "total_flos": 8.149472647147684e+17, | |
| "train_loss": 0.3728780325369111, | |
| "train_runtime": 25402.5047, | |
| "train_samples_per_second": 0.871, | |
| "train_steps_per_second": 0.009 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 224, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "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": 8.149472647147684e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |