| { | |
| "best_global_step": 1445, | |
| "best_metric": 0.8139059304703476, | |
| "best_model_checkpoint": "RALL_RGBCROP_Aug16F-WD001/checkpoint-1445", | |
| "epoch": 8.083477758521086, | |
| "eval_steps": 500, | |
| "global_step": 2601, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0028885037550548816, | |
| "grad_norm": 5.302567481994629, | |
| "learning_rate": 1.2968299711815564e-07, | |
| "loss": 0.7184, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.005777007510109763, | |
| "grad_norm": 6.957106113433838, | |
| "learning_rate": 2.7377521613832854e-07, | |
| "loss": 0.7586, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.008665511265164644, | |
| "grad_norm": 4.999588966369629, | |
| "learning_rate": 4.1786743515850145e-07, | |
| "loss": 0.695, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.011554015020219527, | |
| "grad_norm": 6.24959135055542, | |
| "learning_rate": 5.619596541786745e-07, | |
| "loss": 0.6835, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.014442518775274409, | |
| "grad_norm": 5.431220531463623, | |
| "learning_rate": 7.060518731988474e-07, | |
| "loss": 0.7258, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.01733102253032929, | |
| "grad_norm": 4.120445251464844, | |
| "learning_rate": 8.501440922190203e-07, | |
| "loss": 0.6909, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.020219526285384173, | |
| "grad_norm": 6.550514221191406, | |
| "learning_rate": 9.94236311239193e-07, | |
| "loss": 0.7431, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.023108030040439053, | |
| "grad_norm": 5.101905822753906, | |
| "learning_rate": 1.138328530259366e-06, | |
| "loss": 0.7267, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.025996533795493933, | |
| "grad_norm": 4.306721210479736, | |
| "learning_rate": 1.282420749279539e-06, | |
| "loss": 0.7244, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.028885037550548817, | |
| "grad_norm": 4.6065449714660645, | |
| "learning_rate": 1.426512968299712e-06, | |
| "loss": 0.7197, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0317735413056037, | |
| "grad_norm": 5.6030964851379395, | |
| "learning_rate": 1.5706051873198849e-06, | |
| "loss": 0.656, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.03466204506065858, | |
| "grad_norm": 4.630875110626221, | |
| "learning_rate": 1.7146974063400579e-06, | |
| "loss": 0.7084, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.03755054881571346, | |
| "grad_norm": 5.320244789123535, | |
| "learning_rate": 1.8587896253602309e-06, | |
| "loss": 0.6829, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.040439052570768345, | |
| "grad_norm": 5.340140342712402, | |
| "learning_rate": 2.0028818443804035e-06, | |
| "loss": 0.6657, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.043327556325823226, | |
| "grad_norm": 4.635012149810791, | |
| "learning_rate": 2.1469740634005763e-06, | |
| "loss": 0.6698, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.046216060080878106, | |
| "grad_norm": 4.354490756988525, | |
| "learning_rate": 2.2910662824207495e-06, | |
| "loss": 0.6411, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.049104563835932986, | |
| "grad_norm": 4.6339192390441895, | |
| "learning_rate": 2.4351585014409223e-06, | |
| "loss": 0.6056, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.05199306759098787, | |
| "grad_norm": 5.349210739135742, | |
| "learning_rate": 2.579250720461095e-06, | |
| "loss": 0.6501, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.05488157134604275, | |
| "grad_norm": 4.616723537445068, | |
| "learning_rate": 2.7233429394812683e-06, | |
| "loss": 0.5882, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.057770075101097634, | |
| "grad_norm": 4.99802303314209, | |
| "learning_rate": 2.867435158501441e-06, | |
| "loss": 0.5804, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.060658578856152515, | |
| "grad_norm": 5.823806285858154, | |
| "learning_rate": 3.0115273775216143e-06, | |
| "loss": 0.6305, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.0635470826112074, | |
| "grad_norm": 6.0993781089782715, | |
| "learning_rate": 3.1556195965417867e-06, | |
| "loss": 0.5662, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.06643558636626228, | |
| "grad_norm": 5.885171890258789, | |
| "learning_rate": 3.29971181556196e-06, | |
| "loss": 0.5734, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.06932409012131716, | |
| "grad_norm": 4.648609638214111, | |
| "learning_rate": 3.4438040345821327e-06, | |
| "loss": 0.5429, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.07221259387637204, | |
| "grad_norm": 6.03137731552124, | |
| "learning_rate": 3.587896253602306e-06, | |
| "loss": 0.4988, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.07510109763142692, | |
| "grad_norm": 5.2586989402771, | |
| "learning_rate": 3.7319884726224787e-06, | |
| "loss": 0.5157, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.0779896013864818, | |
| "grad_norm": 5.4461140632629395, | |
| "learning_rate": 3.876080691642652e-06, | |
| "loss": 0.4589, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.08087810514153669, | |
| "grad_norm": 4.95507287979126, | |
| "learning_rate": 4.020172910662825e-06, | |
| "loss": 0.4792, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.08347775852108608, | |
| "eval_accuracy": 0.7157464212678937, | |
| "eval_loss": 0.5589777231216431, | |
| "eval_runtime": 67.5425, | |
| "eval_samples_per_second": 7.24, | |
| "eval_steps_per_second": 0.918, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.0002888503755054, | |
| "grad_norm": 8.520320892333984, | |
| "learning_rate": 4.1642651296829975e-06, | |
| "loss": 0.4655, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.0031773541305604, | |
| "grad_norm": 8.880334854125977, | |
| "learning_rate": 4.30835734870317e-06, | |
| "loss": 0.4589, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0060658578856152, | |
| "grad_norm": 5.965415954589844, | |
| "learning_rate": 4.452449567723343e-06, | |
| "loss": 0.4373, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0089543616406702, | |
| "grad_norm": 7.1385602951049805, | |
| "learning_rate": 4.596541786743517e-06, | |
| "loss": 0.4358, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.011842865395725, | |
