| { | |
| "best_metric": 0.8209169054441261, | |
| "best_model_checkpoint": "videomae-base-finetuned-scratch_1/checkpoint-5280", | |
| "epoch": 35.02580971659919, | |
| "eval_steps": 500, | |
| "global_step": 11856, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 5.349636554718018, | |
| "learning_rate": 4.2158516020236085e-07, | |
| "loss": 0.7141, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 5.897022724151611, | |
| "learning_rate": 8.431703204047217e-07, | |
| "loss": 0.7044, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 4.595437526702881, | |
| "learning_rate": 1.2647554806070827e-06, | |
| "loss": 0.7118, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 4.464962959289551, | |
| "learning_rate": 1.6863406408094434e-06, | |
| "loss": 0.7121, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 4.585201263427734, | |
| "learning_rate": 2.1079258010118047e-06, | |
| "loss": 0.7018, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.25459098815918, | |
| "learning_rate": 2.5295109612141654e-06, | |
| "loss": 0.6952, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.792253017425537, | |
| "learning_rate": 2.951096121416526e-06, | |
| "loss": 0.6813, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.621083974838257, | |
| "learning_rate": 3.372681281618887e-06, | |
| "loss": 0.6814, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.006009578704834, | |
| "learning_rate": 3.794266441821248e-06, | |
| "loss": 0.6896, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.3775813579559326, | |
| "learning_rate": 4.2158516020236095e-06, | |
| "loss": 0.6803, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 5.32543420791626, | |
| "learning_rate": 4.63743676222597e-06, | |
| "loss": 0.6793, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.859120845794678, | |
| "learning_rate": 5.059021922428331e-06, | |
| "loss": 0.6715, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.3114655017852783, | |
| "learning_rate": 5.480607082630692e-06, | |
| "loss": 0.705, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.681263446807861, | |
| "learning_rate": 5.902192242833052e-06, | |
| "loss": 0.7032, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.5089430809020996, | |
| "learning_rate": 6.323777403035413e-06, | |
| "loss": 0.6903, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.625184059143066, | |
| "learning_rate": 6.745362563237774e-06, | |
| "loss": 0.6885, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.5083889961242676, | |
| "learning_rate": 7.166947723440136e-06, | |
| "loss": 0.6671, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.028371334075928, | |
| "learning_rate": 7.588532883642496e-06, | |
| "loss": 0.6822, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.046874046325684, | |
| "learning_rate": 8.010118043844857e-06, | |
| "loss": 0.676, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 7.515069961547852, | |
| "learning_rate": 8.431703204047219e-06, | |
| "loss": 0.6649, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.611435890197754, | |
| "learning_rate": 8.85328836424958e-06, | |
| "loss": 0.6726, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.385730743408203, | |
| "learning_rate": 9.27487352445194e-06, | |
| "loss": 0.6771, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.537276268005371, | |
| "learning_rate": 9.696458684654301e-06, | |
| "loss": 0.6543, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.257579803466797, | |
| "learning_rate": 1.0118043844856662e-05, | |
| "loss": 0.6602, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.346296787261963, | |
| "learning_rate": 1.0539629005059022e-05, | |
| "loss": 0.6293, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 8.624234199523926, | |
| "learning_rate": 1.0961214165261384e-05, | |
| "loss": 0.6735, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.386282920837402, | |
| "learning_rate": 1.1382799325463744e-05, | |
| "loss": 0.6612, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 9.212226867675781, | |
| "learning_rate": 1.1804384485666105e-05, | |
| "loss": 0.673, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.993302822113037, | |
| "learning_rate": 1.2225969645868467e-05, | |
| "loss": 0.6551, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.731637954711914, | |
| "learning_rate": 1.2647554806070827e-05, | |
| "loss": 0.6432, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.617886543273926, | |
| "learning_rate": 1.3069139966273189e-05, | |
| "loss": 0.6777, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 7.462168216705322, | |
| "learning_rate": 1.3490725126475547e-05, | |
| "loss": 0.6153, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 9.60923957824707, | |
| "learning_rate": 1.391231028667791e-05, | |
| "loss": 0.6811, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "eval_accuracy": 0.6547277936962751, | |
| "eval_loss": 0.6314294338226318, | |
| "eval_runtime": 33.9302, | |
| "eval_samples_per_second": 20.572, | |
| "eval_steps_per_second": 1.739, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 9.263134002685547, | |
| "learning_rate": 1.4333895446880271e-05, | |
| "loss": 0.6244, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 8.611780166625977, | |
| "learning_rate": 1.4755480607082632e-05, | |
| "loss": 0.6358, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 8.437411308288574, | |
| "learning_rate": 1.5177065767284992e-05, | |
| "loss": 0.6273, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 6.759949207305908, | |
| "learning_rate": 1.5598650927487355e-05, | |
| "loss": 0.6529, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 15.71631908416748, | |
| "learning_rate": 1.6020236087689714e-05, | |
| "loss": 0.6066, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 17.849533081054688, | |
| "learning_rate": 1.6441821247892076e-05, | |
| "loss": 0.5587, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 4.601895809173584, | |
| "learning_rate": 1.6863406408094438e-05, | |
| "loss": 0.6334, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 6.037390232086182, | |
| "learning_rate": 1.7284991568296797e-05, | |
| "loss": 0.7324, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 4.270590782165527, | |
| "learning_rate": 1.770657672849916e-05, | |
| "loss": 0.6552, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 10.071660995483398, | |
| "learning_rate": 1.812816188870152e-05, | |
| "loss": 0.6441, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 7.655844211578369, | |
| "learning_rate": 1.854974704890388e-05, | |
| "loss": 0.6271, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 7.267734050750732, | |
| "learning_rate": 1.897133220910624e-05, | |
| "loss": 0.6048, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 6.196805953979492, | |
| "learning_rate": 1.9392917369308603e-05, | |
| "loss": 0.6371, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 5.247162342071533, | |
| "learning_rate": 1.981450252951096e-05, | |
| "loss": 0.5485, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 7.14028787612915, | |
| "learning_rate": 2.0236087689713324e-05, | |
| "loss": 0.6503, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 7.4398722648620605, | |
| "learning_rate": 2.0657672849915685e-05, | |
| "loss": 0.5997, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 4.917046070098877, | |
| "learning_rate": 2.1079258010118044e-05, | |
| "loss": 0.6271, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 7.6909565925598145, | |
| "learning_rate": 2.1500843170320406e-05, | |
| "loss": 0.5816, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 5.249167442321777, | |