| "grad_norm": 5.977208614349365, | |
| "learning_rate": 4.740634005763689e-06, | |
| "loss": 0.3755, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.01473136915078, | |
| "grad_norm": 3.8900434970855713, | |
| "learning_rate": 4.884726224783862e-06, | |
| "loss": 0.3756, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.0176198729058348, | |
| "grad_norm": 11.794221878051758, | |
| "learning_rate": 4.996789727126807e-06, | |
| "loss": 0.3363, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.0205083766608896, | |
| "grad_norm": 6.779726982116699, | |
| "learning_rate": 4.980738362760835e-06, | |
| "loss": 0.3522, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.0233968804159446, | |
| "grad_norm": 3.1879703998565674, | |
| "learning_rate": 4.9646869983948645e-06, | |
| "loss": 0.3296, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.0262853841709993, | |
| "grad_norm": 10.850168228149414, | |
| "learning_rate": 4.948635634028893e-06, | |
| "loss": 0.344, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0291738879260544, | |
| "grad_norm": 7.838939666748047, | |
| "learning_rate": 4.9325842696629215e-06, | |
| "loss": 0.3328, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0320623916811091, | |
| "grad_norm": 4.140768527984619, | |
| "learning_rate": 4.916532905296951e-06, | |
| "loss": 0.2825, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0349508954361641, | |
| "grad_norm": 13.219120979309082, | |
| "learning_rate": 4.90048154093098e-06, | |
| "loss": 0.2849, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.037839399191219, | |
| "grad_norm": 9.75869083404541, | |
| "learning_rate": 4.884430176565008e-06, | |
| "loss": 0.2791, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.040727902946274, | |
| "grad_norm": 12.4995756149292, | |
| "learning_rate": 4.868378812199037e-06, | |
| "loss": 0.2865, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.0436164067013287, | |
| "grad_norm": 2.0662763118743896, | |
| "learning_rate": 4.8523274478330665e-06, | |
| "loss": 0.2121, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0465049104563835, | |
| "grad_norm": 8.064370155334473, | |
| "learning_rate": 4.836276083467095e-06, | |
| "loss": 0.3064, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.0493934142114385, | |
| "grad_norm": 2.636781930923462, | |
| "learning_rate": 4.820224719101124e-06, | |
| "loss": 0.1826, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.0522819179664933, | |
| "grad_norm": 9.863945960998535, | |
| "learning_rate": 4.804173354735153e-06, | |
| "loss": 0.2448, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0551704217215483, | |
| "grad_norm": 6.096480846405029, | |
| "learning_rate": 4.788121990369181e-06, | |
| "loss": 0.2191, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.058058925476603, | |
| "grad_norm": 6.044295310974121, | |
| "learning_rate": 4.772070626003211e-06, | |
| "loss": 0.3196, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.060947429231658, | |
| "grad_norm": 5.9240803718566895, | |
| "learning_rate": 4.75601926163724e-06, | |
| "loss": 0.2525, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0638359329867129, | |
| "grad_norm": 3.8338232040405273, | |
| "learning_rate": 4.7399678972712685e-06, | |
| "loss": 0.1739, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0667244367417679, | |
| "grad_norm": 13.795681953430176, | |
| "learning_rate": 4.723916532905297e-06, | |
| "loss": 0.2939, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0696129404968227, | |
| "grad_norm": 4.1247663497924805, | |
| "learning_rate": 4.707865168539326e-06, | |
| "loss": 0.1571, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.0725014442518774, | |
| "grad_norm": 8.093220710754395, | |
| "learning_rate": 4.691813804173355e-06, | |
| "loss": 0.1902, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.0753899480069324, | |
| "grad_norm": 6.401426315307617, | |
| "learning_rate": 4.675762439807384e-06, | |
| "loss": 0.2347, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.0782784517619872, | |
| "grad_norm": 4.840950012207031, | |
| "learning_rate": 4.659711075441413e-06, | |
| "loss": 0.1322, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.0811669555170422, | |
| "grad_norm": 10.177003860473633, | |
| "learning_rate": 4.643659711075442e-06, | |
| "loss": 0.2392, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.0834777585210862, | |
| "eval_accuracy": 0.7955010224948875, | |
| "eval_loss": 0.5009445548057556, | |
| "eval_runtime": 67.0386, | |
| "eval_samples_per_second": 7.294, | |
| "eval_steps_per_second": 0.925, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.000577700751011, | |
| "grad_norm": 5.978249549865723, | |
| "learning_rate": 4.6276083467094705e-06, | |
| "loss": 0.1546, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.003466204506066, | |
| "grad_norm": 7.84554386138916, | |
| "learning_rate": 4.6115569823435e-06, | |
| "loss": 0.1298, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.006354708261121, | |
| "grad_norm": 16.169857025146484, | |
| "learning_rate": 4.595505617977528e-06, | |
| "loss": 0.2104, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.009243212016176, | |
| "grad_norm": 7.568899154663086, | |
| "learning_rate": 4.579454253611557e-06, | |
| "loss": 0.1345, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.0121317157712304, | |
| "grad_norm": 1.727231740951538, | |
| "learning_rate": 4.563402889245586e-06, | |
| "loss": 0.128, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0150202195262854, | |
| "grad_norm": 8.80377197265625, | |
| "learning_rate": 4.5473515248796155e-06, | |
| "loss": 0.123, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0179087232813404, | |
| "grad_norm": 12.737110137939453, | |
| "learning_rate": 4.531300160513644e-06, | |
| "loss": 0.0714, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.020797227036395, | |
| "grad_norm": 1.612018346786499, | |
| "learning_rate": 4.515248796147673e-06, | |
| "loss": 0.1137, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.02368573079145, | |
| "grad_norm": 3.954007625579834, | |
| "learning_rate": 4.499197431781702e-06, | |
| "loss": 0.1542, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.026574234546505, | |