| "learning_rate": 2.1922428330522768e-05, | |
| "loss": 0.6534, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 5.298259735107422, | |
| "learning_rate": 2.2344013490725127e-05, | |
| "loss": 0.6713, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 8.135736465454102, | |
| "learning_rate": 2.276559865092749e-05, | |
| "loss": 0.5999, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 9.37923812866211, | |
| "learning_rate": 2.318718381112985e-05, | |
| "loss": 0.7861, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 4.831663131713867, | |
| "learning_rate": 2.360876897133221e-05, | |
| "loss": 0.605, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 5.473945617675781, | |
| "learning_rate": 2.403035413153457e-05, | |
| "loss": 0.6066, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 5.188632488250732, | |
| "learning_rate": 2.4451939291736933e-05, | |
| "loss": 0.6189, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 4.069911956787109, | |
| "learning_rate": 2.487352445193929e-05, | |
| "loss": 0.6705, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 8.71951675415039, | |
| "learning_rate": 2.5295109612141654e-05, | |
| "loss": 0.6168, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 6.839282989501953, | |
| "learning_rate": 2.5716694772344012e-05, | |
| "loss": 0.6465, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 5.155991077423096, | |
| "learning_rate": 2.6138279932546377e-05, | |
| "loss": 0.6007, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 4.38847541809082, | |
| "learning_rate": 2.6559865092748736e-05, | |
| "loss": 0.6253, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 6.741366386413574, | |
| "learning_rate": 2.6981450252951095e-05, | |
| "loss": 0.614, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 6.325467586517334, | |
| "learning_rate": 2.740303541315346e-05, | |
| "loss": 0.6444, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 7.080298900604248, | |
| "learning_rate": 2.782462057335582e-05, | |
| "loss": 0.6706, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "eval_accuracy": 0.6318051575931232, | |
| "eval_loss": 0.635435938835144, | |
| "eval_runtime": 36.8826, | |
| "eval_samples_per_second": 18.925, | |
| "eval_steps_per_second": 1.6, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 8.330774307250977, | |
| "learning_rate": 2.8246205733558177e-05, | |
| "loss": 0.5594, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.559295654296875, | |
| "learning_rate": 2.8667790893760543e-05, | |
| "loss": 0.6125, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 7.404927730560303, | |
| "learning_rate": 2.90893760539629e-05, | |
| "loss": 0.5856, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 13.15217113494873, | |
| "learning_rate": 2.9510961214165263e-05, | |
| "loss": 0.5817, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 8.084522247314453, | |
| "learning_rate": 2.9932546374367625e-05, | |
| "loss": 0.646, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 10.898458480834961, | |
| "learning_rate": 3.0354131534569984e-05, | |
| "loss": 0.5814, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 4.796360015869141, | |
| "learning_rate": 3.0775716694772346e-05, | |
| "loss": 0.563, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 6.336984157562256, | |
| "learning_rate": 3.119730185497471e-05, | |
| "loss": 0.5837, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 6.342469692230225, | |
| "learning_rate": 3.161888701517707e-05, | |
| "loss": 0.5755, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 3.3841326236724854, | |
| "learning_rate": 3.204047217537943e-05, | |
| "loss": 0.6185, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 3.3901569843292236, | |
| "learning_rate": 3.2462057335581793e-05, | |
| "loss": 0.6117, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 5.979400634765625, | |
| "learning_rate": 3.288364249578415e-05, | |
| "loss": 0.5974, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 8.773338317871094, | |
| "learning_rate": 3.330522765598651e-05, | |
| "loss": 0.6024, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 7.4114298820495605, | |
| "learning_rate": 3.3726812816188876e-05, | |
| "loss": 0.5838, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 8.659492492675781, | |
| "learning_rate": 3.4148397976391235e-05, | |
| "loss": 0.6818, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 4.71567440032959, | |
| "learning_rate": 3.456998313659359e-05, | |
| "loss": 0.6604, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 4.012688159942627, | |
| "learning_rate": 3.499156829679596e-05, | |
| "loss": 0.6041, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 5.472045421600342, | |
| "learning_rate": 3.541315345699832e-05, | |
| "loss": 0.573, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 3.6241841316223145, | |
| "learning_rate": 3.5834738617200676e-05, | |
| "loss": 0.6385, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 3.782322645187378, | |
| "learning_rate": 3.625632377740304e-05, | |
| "loss": 0.6425, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 5.188979148864746, | |
| "learning_rate": 3.66779089376054e-05, | |
| "loss": 0.5915, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 7.798283576965332, | |
| "learning_rate": 3.709949409780776e-05, | |
| "loss": 0.6863, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 4.644458293914795, | |
| "learning_rate": 3.7521079258010123e-05, | |
| "loss": 0.6003, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 7.098191738128662, | |
| "learning_rate": 3.794266441821248e-05, | |
| "loss": 0.5639, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 5.175333499908447, | |
| "learning_rate": 3.836424957841484e-05, | |
| "loss": 0.5598, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 7.205776691436768, | |
| "learning_rate": 3.8785834738617206e-05, | |
| "loss": 0.6401, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 4.420816898345947, | |
| "learning_rate": 3.9207419898819565e-05, | |
| "loss": 0.6388, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 3.2338483333587646, | |
| "learning_rate": 3.962900505902192e-05, | |
| "loss": 0.5893, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 3.816267251968384, | |
| "learning_rate": 4.005059021922429e-05, | |
| "loss": 0.6419, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 3.6634039878845215, | |
| "learning_rate": 4.047217537942665e-05, | |
| "loss": 0.659, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 2.776705503463745, | |
| "learning_rate": 4.0893760539629006e-05, | |
| "loss": 0.6155, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 3.137341260910034, | |
| "learning_rate": 4.131534569983137e-05, | |
| "loss": 0.6158, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 6.1107707023620605, | |
| "learning_rate": 4.173693086003373e-05, | |
| "loss": 0.6581, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "eval_accuracy": 0.6375358166189111, | |
| "eval_loss": 0.6327700018882751, | |
| "eval_runtime": 33.7114, | |
| "eval_samples_per_second": 20.705, | |
| "eval_steps_per_second": 1.75, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.0415008068084717, | |
| "learning_rate": 4.215851602023609e-05, | |
| "loss": 0.6034, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 6.006179332733154, | |
| "learning_rate": 4.2580101180438453e-05, | |
| "loss": 0.4955, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 6.862912654876709, | |
| "learning_rate": 4.300168634064081e-05, | |
| "loss": 0.4918, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 11.918879508972168, | |