| "grad_norm": 12.517119407653809, | |
| "learning_rate": 4.48314606741573e-06, | |
| "loss": 0.1664, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.02946273830156, | |
| "grad_norm": 10.5608549118042, | |
| "learning_rate": 4.46709470304976e-06, | |
| "loss": 0.1145, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.0323512420566145, | |
| "grad_norm": 0.9301741123199463, | |
| "learning_rate": 4.451043338683789e-06, | |
| "loss": 0.0606, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.0352397458116696, | |
| "grad_norm": 6.918083667755127, | |
| "learning_rate": 4.4349919743178176e-06, | |
| "loss": 0.08, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.0381282495667246, | |
| "grad_norm": 12.93762493133545, | |
| "learning_rate": 4.418940609951846e-06, | |
| "loss": 0.0869, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.041016753321779, | |
| "grad_norm": 0.3353147506713867, | |
| "learning_rate": 4.402889245585875e-06, | |
| "loss": 0.0313, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.043905257076834, | |
| "grad_norm": 0.9309785962104797, | |
| "learning_rate": 4.386837881219904e-06, | |
| "loss": 0.1997, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.046793760831889, | |
| "grad_norm": 4.111329078674316, | |
| "learning_rate": 4.370786516853933e-06, | |
| "loss": 0.1582, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.049682264586944, | |
| "grad_norm": 1.4403131008148193, | |
| "learning_rate": 4.354735152487962e-06, | |
| "loss": 0.0591, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.0525707683419987, | |
| "grad_norm": 11.65287971496582, | |
| "learning_rate": 4.33868378812199e-06, | |
| "loss": 0.0663, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.0554592720970537, | |
| "grad_norm": 5.396105766296387, | |
| "learning_rate": 4.32263242375602e-06, | |
| "loss": 0.0932, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.0583477758521087, | |
| "grad_norm": 31.526716232299805, | |
| "learning_rate": 4.306581059390049e-06, | |
| "loss": 0.1193, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.0612362796071633, | |
| "grad_norm": 0.1906011402606964, | |
| "learning_rate": 4.290529695024077e-06, | |
| "loss": 0.0971, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.0641247833622183, | |
| "grad_norm": 0.8602182865142822, | |
| "learning_rate": 4.274478330658106e-06, | |
| "loss": 0.0882, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.0670132871172733, | |
| "grad_norm": 6.391236305236816, | |
| "learning_rate": 4.258426966292135e-06, | |
| "loss": 0.2185, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.0699017908723283, | |
| "grad_norm": 25.014951705932617, | |
| "learning_rate": 4.242375601926164e-06, | |
| "loss": 0.0601, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.072790294627383, | |
| "grad_norm": 6.7296247482299805, | |
| "learning_rate": 4.226324237560193e-06, | |
| "loss": 0.0896, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.075678798382438, | |
| "grad_norm": 6.794140815734863, | |
| "learning_rate": 4.210272873194222e-06, | |
| "loss": 0.1055, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.078567302137493, | |
| "grad_norm": 0.9049288630485535, | |
| "learning_rate": 4.194221508828251e-06, | |
| "loss": 0.0245, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.081455805892548, | |
| "grad_norm": 11.116994857788086, | |
| "learning_rate": 4.1781701444622794e-06, | |
| "loss": 0.0414, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.083477758521086, | |
| "eval_accuracy": 0.8016359918200409, | |
| "eval_loss": 0.6814729571342468, | |
| "eval_runtime": 61.2995, | |
| "eval_samples_per_second": 7.977, | |
| "eval_steps_per_second": 1.011, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 3.0008665511265167, | |
| "grad_norm": 6.378897666931152, | |
| "learning_rate": 4.162118780096309e-06, | |
| "loss": 0.1386, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.0037550548815712, | |
| "grad_norm": 0.5761991143226624, | |
| "learning_rate": 4.146067415730337e-06, | |
| "loss": 0.0811, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.0066435586366262, | |
| "grad_norm": 3.8463475704193115, | |
| "learning_rate": 4.130016051364366e-06, | |
| "loss": 0.0074, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.0095320623916813, | |
| "grad_norm": 4.725830078125, | |
| "learning_rate": 4.113964686998395e-06, | |
| "loss": 0.0338, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.012420566146736, | |
| "grad_norm": 15.801431655883789, | |
| "learning_rate": 4.0979133226324245e-06, | |
| "loss": 0.076, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.015309069901791, | |
| "grad_norm": 3.2666287422180176, | |
| "learning_rate": 4.081861958266453e-06, | |
| "loss": 0.0165, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.018197573656846, | |
| "grad_norm": 0.23704250156879425, | |
| "learning_rate": 4.0658105939004815e-06, | |
| "loss": 0.0655, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.021086077411901, | |
| "grad_norm": 0.14014659821987152, | |
| "learning_rate": 4.049759229534511e-06, | |
| "loss": 0.0395, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.0239745811669554, | |
| "grad_norm": 0.4789138436317444, | |
| "learning_rate": 4.033707865168539e-06, | |
| "loss": 0.016, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.0268630849220104, | |
| "grad_norm": 0.08817695081233978, | |
| "learning_rate": 4.017656500802569e-06, | |
| "loss": 0.0169, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.0297515886770654, | |
| "grad_norm": 1.1655572652816772, | |
| "learning_rate": 4.001605136436598e-06, | |
| "loss": 0.0705, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.03264009243212, | |
| "grad_norm": 1.2523001432418823, | |
| "learning_rate": 3.9855537720706265e-06, | |
| "loss": 0.0077, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.035528596187175, | |
| "grad_norm": 0.08337568491697311, | |
| "learning_rate": 3.969502407704655e-06, | |
| "loss": 0.0053, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.03841709994223, | |
| "grad_norm": 0.1509549468755722, | |
| "learning_rate": 3.953451043338684e-06, | |
| "loss": 0.0418, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.041305603697285, | |