| "learning_rate": 4.342327150084317e-05, | |
| "loss": 0.5643, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 8.16517448425293, | |
| "learning_rate": 4.3844856661045536e-05, | |
| "loss": 0.638, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 6.384884357452393, | |
| "learning_rate": 4.4266441821247895e-05, | |
| "loss": 0.5995, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 5.2973127365112305, | |
| "learning_rate": 4.468802698145025e-05, | |
| "loss": 0.639, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 3.8324005603790283, | |
| "learning_rate": 4.510961214165262e-05, | |
| "loss": 0.572, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 4.97128963470459, | |
| "learning_rate": 4.553119730185498e-05, | |
| "loss": 0.597, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 5.064922332763672, | |
| "learning_rate": 4.5952782462057336e-05, | |
| "loss": 0.5617, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 3.734870672225952, | |
| "learning_rate": 4.63743676222597e-05, | |
| "loss": 0.5439, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 12.495979309082031, | |
| "learning_rate": 4.679595278246206e-05, | |
| "loss": 0.5659, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 3.5089731216430664, | |
| "learning_rate": 4.721753794266442e-05, | |
| "loss": 0.517, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 5.993117332458496, | |
| "learning_rate": 4.7639123102866784e-05, | |
| "loss": 0.6202, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 3.431788444519043, | |
| "learning_rate": 4.806070826306914e-05, | |
| "loss": 0.5767, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 5.00049352645874, | |
| "learning_rate": 4.84822934232715e-05, | |
| "loss": 0.5623, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 5.571552276611328, | |
| "learning_rate": 4.8903878583473866e-05, | |
| "loss": 0.5445, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 3.7177913188934326, | |
| "learning_rate": 4.9325463743676225e-05, | |
| "loss": 0.5748, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 4.8672194480896, | |
| "learning_rate": 4.974704890387858e-05, | |
| "loss": 0.6161, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 4.830544948577881, | |
| "learning_rate": 4.9981255857544517e-05, | |
| "loss": 0.587, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 5.276002407073975, | |
| "learning_rate": 4.993439550140581e-05, | |
| "loss": 0.5564, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 6.24776029586792, | |
| "learning_rate": 4.988753514526711e-05, | |
| "loss": 0.5164, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 4.341228008270264, | |
| "learning_rate": 4.98406747891284e-05, | |
| "loss": 0.501, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 3.2242419719696045, | |
| "learning_rate": 4.979381443298969e-05, | |
| "loss": 0.5489, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 4.790765285491943, | |
| "learning_rate": 4.9746954076850985e-05, | |
| "loss": 0.5591, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 3.7129054069519043, | |
| "learning_rate": 4.9700093720712284e-05, | |
| "loss": 0.612, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 7.660614013671875, | |
| "learning_rate": 4.9653233364573576e-05, | |
| "loss": 0.5194, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 5.343811511993408, | |
| "learning_rate": 4.960637300843486e-05, | |
| "loss": 0.6074, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 4.8525590896606445, | |
| "learning_rate": 4.955951265229616e-05, | |
| "loss": 0.548, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 4.513533592224121, | |
| "learning_rate": 4.951265229615745e-05, | |
| "loss": 0.5818, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 3.897047996520996, | |
| "learning_rate": 4.9465791940018746e-05, | |
| "loss": 0.5784, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 3.294161796569824, | |
| "learning_rate": 4.941893158388004e-05, | |
| "loss": 0.5605, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 6.77754545211792, | |
| "learning_rate": 4.937207122774134e-05, | |
| "loss": 0.5614, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "eval_accuracy": 0.7134670487106017, | |
| "eval_loss": 0.5506558418273926, | |
| "eval_runtime": 34.5926, | |
| "eval_samples_per_second": 20.178, | |
| "eval_steps_per_second": 1.706, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.854128360748291, | |
| "learning_rate": 4.932521087160263e-05, | |
| "loss": 0.5509, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 6.327508449554443, | |
| "learning_rate": 4.927835051546392e-05, | |
| "loss": 0.5109, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 5.180001258850098, | |
| "learning_rate": 4.9231490159325214e-05, | |
| "loss": 0.5522, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 6.649283409118652, | |
| "learning_rate": 4.9184629803186506e-05, | |
| "loss": 0.4932, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.401059150695801, | |
| "learning_rate": 4.91377694470478e-05, | |
| "loss": 0.5446, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 4.1620635986328125, | |
| "learning_rate": 4.909090909090909e-05, | |
| "loss": 0.4964, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 5.955656051635742, | |
| "learning_rate": 4.904404873477039e-05, | |
| "loss": 0.4685, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 2.9708614349365234, | |
| "learning_rate": 4.899718837863168e-05, | |
| "loss": 0.4874, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 4.681861877441406, | |
| "learning_rate": 4.8950328022492975e-05, | |
| "loss": 0.5001, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 10.97381591796875, | |
| "learning_rate": 4.890346766635427e-05, | |
| "loss": 0.4289, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 4.6500349044799805, | |
| "learning_rate": 4.885660731021556e-05, | |
| "loss": 0.4725, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 7.461334705352783, | |
| "learning_rate": 4.880974695407685e-05, | |
| "loss": 0.5846, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 4.3084940910339355, | |
| "learning_rate": 4.8762886597938144e-05, | |
| "loss": 0.5864, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 8.88397216796875, | |
| "learning_rate": 4.8716026241799436e-05, | |
| "loss": 0.5392, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 6.398833274841309, | |
| "learning_rate": 4.8669165885660735e-05, | |
| "loss": 0.5501, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 6.638328552246094, | |
| "learning_rate": 4.862230552952203e-05, | |
| "loss": 0.5494, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 5.006091594696045, | |
| "learning_rate": 4.857544517338332e-05, | |
| "loss": 0.6367, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 5.209989547729492, | |
| "learning_rate": 4.852858481724461e-05, | |
| "loss": 0.5773, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 5.720169544219971, | |
| "learning_rate": 4.8481724461105905e-05, | |
| "loss": 0.5022, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 6.709425449371338, | |
| "learning_rate": 4.84348641049672e-05, | |
| "loss": 0.4968, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 6.903029918670654, | |
| "learning_rate": 4.838800374882849e-05, | |
| "loss": 0.5384, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 3.307621479034424, | |
| "learning_rate": 4.834114339268979e-05, | |
| "loss": 0.5675, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 6.119607925415039, | |
| "learning_rate": 4.829428303655108e-05, | |