| "grad_norm": 0.20127534866333008, | |
| "learning_rate": 3.937399678972713e-06, | |
| "loss": 0.0123, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.0441941074523395, | |
| "grad_norm": 0.7579932808876038, | |
| "learning_rate": 3.921348314606742e-06, | |
| "loss": 0.025, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.0470826112073945, | |
| "grad_norm": 0.09110520780086517, | |
| "learning_rate": 3.905296950240771e-06, | |
| "loss": 0.0108, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.0499711149624495, | |
| "grad_norm": 1.204427719116211, | |
| "learning_rate": 3.8892455858748e-06, | |
| "loss": 0.0165, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.0528596187175046, | |
| "grad_norm": 0.6164229512214661, | |
| "learning_rate": 3.8731942215088285e-06, | |
| "loss": 0.0018, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.055748122472559, | |
| "grad_norm": 0.09581731259822845, | |
| "learning_rate": 3.857142857142858e-06, | |
| "loss": 0.0659, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.058636626227614, | |
| "grad_norm": 0.09253386408090591, | |
| "learning_rate": 3.841091492776886e-06, | |
| "loss": 0.0227, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.061525129982669, | |
| "grad_norm": 0.06041618436574936, | |
| "learning_rate": 3.825040128410915e-06, | |
| "loss": 0.0343, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.0644136337377237, | |
| "grad_norm": 0.0652775913476944, | |
| "learning_rate": 3.808988764044944e-06, | |
| "loss": 0.0233, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.0673021374927787, | |
| "grad_norm": 1.747198462486267, | |
| "learning_rate": 3.792937399678973e-06, | |
| "loss": 0.0054, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.0701906412478337, | |
| "grad_norm": 0.5364682078361511, | |
| "learning_rate": 3.776886035313002e-06, | |
| "loss": 0.0353, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.0730791450028887, | |
| "grad_norm": 0.12140754610300064, | |
| "learning_rate": 3.7608346709470305e-06, | |
| "loss": 0.0116, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.0759676487579433, | |
| "grad_norm": 0.045340485870838165, | |
| "learning_rate": 3.7447833065810594e-06, | |
| "loss": 0.0031, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.0788561525129983, | |
| "grad_norm": 0.05959215760231018, | |
| "learning_rate": 3.7287319422150888e-06, | |
| "loss": 0.0078, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.0817446562680533, | |
| "grad_norm": 0.2539525628089905, | |
| "learning_rate": 3.7126805778491177e-06, | |
| "loss": 0.0484, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.083477758521086, | |
| "eval_accuracy": 0.8016359918200409, | |
| "eval_loss": 0.8760994672775269, | |
| "eval_runtime": 64.4473, | |
| "eval_samples_per_second": 7.588, | |
| "eval_steps_per_second": 0.962, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 4.001155401502022, | |
| "grad_norm": 0.27617499232292175, | |
| "learning_rate": 3.6966292134831466e-06, | |
| "loss": 0.0164, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.004043905257077, | |
| "grad_norm": 0.09850801527500153, | |
| "learning_rate": 3.680577849117175e-06, | |
| "loss": 0.0043, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.006932409012132, | |
| "grad_norm": 8.809159278869629, | |
| "learning_rate": 3.664526484751204e-06, | |
| "loss": 0.0105, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.009820912767187, | |
| "grad_norm": 0.16609108448028564, | |
| "learning_rate": 3.648475120385233e-06, | |
| "loss": 0.0015, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.012709416522242, | |
| "grad_norm": 0.061846837401390076, | |
| "learning_rate": 3.6324237560192623e-06, | |
| "loss": 0.021, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.015597920277297, | |
| "grad_norm": 0.07643674314022064, | |
| "learning_rate": 3.6163723916532904e-06, | |
| "loss": 0.0021, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.018486424032352, | |
| "grad_norm": 0.24589237570762634, | |
| "learning_rate": 3.6003210272873197e-06, | |
| "loss": 0.0032, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.021374927787406, | |
| "grad_norm": 0.12212871015071869, | |
| "learning_rate": 3.5842696629213486e-06, | |
| "loss": 0.0018, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.024263431542461, | |
| "grad_norm": 0.05846753343939781, | |
| "learning_rate": 3.5682182985553776e-06, | |
| "loss": 0.0021, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.027151935297516, | |
| "grad_norm": 0.05186119303107262, | |
| "learning_rate": 3.5521669341894065e-06, | |
| "loss": 0.0023, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.030040439052571, | |
| "grad_norm": 0.060669466853141785, | |
| "learning_rate": 3.536115569823435e-06, | |
| "loss": 0.0609, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.032928942807626, | |
| "grad_norm": 0.18004949390888214, | |
| "learning_rate": 3.520064205457464e-06, | |
| "loss": 0.0009, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.035817446562681, | |
| "grad_norm": 0.8595585823059082, | |
| "learning_rate": 3.5040128410914932e-06, | |
| "loss": 0.0014, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.038705950317736, | |
| "grad_norm": 0.02049427106976509, | |
| "learning_rate": 3.487961476725522e-06, | |
| "loss": 0.0359, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.04159445407279, | |
| "grad_norm": 0.1333732008934021, | |
| "learning_rate": 3.471910112359551e-06, | |
| "loss": 0.001, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.044482957827845, | |
| "grad_norm": 0.19866865873336792, | |
| "learning_rate": 3.4558587479935796e-06, | |
| "loss": 0.0015, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.0473714615829, | |
| "grad_norm": 0.13261398673057556, | |
| "learning_rate": 3.4398073836276085e-06, | |
| "loss": 0.0009, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.050259965337955, | |
| "grad_norm": 0.041597481817007065, | |
| "learning_rate": 3.4237560192616374e-06, | |
| "loss": 0.0007, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.05314846909301, | |
| "grad_norm": 0.05136357620358467, | |
| "learning_rate": 3.4077046548956668e-06, | |
| "loss": 0.0014, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.056036972848065, | |
| "grad_norm": 0.01516722422093153, | |
| "learning_rate": 3.391653290529695e-06, | |
| "loss": 0.0021, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.05892547660312, | |
| "grad_norm": 0.10922623425722122, | |
| "learning_rate": 3.375601926163724e-06, | |
| "loss": 0.0018, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.061813980358174, | |
| "grad_norm": 0.033754732459783554, | |
| "learning_rate": 3.359550561797753e-06, | |
| "loss": 0.0008, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.064702484113229, | |
| "grad_norm": 0.025863097980618477, | |
| "learning_rate": 3.343499197431782e-06, | |
| "loss": 0.0064, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.067590987868284, | |
| "grad_norm": 0.7977182865142822, | |
| "learning_rate": 3.327447833065811e-06, | |
| "loss": 0.0113, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.070479491623339, | |
| "grad_norm": 0.10724899172782898, | |
| "learning_rate": 3.3113964686998394e-06, | |
| "loss": 0.0009, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.073367995378394, | |
| "grad_norm": 0.06613096594810486, | |
| "learning_rate": 3.2953451043338684e-06, | |
| "loss": 0.0014, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.076256499133449, | |
| "grad_norm": 0.166402205824852, | |
| "learning_rate": 3.2792937399678977e-06, | |
| "loss": 0.0017, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.079145002888504, | |
| "grad_norm": 0.032608021050691605, | |
| "learning_rate": 3.2632423756019266e-06, | |
| "loss": 0.037, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.082033506643558, | |
| "grad_norm": 0.21126851439476013, | |
| "learning_rate": 3.2471910112359555e-06, | |
| "loss": 0.0032, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.083477758521086, | |
| "eval_accuracy": 0.8139059304703476, | |
| "eval_loss": 0.9753301739692688, | |
| "eval_runtime": 64.1258, | |
| "eval_samples_per_second": 7.626, | |
| "eval_steps_per_second": 0.967, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 5.001444251877527, | |
| "grad_norm": 0.030852673575282097, | |
| "learning_rate": 3.231139646869984e-06, | |
| "loss": 0.0006, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 5.004332755632582, | |
| "grad_norm": 0.12117096781730652, | |
| "learning_rate": 3.215088282504013e-06, | |
| "loss": 0.0007, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 5.007221259387637, | |
| "grad_norm": 0.19616347551345825, | |
| "learning_rate": 3.199036918138042e-06, | |
| "loss": 0.0024, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 5.010109763142692, | |
| "grad_norm": 0.05154794454574585, | |
| "learning_rate": 3.1829855537720712e-06, | |
| "loss": 0.0084, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 5.012998266897747, | |
| "grad_norm": 0.01491555105894804, | |
| "learning_rate": 3.1669341894060997e-06, | |
| "loss": 0.0008, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 5.015886770652802, | |
| "grad_norm": 0.6864179968833923, | |
| "learning_rate": 3.1508828250401286e-06, | |
| "loss": 0.0007, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.018775274407857, | |
| "grad_norm": 1.6264543533325195, | |
| "learning_rate": 3.1348314606741576e-06, | |
| "loss": 0.0187, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 5.021663778162911, | |
| "grad_norm": 0.017343902960419655, | |
| "learning_rate": 3.1187800963081865e-06, | |
| "loss": 0.0011, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 5.024552281917966, | |
| "grad_norm": 0.11245548725128174, | |
| "learning_rate": 3.1027287319422154e-06, | |
| "loss": 0.0016, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 5.027440785673021, | |
| "grad_norm": 0.038954492658376694, | |
| "learning_rate": 3.086677367576244e-06, | |
| "loss": 0.0005, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 5.030329289428076, | |
| "grad_norm": 0.023842019960284233, | |
| "learning_rate": 3.0706260032102732e-06, | |
| "loss": 0.0006, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 5.033217793183131, | |
| "grad_norm": 0.042979270219802856, | |
| "learning_rate": 3.054574638844302e-06, | |
| "loss": 0.0012, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 5.036106296938186, | |
| "grad_norm": 11.767372131347656, | |
| "learning_rate": 3.038523274478331e-06, | |
| "loss": 0.0263, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 5.038994800693241, | |
| "grad_norm": 0.03212727978825569, | |
| "learning_rate": 3.02247191011236e-06, | |
| "loss": 0.0006, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 5.041883304448295, | |
| "grad_norm": 0.11572208255529404, | |
| "learning_rate": 3.0064205457463885e-06, | |
| "loss": 0.0009, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.04477180820335, | |
| "grad_norm": 0.04862317815423012, | |
| "learning_rate": 2.9903691813804174e-06, | |
| "loss": 0.0007, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.047660311958405, | |
| "grad_norm": 0.01599649153649807, | |
| "learning_rate": 2.9743178170144463e-06, | |
| "loss": 0.0005, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 5.05054881571346, | |
| "grad_norm": 0.10585848242044449, | |
| "learning_rate": 2.9582664526484757e-06, | |
| "loss": 0.0006, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 5.053437319468515, | |
| "grad_norm": 0.015074520371854305, | |
| "learning_rate": 2.942215088282504e-06, | |
| "loss": 0.0004, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 5.05632582322357, | |
| "grad_norm": 0.06937854737043381, | |
| "learning_rate": 2.926163723916533e-06, | |
| "loss": 0.0005, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.059214326978625, | |
| "grad_norm": 0.028225358575582504, | |
| "learning_rate": 2.910112359550562e-06, | |
| "loss": 0.0005, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 5.06210283073368, | |
| "grad_norm": 0.02525603584945202, | |
| "learning_rate": 2.894060995184591e-06, | |
| "loss": 0.0003, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 5.0649913344887345, | |
| "grad_norm": 7.610655784606934, | |
| "learning_rate": 2.87800963081862e-06, | |
| "loss": 0.0134, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 5.0678798382437895, | |
| "grad_norm": 0.037229206413030624, | |