| "loss": 0.535, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 2.2412047386169434, | |
| "learning_rate": 4.824742268041237e-05, | |
| "loss": 0.4636, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 3.960838794708252, | |
| "learning_rate": 4.8200562324273665e-05, | |
| "loss": 0.4383, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 3.922579765319824, | |
| "learning_rate": 4.8153701968134964e-05, | |
| "loss": 0.4583, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 6.935645580291748, | |
| "learning_rate": 4.810684161199626e-05, | |
| "loss": 0.4757, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 6.558530807495117, | |
| "learning_rate": 4.805998125585754e-05, | |
| "loss": 0.5097, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 4.9284186363220215, | |
| "learning_rate": 4.801312089971884e-05, | |
| "loss": 0.5218, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 8.183663368225098, | |
| "learning_rate": 4.7966260543580134e-05, | |
| "loss": 0.4462, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 6.276401996612549, | |
| "learning_rate": 4.7919400187441426e-05, | |
| "loss": 0.7016, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 4.575777530670166, | |
| "learning_rate": 4.787253983130272e-05, | |
| "loss": 0.4815, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 2.3570098876953125, | |
| "learning_rate": 4.782567947516402e-05, | |
| "loss": 0.438, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "eval_accuracy": 0.7263610315186246, | |
| "eval_loss": 0.5439700484275818, | |
| "eval_runtime": 33.443, | |
| "eval_samples_per_second": 20.871, | |
| "eval_steps_per_second": 1.764, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 7.316202163696289, | |
| "learning_rate": 4.777881911902531e-05, | |
| "loss": 0.3581, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 6.561455249786377, | |
| "learning_rate": 4.77319587628866e-05, | |
| "loss": 0.6762, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.9346327781677246, | |
| "learning_rate": 4.7685098406747894e-05, | |
| "loss": 0.4135, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.1179192066192627, | |
| "learning_rate": 4.7638238050609187e-05, | |
| "loss": 0.4336, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 6.217395305633545, | |
| "learning_rate": 4.759137769447048e-05, | |
| "loss": 0.4875, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 5.203670978546143, | |
| "learning_rate": 4.754451733833177e-05, | |
| "loss": 0.4241, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 6.5499420166015625, | |
| "learning_rate": 4.749765698219307e-05, | |
| "loss": 0.4823, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 4.842621803283691, | |
| "learning_rate": 4.745079662605436e-05, | |
| "loss": 0.3871, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 5.887164115905762, | |
| "learning_rate": 4.7403936269915655e-05, | |
| "loss": 0.4075, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 7.392530918121338, | |
| "learning_rate": 4.735707591377695e-05, | |
| "loss": 0.4504, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 6.812521934509277, | |
| "learning_rate": 4.731021555763824e-05, | |
| "loss": 0.5216, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 5.895148754119873, | |
| "learning_rate": 4.726335520149953e-05, | |
| "loss": 0.455, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 7.757108688354492, | |
| "learning_rate": 4.7216494845360824e-05, | |
| "loss": 0.488, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 9.14256477355957, | |
| "learning_rate": 4.7169634489222116e-05, | |
| "loss": 0.5042, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 5.17966890335083, | |
| "learning_rate": 4.7122774133083416e-05, | |
| "loss": 0.4534, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 6.081084728240967, | |
| "learning_rate": 4.707591377694471e-05, | |
| "loss": 0.4221, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 3.0460875034332275, | |
| "learning_rate": 4.7029053420806e-05, | |
| "loss": 0.4783, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 5.692960262298584, | |
| "learning_rate": 4.698219306466729e-05, | |
| "loss": 0.5559, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 5.824710845947266, | |
| "learning_rate": 4.6935332708528585e-05, | |
| "loss": 0.486, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 8.523110389709473, | |
| "learning_rate": 4.688847235238988e-05, | |
| "loss": 0.4036, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 13.942612648010254, | |
| "learning_rate": 4.684161199625117e-05, | |
| "loss": 0.4903, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 6.322441577911377, | |
| "learning_rate": 4.679475164011247e-05, | |
| "loss": 0.4573, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 3.1088461875915527, | |
| "learning_rate": 4.674789128397376e-05, | |
| "loss": 0.3915, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 7.44355583190918, | |
| "learning_rate": 4.670103092783505e-05, | |
| "loss": 0.4966, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 5.484576225280762, | |
| "learning_rate": 4.6654170571696345e-05, | |
| "loss": 0.4336, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 6.856348037719727, | |
| "learning_rate": 4.6607310215557645e-05, | |
| "loss": 0.5151, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 6.7076849937438965, | |
| "learning_rate": 4.656044985941894e-05, | |
| "loss": 0.4332, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 6.39366340637207, | |
| "learning_rate": 4.651358950328022e-05, | |
| "loss": 0.4858, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 7.002589225769043, | |
| "learning_rate": 4.646672914714152e-05, | |
| "loss": 0.4684, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 10.67752456665039, | |
| "learning_rate": 4.6419868791002814e-05, | |
| "loss": 0.3976, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 12.617927551269531, | |
| "learning_rate": 4.6373008434864106e-05, | |
| "loss": 0.4468, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 6.047290802001953, | |
| "learning_rate": 4.63261480787254e-05, | |
| "loss": 0.5431, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 6.107527256011963, | |
| "learning_rate": 4.62792877225867e-05, | |
| "loss": 0.4569, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "eval_accuracy": 0.7277936962750716, | |
| "eval_loss": 0.5531865954399109, | |
| "eval_runtime": 34.546, | |
| "eval_samples_per_second": 20.205, | |
| "eval_steps_per_second": 1.708, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 5.367207050323486, | |
| "learning_rate": 4.623242736644799e-05, | |
| "loss": 0.37, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 5.771437644958496, | |
| "learning_rate": 4.618556701030928e-05, | |
| "loss": 0.3602, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 9.664875984191895, | |
| "learning_rate": 4.6138706654170575e-05, | |
| "loss": 0.3343, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 6.788029193878174, | |
| "learning_rate": 4.609184629803187e-05, | |
| "loss": 0.4078, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 7.247564315795898, | |
| "learning_rate": 4.604498594189316e-05, | |
| "loss": 0.4155, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 4.327235698699951, | |
| "learning_rate": 4.599812558575445e-05, | |
| "loss": 0.3966, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 7.491323947906494, | |
| "learning_rate": 4.595126522961575e-05, | |
| "loss": 0.349, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 9.212530136108398, | |
| "learning_rate": 4.590440487347704e-05, | |