| "learning_rate": 2.8619582664526484e-06, | |
| "loss": 0.0005, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 5.0707683419988445, | |
| "grad_norm": 0.031067833304405212, | |
| "learning_rate": 2.8459069020866777e-06, | |
| "loss": 0.0005, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 5.0736568457538995, | |
| "grad_norm": 0.053743135184049606, | |
| "learning_rate": 2.8298555377207066e-06, | |
| "loss": 0.0007, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.0765453495089545, | |
| "grad_norm": 1.9661149978637695, | |
| "learning_rate": 2.8138041733547356e-06, | |
| "loss": 0.0018, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 5.0794338532640095, | |
| "grad_norm": 0.024581482633948326, | |
| "learning_rate": 2.797752808988764e-06, | |
| "loss": 0.0008, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 5.0823223570190645, | |
| "grad_norm": 0.028522053733468056, | |
| "learning_rate": 2.781701444622793e-06, | |
| "loss": 0.0004, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 5.083477758521086, | |
| "eval_accuracy": 0.8057259713701431, | |
| "eval_loss": 1.045902967453003, | |
| "eval_runtime": 61.8445, | |
| "eval_samples_per_second": 7.907, | |
| "eval_steps_per_second": 1.003, | |
| "step": 1734 | |
| }, | |
| { | |
| "epoch": 6.001733102253033, | |
| "grad_norm": 0.009279723279178143, | |
| "learning_rate": 2.765650080256822e-06, | |
| "loss": 0.0004, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 6.0046216060080875, | |
| "grad_norm": 0.013963778503239155, | |
| "learning_rate": 2.7495987158908512e-06, | |
| "loss": 0.0004, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 6.0075101097631425, | |
| "grad_norm": 0.02687692642211914, | |
| "learning_rate": 2.73354735152488e-06, | |
| "loss": 0.0004, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 6.0103986135181975, | |
| "grad_norm": 1.6647899150848389, | |
| "learning_rate": 2.7174959871589087e-06, | |
| "loss": 0.0196, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 6.0132871172732525, | |
| "grad_norm": 0.024079566821455956, | |
| "learning_rate": 2.7014446227929376e-06, | |
| "loss": 0.0005, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 6.0161756210283075, | |
| "grad_norm": 0.08010951429605484, | |
| "learning_rate": 2.6853932584269665e-06, | |
| "loss": 0.0004, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 6.0190641247833625, | |
| "grad_norm": 0.03973499685525894, | |
| "learning_rate": 2.6693418940609954e-06, | |
| "loss": 0.0016, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 6.0219526285384175, | |
| "grad_norm": 0.012200037948787212, | |
| "learning_rate": 2.6532905296950243e-06, | |
| "loss": 0.0039, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 6.024841132293472, | |
| "grad_norm": 0.020096924155950546, | |
| "learning_rate": 2.637239165329053e-06, | |
| "loss": 0.001, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 6.027729636048527, | |
| "grad_norm": 0.037991344928741455, | |
| "learning_rate": 2.621187800963082e-06, | |
| "loss": 0.0009, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 6.030618139803582, | |
| "grad_norm": 0.01323782093822956, | |
| "learning_rate": 2.605136436597111e-06, | |
| "loss": 0.0003, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 6.033506643558637, | |
| "grad_norm": 0.015599347651004791, | |
| "learning_rate": 2.58908507223114e-06, | |
| "loss": 0.0004, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 6.036395147313692, | |
| "grad_norm": 0.05168904364109039, | |
| "learning_rate": 2.5730337078651685e-06, | |
| "loss": 0.0004, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 6.039283651068747, | |
| "grad_norm": 0.019163014367222786, | |
| "learning_rate": 2.5569823434991974e-06, | |
| "loss": 0.0004, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 6.042172154823802, | |
| "grad_norm": 0.025135764852166176, | |
| "learning_rate": 2.5409309791332264e-06, | |
| "loss": 0.0003, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 6.045060658578856, | |
| "grad_norm": 0.030841296538710594, | |
| "learning_rate": 2.5248796147672557e-06, | |
| "loss": 0.0004, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 6.047949162333911, | |
| "grad_norm": 0.01263574417680502, | |
| "learning_rate": 2.5088282504012846e-06, | |
| "loss": 0.0003, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 6.050837666088966, | |
| "grad_norm": 0.015735147520899773, | |
| "learning_rate": 2.492776886035313e-06, | |
| "loss": 0.0004, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 6.053726169844021, | |
| "grad_norm": 0.00912562571465969, | |
| "learning_rate": 2.476725521669342e-06, | |
| "loss": 0.0062, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 6.056614673599076, | |
| "grad_norm": 0.017274271696805954, | |
| "learning_rate": 2.460674157303371e-06, | |
| "loss": 0.0003, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 6.059503177354131, | |
| "grad_norm": 0.4504781663417816, | |
| "learning_rate": 2.4446227929374e-06, | |
| "loss": 0.0005, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 6.062391681109186, | |
| "grad_norm": 0.033242367208004, | |
| "learning_rate": 2.428571428571429e-06, | |
| "loss": 0.0008, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 6.06528018486424, | |
| "grad_norm": 0.03502265736460686, | |
| "learning_rate": 2.4125200642054577e-06, | |
| "loss": 0.0003, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 6.068168688619295, | |
| "grad_norm": 0.033378973603248596, | |
| "learning_rate": 2.3964686998394866e-06, | |
| "loss": 0.0123, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 6.07105719237435, | |
| "grad_norm": 0.01710161752998829, | |
| "learning_rate": 2.3804173354735156e-06, | |
| "loss": 0.0003, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 6.073945696129405, | |
| "grad_norm": 0.02805742621421814, | |
| "learning_rate": 2.364365971107544e-06, | |
| "loss": 0.0003, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 6.07683419988446, | |
| "grad_norm": 2.4002020359039307, | |
| "learning_rate": 2.3483146067415734e-06, | |
| "loss": 0.0041, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.079722703639515, | |
| "grad_norm": 0.034336935728788376, | |
| "learning_rate": 2.3322632423756023e-06, | |