| "loss": 0.4265, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 13.466367721557617, | |
| "learning_rate": 4.5857544517338335e-05, | |
| "loss": 0.4375, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 7.694874286651611, | |
| "learning_rate": 4.581068416119963e-05, | |
| "loss": 0.3759, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 8.024492263793945, | |
| "learning_rate": 4.576382380506092e-05, | |
| "loss": 0.3936, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 5.053406238555908, | |
| "learning_rate": 4.571696344892221e-05, | |
| "loss": 0.3063, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 7.138169288635254, | |
| "learning_rate": 4.5670103092783504e-05, | |
| "loss": 0.4692, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 3.2747809886932373, | |
| "learning_rate": 4.56232427366448e-05, | |
| "loss": 0.3421, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 5.716726303100586, | |
| "learning_rate": 4.5576382380506096e-05, | |
| "loss": 0.4158, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 5.238852500915527, | |
| "learning_rate": 4.552952202436739e-05, | |
| "loss": 0.3768, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 9.766545295715332, | |
| "learning_rate": 4.548266166822868e-05, | |
| "loss": 0.3791, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 3.8693816661834717, | |
| "learning_rate": 4.543580131208997e-05, | |
| "loss": 0.334, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 9.960877418518066, | |
| "learning_rate": 4.538894095595127e-05, | |
| "loss": 0.4487, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 4.830340385437012, | |
| "learning_rate": 4.534208059981256e-05, | |
| "loss": 0.4052, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 4.533899307250977, | |
| "learning_rate": 4.529522024367385e-05, | |
| "loss": 0.3292, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 4.12489652633667, | |
| "learning_rate": 4.524835988753515e-05, | |
| "loss": 0.342, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 7.824032306671143, | |
| "learning_rate": 4.520149953139644e-05, | |
| "loss": 0.4713, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 6.992900371551514, | |
| "learning_rate": 4.5154639175257733e-05, | |
| "loss": 0.4517, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 2.9177560806274414, | |
| "learning_rate": 4.5107778819119026e-05, | |
| "loss": 0.346, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 5.294494152069092, | |
| "learning_rate": 4.5060918462980325e-05, | |
| "loss": 0.4555, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 3.4280591011047363, | |
| "learning_rate": 4.501405810684162e-05, | |
| "loss": 0.3906, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 5.682974338531494, | |
| "learning_rate": 4.49671977507029e-05, | |
| "loss": 0.3501, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 14.726093292236328, | |
| "learning_rate": 4.49203373945642e-05, | |
| "loss": 0.6106, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "grad_norm": 6.266767501831055, | |
| "learning_rate": 4.4873477038425494e-05, | |
| "loss": 0.4206, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "grad_norm": 5.161693572998047, | |
| "learning_rate": 4.4826616682286786e-05, | |
| "loss": 0.5073, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "grad_norm": 3.678534746170044, | |
| "learning_rate": 4.477975632614808e-05, | |
| "loss": 0.4903, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "grad_norm": 3.0561232566833496, | |
| "learning_rate": 4.473289597000938e-05, | |
| "loss": 0.3614, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "eval_accuracy": 0.7363896848137536, | |
| "eval_loss": 0.5283002853393555, | |
| "eval_runtime": 34.2167, | |
| "eval_samples_per_second": 20.399, | |
| "eval_steps_per_second": 1.724, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 4.80795431137085, | |
| "learning_rate": 4.468603561387067e-05, | |
| "loss": 0.342, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 5.0758748054504395, | |
| "learning_rate": 4.463917525773196e-05, | |
| "loss": 0.3799, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 7.944527626037598, | |
| "learning_rate": 4.4592314901593255e-05, | |
| "loss": 0.3295, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 8.439969062805176, | |
| "learning_rate": 4.454545454545455e-05, | |
| "loss": 0.4199, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 3.974719762802124, | |
| "learning_rate": 4.449859418931584e-05, | |
| "loss": 0.3785, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 6.712285995483398, | |
| "learning_rate": 4.445173383317713e-05, | |
| "loss": 0.3202, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 0.3726706802845001, | |
| "learning_rate": 4.440487347703843e-05, | |
| "loss": 0.1409, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 9.558029174804688, | |
| "learning_rate": 4.435801312089972e-05, | |
| "loss": 0.4147, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 3.2590224742889404, | |
| "learning_rate": 4.4311152764761015e-05, | |
| "loss": 0.2469, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 8.232924461364746, | |
| "learning_rate": 4.426429240862231e-05, | |
| "loss": 0.229, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 11.747199058532715, | |
| "learning_rate": 4.42174320524836e-05, | |
| "loss": 0.3219, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 12.6619234085083, | |
| "learning_rate": 4.417057169634489e-05, | |
| "loss": 0.4654, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 10.572305679321289, | |
| "learning_rate": 4.4123711340206185e-05, | |
| "loss": 0.4145, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 7.648813247680664, | |
| "learning_rate": 4.407685098406748e-05, | |
| "loss": 0.3939, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 7.876673698425293, | |
| "learning_rate": 4.4029990627928776e-05, | |
| "loss": 0.4002, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 2.9009244441986084, | |
| "learning_rate": 4.398313027179007e-05, | |
| "loss": 0.3665, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 4.1332597732543945, | |
| "learning_rate": 4.393626991565136e-05, | |
| "loss": 0.315, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 8.535113334655762, | |
| "learning_rate": 4.388940955951265e-05, | |
| "loss": 0.3637, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 6.837998390197754, | |
| "learning_rate": 4.384254920337395e-05, | |
| "loss": 0.2805, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 9.94541072845459, | |
| "learning_rate": 4.379568884723524e-05, | |
| "loss": 0.3765, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 8.577096939086914, | |
| "learning_rate": 4.374882849109653e-05, | |
| "loss": 0.3092, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 4.8793253898620605, | |
| "learning_rate": 4.370196813495783e-05, | |
| "loss": 0.2715, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 6.965157508850098, | |
| "learning_rate": 4.365510777881912e-05, | |
| "loss": 0.4944, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 2.5626721382141113, | |
| "learning_rate": 4.3608247422680414e-05, | |
| "loss": 0.311, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 5.375543594360352, | |
| "learning_rate": 4.3561387066541706e-05, | |
| "loss": 0.4302, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 8.762489318847656, | |
| "learning_rate": 4.3514526710403005e-05, | |
| "loss": 0.2687, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 6.056619167327881, | |
| "learning_rate": 4.34676663542643e-05, | |
| "loss": 0.3575, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 5.398069858551025, | |
| "learning_rate": 4.342080599812558e-05, | |
| "loss": 0.2445, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 6.936893939971924, | |
| "learning_rate": 4.337394564198688e-05, | |
| "loss": 0.3377, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "grad_norm": 6.378572463989258, | |
| "learning_rate": 4.3327085285848174e-05, | |
| "loss": 0.3935, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "grad_norm": 15.028035163879395, | |
| "learning_rate": 4.328022492970947e-05, | |
| "loss": 0.238, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "grad_norm": 4.001184463500977, | |
| "learning_rate": 4.323336457357076e-05, | |
| "loss": 0.3672, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "grad_norm": 0.5959993600845337, | |
| "learning_rate": 4.318650421743206e-05, | |
| "loss": 0.3514, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "eval_accuracy": 0.7406876790830945, | |
| "eval_loss": 0.6699703335762024, | |
| "eval_runtime": 33.8438, | |
| "eval_samples_per_second": 20.624, | |
| "eval_steps_per_second": 1.743, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 6.682882308959961, | |
| "learning_rate": 4.313964386129335e-05, | |
| "loss": 0.3878, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 8.971117973327637, | |
| "learning_rate": 4.309278350515464e-05, | |
| "loss": 0.3186, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 7.282140731811523, | |
| "learning_rate": 4.3045923149015935e-05, | |
| "loss": 0.2842, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 10.455528259277344, | |
| "learning_rate": 4.299906279287723e-05, | |
| "loss": 0.3175, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 8.218666076660156, | |
| "learning_rate": 4.295220243673852e-05, | |
| "loss": 0.3408, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 8.701309204101562, | |
| "learning_rate": 4.290534208059981e-05, | |
| "loss": 0.2705, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 7.51228666305542, | |
| "learning_rate": 4.285848172446111e-05, | |
| "loss": 0.3086, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 1.298295497894287, | |
| "learning_rate": 4.2811621368322403e-05, | |
| "loss": 0.2298, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 9.289319038391113, | |
| "learning_rate": 4.2764761012183696e-05, | |
| "loss": 0.2307, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 10.216160774230957, | |
| "learning_rate": 4.271790065604499e-05, | |
| "loss": 0.2715, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 0.6326087117195129, | |
| "learning_rate": 4.267104029990628e-05, | |
| "loss": 0.224, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 0.4399392306804657, | |
| "learning_rate": 4.262417994376757e-05, | |
| "loss": 0.2329, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 18.67351722717285, | |
| "learning_rate": 4.2577319587628865e-05, | |
| "loss": 0.3202, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 11.18542194366455, | |
| "learning_rate": 4.253045923149016e-05, | |
| "loss": 0.3705, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 1.94206702709198, | |
| "learning_rate": 4.2483598875351456e-05, | |
| "loss": 0.4143, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 6.226709365844727, | |
| "learning_rate": 4.243673851921275e-05, | |
| "loss": 0.31, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 14.290252685546875, | |
| "learning_rate": 4.238987816307404e-05, | |
| "loss": 0.2697, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 14.262323379516602, | |
| "learning_rate": 4.234301780693533e-05, | |
| "loss": 0.3874, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 9.656635284423828, | |
| "learning_rate": 4.229615745079663e-05, | |
| "loss": 0.2919, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 3.8038530349731445, | |
| "learning_rate": 4.224929709465792e-05, | |
| "loss": 0.3939, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 3.729642152786255, | |
| "learning_rate": 4.220243673851921e-05, | |
| "loss": 0.2423, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 4.79857873916626, | |
| "learning_rate": 4.215557638238051e-05, | |
| "loss": 0.3128, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 6.748260498046875, | |
| "learning_rate": 4.21087160262418e-05, | |
| "loss": 0.3235, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 10.759824752807617, | |
| "learning_rate": 4.2061855670103094e-05, | |
| "loss": 0.285, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 6.050924777984619, | |
| "learning_rate": 4.2014995313964386e-05, | |
| "loss": 0.3038, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 3.649509906768799, | |
| "learning_rate": 4.1968134957825685e-05, | |
| "loss": 0.2936, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 6.332077980041504, | |
| "learning_rate": 4.192127460168698e-05, | |
| "loss": 0.3153, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 3.2886719703674316, | |
| "learning_rate": 4.187441424554826e-05, | |
| "loss": 0.2492, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 11.733613014221191, | |
| "learning_rate": 4.182755388940956e-05, | |
| "loss": 0.2669, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 8.03, | |
| "grad_norm": 3.867345094680786, | |
| "learning_rate": 4.1780693533270855e-05, | |
| "loss": 0.3001, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 8.03, | |
| "grad_norm": 7.281511306762695, | |
| "learning_rate": 4.173383317713215e-05, | |
| "loss": 0.3297, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 8.03, | |
| "grad_norm": 10.021977424621582, | |
| "learning_rate": 4.168697282099344e-05, | |
| "loss": 0.2637, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 8.03, | |
| "grad_norm": 7.1175971031188965, | |
| "learning_rate": 4.164011246485474e-05, | |
| "loss": 0.2796, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 8.03, | |
| "eval_accuracy": 0.7177650429799427, | |
| "eval_loss": 0.7974393367767334, | |
| "eval_runtime": 34.0885, | |
| "eval_samples_per_second": 20.476, | |
| "eval_steps_per_second": 1.731, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 7.273170471191406, | |
| "learning_rate": 4.159325210871603e-05, | |
| "loss": 0.2985, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 12.718881607055664, | |
| "learning_rate": 4.154639175257732e-05, | |
| "loss": 0.2799, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 12.384184837341309, | |
| "learning_rate": 4.1499531396438615e-05, | |
| "loss": 0.2361, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 4.5740156173706055, | |
| "learning_rate": 4.145267104029991e-05, | |
| "loss": 0.1935, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 4.967354774475098, | |
| "learning_rate": 4.14058106841612e-05, | |
| "loss": 0.2195, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 2.65854811668396, | |
| "learning_rate": 4.135895032802249e-05, | |
| "loss": 0.1492, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 22.392955780029297, | |
| "learning_rate": 4.131208997188379e-05, | |
| "loss": 0.3631, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 8.278757095336914, | |
| "learning_rate": 4.1265229615745084e-05, | |
| "loss": 0.201, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 10.572164535522461, | |
| "learning_rate": 4.1218369259606376e-05, | |
| "loss": 0.285, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 7.999050617218018, | |
| "learning_rate": 4.117150890346767e-05, | |
| "loss": 0.202, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 0.6098045110702515, | |
| "learning_rate": 4.112464854732897e-05, | |
| "loss": 0.3136, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 13.311070442199707, | |
| "learning_rate": 4.107778819119025e-05, | |
| "loss": 0.2286, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 14.218550682067871, | |
| "learning_rate": 4.1030927835051545e-05, | |
| "loss": 0.1248, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 19.886507034301758, | |
| "learning_rate": 4.098406747891284e-05, | |
| "loss": 0.2682, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 15.638948440551758, | |
| "learning_rate": 4.093720712277414e-05, | |
| "loss": 0.2751, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 16.060192108154297, | |
| "learning_rate": 4.089034676663543e-05, | |
| "loss": 0.1631, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 16.14665412902832, | |
| "learning_rate": 4.084348641049672e-05, | |
| "loss": 0.21, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 12.38498592376709, | |
| "learning_rate": 4.079662605435802e-05, | |
| "loss": 0.2409, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 21.296875, | |
| "learning_rate": 4.074976569821931e-05, | |
| "loss": 0.3526, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 12.729432106018066, | |
| "learning_rate": 4.07029053420806e-05, | |
| "loss": 0.3236, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 5.113853931427002, | |
| "learning_rate": 4.065604498594189e-05, | |
| "loss": 0.116, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 15.853752136230469, | |
| "learning_rate": 4.060918462980319e-05, | |
| "loss": 0.3694, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 3.868764877319336, | |
| "learning_rate": 4.056232427366448e-05, | |
| "loss": 0.1406, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 3.905461072921753, | |
| "learning_rate": 4.0515463917525774e-05, | |
| "loss": 0.2263, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 9.867708206176758, | |
| "learning_rate": 4.0468603561387067e-05, | |
| "loss": 0.241, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 7.449389457702637, | |
| "learning_rate": 4.0421743205248366e-05, | |
| "loss": 0.3362, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 6.9868292808532715, | |
| "learning_rate": 4.037488284910966e-05, | |
| "loss": 0.2709, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 9.967850685119629, | |
| "learning_rate": 4.0328022492970944e-05, | |
| "loss": 0.3062, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 6.262727737426758, | |
| "learning_rate": 4.028116213683224e-05, | |
| "loss": 0.2394, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 9.03, | |
| "grad_norm": 12.531270980834961, | |
| "learning_rate": 4.0234301780693535e-05, | |
| "loss": 0.3621, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 9.03, | |
| "grad_norm": 8.992213249206543, | |
| "learning_rate": 4.018744142455483e-05, | |
| "loss": 0.1948, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 9.03, | |
| "grad_norm": 6.379027843475342, | |
| "learning_rate": 4.014058106841612e-05, | |
| "loss": 0.1911, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 9.03, | |
| "grad_norm": 23.431751251220703, | |
| "learning_rate": 4.009372071227742e-05, | |
| "loss": 0.236, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 9.03, | |
| "eval_accuracy": 0.7722063037249284, | |
| "eval_loss": 0.6850898265838623, | |
| "eval_runtime": 33.99, | |
| "eval_samples_per_second": 20.535, | |
| "eval_steps_per_second": 1.736, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 14.146215438842773, | |
| "learning_rate": 4.004686035613871e-05, | |
| "loss": 0.2476, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 6.308194160461426, | |
| "learning_rate": 4e-05, | |
| "loss": 0.1314, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 12.289512634277344, | |
| "learning_rate": 3.9953139643861296e-05, | |
| "loss": 0.1203, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 15.842235565185547, | |
| "learning_rate": 3.990627928772259e-05, | |
| "loss": 0.2797, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.5130534172058105, | |
| "learning_rate": 3.985941893158388e-05, | |
| "loss": 0.1283, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 11.302055358886719, | |
| "learning_rate": 3.981255857544517e-05, | |
| "loss": 0.3701, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 15.355897903442383, | |
| "learning_rate": 3.976569821930647e-05, | |
| "loss": 0.4371, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 12.716096878051758, | |
| "learning_rate": 3.9718837863167764e-05, | |
| "loss": 0.2433, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 3.5532474517822266, | |
| "learning_rate": 3.9671977507029056e-05, | |
| "loss": 0.0855, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 1.783897876739502, | |
| "learning_rate": 3.962511715089035e-05, | |
| "loss": 0.1193, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 8.46875286102295, | |
| "learning_rate": 3.957825679475165e-05, | |
| "loss": 0.2303, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 16.5199031829834, | |
| "learning_rate": 3.953139643861293e-05, | |
| "loss": 0.254, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 0.3168381452560425, | |
| "learning_rate": 3.9484536082474226e-05, | |
| "loss": 0.1953, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 16.903932571411133, | |
| "learning_rate": 3.943767572633552e-05, | |
| "loss": 0.35, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 8.773260116577148, | |
| "learning_rate": 3.939081537019682e-05, | |
| "loss": 0.2684, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 12.934499740600586, | |
| "learning_rate": 3.934395501405811e-05, | |
| "loss": 0.2493, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "grad_norm": 11.468120574951172, | |
| "learning_rate": 3.92970946579194e-05, | |
| "loss": 0.154, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 0.31763386726379395, | |
| "learning_rate": 3.92502343017807e-05, | |
| "loss": 0.1635, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 0.5305169224739075, | |
| "learning_rate": 3.920337394564199e-05, | |
| "loss": 0.1666, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 23.478593826293945, | |
| "learning_rate": 3.915651358950328e-05, | |
| "loss": 0.1499, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 0.9787814021110535, | |
| "learning_rate": 3.910965323336457e-05, | |
| "loss": 0.3277, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 10.264555931091309, | |
| "learning_rate": 3.906279287722587e-05, | |
| "loss": 0.2312, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 1.865168571472168, | |
| "learning_rate": 3.901593252108716e-05, | |
| "loss": 0.2381, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 13.4818754196167, | |
| "learning_rate": 3.8969072164948455e-05, | |
| "loss": 0.263, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 11.58323860168457, | |
| "learning_rate": 3.892221180880975e-05, | |
| "loss": 0.2925, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 6.375723838806152, | |
| "learning_rate": 3.8875351452671046e-05, | |
| "loss": 0.2063, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 6.340160846710205, | |
| "learning_rate": 3.882849109653234e-05, | |
| "loss": 0.179, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 1.2288185358047485, | |
| "learning_rate": 3.8781630740393624e-05, | |
| "loss": 0.1231, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 4.222806453704834, | |
| "learning_rate": 3.873477038425492e-05, | |
| "loss": 0.198, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 10.03, | |
| "grad_norm": 12.188648223876953, | |
| "learning_rate": 3.8687910028116215e-05, | |