| "loss": 0.0003, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 6.08261120739457, | |
| "grad_norm": 0.03934768959879875, | |
| "learning_rate": 2.316211878009631e-06, | |
| "loss": 0.002, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 6.083477758521086, | |
| "eval_accuracy": 0.7914110429447853, | |
| "eval_loss": 1.153733730316162, | |
| "eval_runtime": 66.131, | |
| "eval_samples_per_second": 7.394, | |
| "eval_steps_per_second": 0.938, | |
| "step": 2023 | |
| }, | |
| { | |
| "epoch": 7.002021952628539, | |
| "grad_norm": 0.026671981438994408, | |
| "learning_rate": 2.30016051364366e-06, | |
| "loss": 0.0003, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 7.004910456383593, | |
| "grad_norm": 0.022940438240766525, | |
| "learning_rate": 2.2841091492776887e-06, | |
| "loss": 0.0003, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 7.007798960138648, | |
| "grad_norm": 0.013357292860746384, | |
| "learning_rate": 2.2680577849117176e-06, | |
| "loss": 0.0003, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 7.010687463893703, | |
| "grad_norm": 0.18174605071544647, | |
| "learning_rate": 2.2520064205457465e-06, | |
| "loss": 0.0003, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 7.013575967648758, | |
| "grad_norm": 0.03782898932695389, | |
| "learning_rate": 2.2359550561797754e-06, | |
| "loss": 0.0003, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 7.016464471403813, | |
| "grad_norm": 0.01123202033340931, | |
| "learning_rate": 2.2199036918138043e-06, | |
| "loss": 0.0003, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 7.019352975158868, | |
| "grad_norm": 0.016729041934013367, | |
| "learning_rate": 2.2038523274478333e-06, | |
| "loss": 0.0002, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 7.022241478913923, | |
| "grad_norm": 0.014862176962196827, | |
| "learning_rate": 2.187800963081862e-06, | |
| "loss": 0.0008, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 7.025129982668977, | |
| "grad_norm": 0.014939128421247005, | |
| "learning_rate": 2.171749598715891e-06, | |
| "loss": 0.0003, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 7.028018486424032, | |
| "grad_norm": 0.03202977776527405, | |
| "learning_rate": 2.15569823434992e-06, | |
| "loss": 0.0002, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 7.030906990179087, | |
| "grad_norm": 0.01170146744698286, | |
| "learning_rate": 2.1396468699839485e-06, | |
| "loss": 0.0002, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 7.033795493934142, | |
| "grad_norm": 0.01243497896939516, | |
| "learning_rate": 2.123595505617978e-06, | |
| "loss": 0.0003, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 7.036683997689197, | |
| "grad_norm": 0.0059522357769310474, | |
| "learning_rate": 2.1075441412520064e-06, | |
| "loss": 0.0076, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 7.039572501444252, | |
| "grad_norm": 0.009490188211202621, | |
| "learning_rate": 2.0914927768860353e-06, | |
| "loss": 0.0002, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 7.042461005199307, | |
| "grad_norm": 0.028632614761590958, | |
| "learning_rate": 2.0754414125200646e-06, | |
| "loss": 0.0046, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 7.045349508954362, | |
| "grad_norm": 0.014610586687922478, | |
| "learning_rate": 2.059390048154093e-06, | |
| "loss": 0.0016, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 7.048238012709416, | |
| "grad_norm": 0.009654671885073185, | |
| "learning_rate": 2.043338683788122e-06, | |
| "loss": 0.0002, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 7.051126516464471, | |
| "grad_norm": 0.0431351475417614, | |
| "learning_rate": 2.027287319422151e-06, | |
| "loss": 0.0011, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 7.054015020219526, | |
| "grad_norm": 0.037138573825359344, | |
| "learning_rate": 2.01123595505618e-06, | |
| "loss": 0.0003, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 7.056903523974581, | |
| "grad_norm": 0.041600652039051056, | |
| "learning_rate": 1.995184590690209e-06, | |
| "loss": 0.001, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 7.059792027729636, | |
| "grad_norm": 0.030539963394403458, | |
| "learning_rate": 1.9791332263242377e-06, | |
| "loss": 0.0002, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 7.062680531484691, | |
| "grad_norm": 0.06598862260580063, | |
| "learning_rate": 1.9630818619582666e-06, | |
| "loss": 0.0004, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 7.065569035239746, | |
| "grad_norm": 0.09313968569040298, | |
| "learning_rate": 1.9470304975922956e-06, | |
| "loss": 0.0003, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 7.0684575389948, | |
| "grad_norm": 0.004848283249884844, | |
| "learning_rate": 1.9309791332263245e-06, | |
| "loss": 0.0003, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 7.071346042749855, | |
| "grad_norm": 0.009899858385324478, | |
| "learning_rate": 1.9149277688603534e-06, | |
| "loss": 0.0003, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 7.07423454650491, | |
| "grad_norm": 0.010704581625759602, | |
| "learning_rate": 1.8988764044943821e-06, | |
| "loss": 0.0002, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 7.077123050259965, | |
| "grad_norm": 0.009377597831189632, | |
| "learning_rate": 1.882825040128411e-06, | |
| "loss": 0.0002, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 7.08001155401502, | |
| "grad_norm": 0.007483753375709057, | |
| "learning_rate": 1.86677367576244e-06, | |
| "loss": 0.0002, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.082900057770075, | |
| "grad_norm": 0.005416968371719122, | |
| "learning_rate": 1.8507223113964689e-06, | |
| "loss": 0.0002, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 7.083477758521086, | |
| "eval_accuracy": 0.8016359918200409, | |
| "eval_loss": 1.142956018447876, | |
| "eval_runtime": 67.63, | |
| "eval_samples_per_second": 7.231, | |
| "eval_steps_per_second": 0.917, | |
| "step": 2312 | |
| }, | |
| { | |
| "epoch": 8.002310803004043, | |
| "grad_norm": 2.4457383155822754, | |
| "learning_rate": 1.8346709470304978e-06, | |
| "loss": 0.0028, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 8.0051993067591, | |
| "grad_norm": 0.013131977990269661, | |
| "learning_rate": 1.8186195826645267e-06, | |