| "loss": 0.0827, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 10.03, | |
| "grad_norm": 16.154399871826172, | |
| "learning_rate": 3.864104967197751e-05, | |
| "loss": 0.2194, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 10.03, | |
| "grad_norm": 25.385496139526367, | |
| "learning_rate": 3.85941893158388e-05, | |
| "loss": 0.2496, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 10.03, | |
| "grad_norm": 0.44989219307899475, | |
| "learning_rate": 3.85473289597001e-05, | |
| "loss": 0.2066, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 10.03, | |
| "eval_accuracy": 0.7707736389684814, | |
| "eval_loss": 0.7625312805175781, | |
| "eval_runtime": 34.4357, | |
| "eval_samples_per_second": 20.27, | |
| "eval_steps_per_second": 1.713, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 15.097335815429688, | |
| "learning_rate": 3.850046860356139e-05, | |
| "loss": 0.1422, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 8.706366539001465, | |
| "learning_rate": 3.8453608247422684e-05, | |
| "loss": 0.1621, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 6.461037635803223, | |
| "learning_rate": 3.8406747891283976e-05, | |
| "loss": 0.2252, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 18.602127075195312, | |
| "learning_rate": 3.835988753514527e-05, | |
| "loss": 0.3387, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 13.997271537780762, | |
| "learning_rate": 3.831302717900656e-05, | |
| "loss": 0.2751, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 6.304378986358643, | |
| "learning_rate": 3.826616682286785e-05, | |
| "loss": 0.0713, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 14.0118408203125, | |
| "learning_rate": 3.821930646672915e-05, | |
| "loss": 0.1648, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 22.292020797729492, | |
| "learning_rate": 3.8172446110590444e-05, | |
| "loss": 0.187, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 3.986797332763672, | |
| "learning_rate": 3.8125585754451737e-05, | |
| "loss": 0.244, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 14.314979553222656, | |
| "learning_rate": 3.807872539831303e-05, | |
| "loss": 0.2893, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 5.869822978973389, | |
| "learning_rate": 3.803186504217433e-05, | |
| "loss": 0.2231, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 2.076101303100586, | |
| "learning_rate": 3.7985004686035613e-05, | |
| "loss": 0.2073, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 0.970169186592102, | |
| "learning_rate": 3.7938144329896906e-05, | |
| "loss": 0.0698, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 6.208909034729004, | |
| "learning_rate": 3.7891283973758205e-05, | |
| "loss": 0.1289, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 17.37769317626953, | |
| "learning_rate": 3.78444236176195e-05, | |
| "loss": 0.1264, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 4.121638298034668, | |
| "learning_rate": 3.779756326148079e-05, | |
| "loss": 0.2226, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 0.10288643091917038, | |
| "learning_rate": 3.775070290534208e-05, | |
| "loss": 0.2066, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 7.0988240242004395, | |
| "learning_rate": 3.770384254920338e-05, | |
| "loss": 0.2174, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 10.114404678344727, | |
| "learning_rate": 3.765698219306467e-05, | |
| "loss": 0.3379, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 4.689489364624023, | |
| "learning_rate": 3.761012183692596e-05, | |
| "loss": 0.1015, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 16.416460037231445, | |
| "learning_rate": 3.756326148078725e-05, | |
| "loss": 0.1955, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 23.091764450073242, | |
| "learning_rate": 3.751640112464855e-05, | |
| "loss": 0.3026, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 8.67167854309082, | |
| "learning_rate": 3.746954076850984e-05, | |
| "loss": 0.197, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 6.001295566558838, | |
| "learning_rate": 3.7422680412371135e-05, | |
| "loss": 0.2166, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 7.23835563659668, | |
| "learning_rate": 3.737582005623243e-05, | |
| "loss": 0.3391, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 5.40344762802124, | |
| "learning_rate": 3.7328959700093726e-05, | |
| "loss": 0.1672, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 6.540796756744385, | |
| "learning_rate": 3.728209934395502e-05, | |
| "loss": 0.0776, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 0.964383065700531, | |
| "learning_rate": 3.7235238987816304e-05, | |
| "loss": 0.1802, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 0.4158385396003723, | |
| "learning_rate": 3.71883786316776e-05, | |
| "loss": 0.0695, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "grad_norm": 17.69159698486328, | |
| "learning_rate": 3.7141518275538895e-05, | |
| "loss": 0.2484, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "grad_norm": 13.761899948120117, | |
| "learning_rate": 3.709465791940019e-05, | |
| "loss": 0.2431, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "grad_norm": 0.3444725275039673, | |
| "learning_rate": 3.704779756326148e-05, | |
| "loss": 0.2122, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "grad_norm": 3.0498321056365967, | |
| "learning_rate": 3.700093720712278e-05, | |
| "loss": 0.2831, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "eval_accuracy": 0.7707736389684814, | |
| "eval_loss": 0.79632568359375, | |
| "eval_runtime": 33.5427, | |
| "eval_samples_per_second": 20.809, | |
| "eval_steps_per_second": 1.759, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.13173483312129974, | |
| "learning_rate": 3.695407685098407e-05, | |
| "loss": 0.0471, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 18.985239028930664, | |
| "learning_rate": 3.6907216494845364e-05, | |
| "loss": 0.0531, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 9.530562400817871, | |
| "learning_rate": 3.6860356138706656e-05, | |
| "loss": 0.2334, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 4.545018672943115, | |
| "learning_rate": 3.681349578256795e-05, | |
| "loss": 0.0164, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 11.210469245910645, | |
| "learning_rate": 3.676663542642924e-05, | |
| "loss": 0.2759, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 15.611169815063477, | |
| "learning_rate": 3.671977507029053e-05, | |
| "loss": 0.1851, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 12.470063209533691, | |
| "learning_rate": 3.667291471415183e-05, | |
| "loss": 0.2275, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 9.493141174316406, | |
| "learning_rate": 3.6626054358013124e-05, | |
| "loss": 0.1243, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 0.25132325291633606, | |
| "learning_rate": 3.657919400187442e-05, | |
| "loss": 0.0885, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 28.46393394470215, | |
| "learning_rate": 3.653233364573571e-05, | |
| "loss": 0.261, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 21.371795654296875, | |
| "learning_rate": 3.648547328959701e-05, | |
| "loss": 0.1305, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 0.9401485323905945, | |
| "learning_rate": 3.6438612933458294e-05, | |
| "loss": 0.1798, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 23.651334762573242, | |
| "learning_rate": 3.6391752577319586e-05, | |
| "loss": 0.1997, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 0.2696411609649658, | |
| "learning_rate": 3.6344892221180885e-05, | |
| "loss": 0.