| "loss": 0.0004, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 8.008087810514153, | |
| "grad_norm": 0.014833220280706882, | |
| "learning_rate": 1.8025682182985554e-06, | |
| "loss": 0.0002, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 8.01097631426921, | |
| "grad_norm": 0.016327854245901108, | |
| "learning_rate": 1.7865168539325846e-06, | |
| "loss": 0.0002, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 8.013864818024263, | |
| "grad_norm": 0.016063909977674484, | |
| "learning_rate": 1.7704654895666133e-06, | |
| "loss": 0.0002, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 8.016753321779317, | |
| "grad_norm": 0.049486108124256134, | |
| "learning_rate": 1.7544141252006422e-06, | |
| "loss": 0.0002, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 8.019641825534373, | |
| "grad_norm": 0.0232455562800169, | |
| "learning_rate": 1.738362760834671e-06, | |
| "loss": 0.0002, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 8.022530329289427, | |
| "grad_norm": 0.007956156507134438, | |
| "learning_rate": 1.7223113964687e-06, | |
| "loss": 0.0002, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 8.025418833044483, | |
| "grad_norm": 0.019017890095710754, | |
| "learning_rate": 1.706260032102729e-06, | |
| "loss": 0.0002, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 8.028307336799537, | |
| "grad_norm": 0.011254646815359592, | |
| "learning_rate": 1.6902086677367576e-06, | |
| "loss": 0.0004, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 8.031195840554593, | |
| "grad_norm": 0.008488072082400322, | |
| "learning_rate": 1.6741573033707868e-06, | |
| "loss": 0.0003, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 8.034084344309647, | |
| "grad_norm": 0.008714953437447548, | |
| "learning_rate": 1.6581059390048155e-06, | |
| "loss": 0.0002, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 8.036972848064703, | |
| "grad_norm": 0.06316517293453217, | |
| "learning_rate": 1.6420545746388444e-06, | |
| "loss": 0.0002, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 8.039861351819757, | |
| "grad_norm": 0.014625471085309982, | |
| "learning_rate": 1.6260032102728735e-06, | |
| "loss": 0.0003, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 8.042749855574812, | |
| "grad_norm": 0.015664029866456985, | |
| "learning_rate": 1.6099518459069023e-06, | |
| "loss": 0.0002, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 8.045638359329867, | |
| "grad_norm": 0.007262027356773615, | |
| "learning_rate": 1.5939004815409312e-06, | |
| "loss": 0.0002, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 8.048526863084922, | |
| "grad_norm": 0.008894843980669975, | |
| "learning_rate": 1.5778491171749599e-06, | |
| "loss": 0.0002, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 8.051415366839977, | |
| "grad_norm": 0.010352100245654583, | |
| "learning_rate": 1.561797752808989e-06, | |
| "loss": 0.0002, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 8.054303870595032, | |
| "grad_norm": 0.005582538899034262, | |
| "learning_rate": 1.5457463884430177e-06, | |
| "loss": 0.0002, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 8.057192374350088, | |
| "grad_norm": 0.01137503795325756, | |
| "learning_rate": 1.5296950240770466e-06, | |
| "loss": 0.0002, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 8.060080878105142, | |
| "grad_norm": 0.005585261154919863, | |
| "learning_rate": 1.5136436597110758e-06, | |
| "loss": 0.0002, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 8.062969381860196, | |
| "grad_norm": 0.015739308670163155, | |
| "learning_rate": 1.4975922953451045e-06, | |
| "loss": 0.0002, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 8.065857885615252, | |
| "grad_norm": 0.006552826147526503, | |
| "learning_rate": 1.4815409309791334e-06, | |
| "loss": 0.0002, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 8.068746389370306, | |
| "grad_norm": 0.024894963949918747, | |
| "learning_rate": 1.4654895666131621e-06, | |
| "loss": 0.0002, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 8.071634893125362, | |
| "grad_norm": 0.047090210020542145, | |
| "learning_rate": 1.4494382022471912e-06, | |
| "loss": 0.0002, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 8.074523396880416, | |
| "grad_norm": 0.005788912996649742, | |
| "learning_rate": 1.43338683788122e-06, | |
| "loss": 0.0002, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 8.077411900635472, | |
| "grad_norm": 0.008126024156808853, | |
| "learning_rate": 1.4173354735152489e-06, | |
| "loss": 0.0002, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 8.080300404390526, | |
| "grad_norm": 0.13917505741119385, | |
| "learning_rate": 1.401284109149278e-06, | |
| "loss": 0.0002, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 8.08318890814558, | |
| "grad_norm": 0.04815088212490082, | |
| "learning_rate": 1.3852327447833067e-06, | |
| "loss": 0.0002, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 8.083477758521086, | |
| "eval_accuracy": 0.7995910020449898, | |
| "eval_loss": 1.1875964403152466, | |
| "eval_runtime": 71.2485, | |
| "eval_samples_per_second": 6.863, | |
| "eval_steps_per_second": 0.87, | |
| "step": 2601 | |
| }, | |
| { | |
| "epoch": 8.083477758521086, | |
| "step": 2601, | |
| "total_flos": 2.5905682815117558e+19, | |
| "train_loss": 0.11906857130725572, | |
| "train_runtime": 6452.2181, | |
| "train_samples_per_second": 4.292, | |
| "train_steps_per_second": 0.537 | |
| }, | |
| { | |
| "epoch": 8.083477758521086, | |
| "eval_accuracy": 0.8393574297188755, | |
| "eval_loss": 0.6800851225852966, | |
| "eval_runtime": 69.7987, | |
| "eval_samples_per_second": 7.135, | |
| "eval_steps_per_second": 0.903, | |
| "step": 2601 | |
| }, | |
| { | |
| "epoch": 8.083477758521086, | |
| "eval_accuracy": 0.8393574297188755, | |
| "eval_loss": 0.6800851821899414, | |
| "eval_runtime": 68.1415, | |
| "eval_samples_per_second": 7.308, | |
| "eval_steps_per_second": 0.925, | |
| "step": 2601 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3462, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 4, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 4 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.5905682815117558e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |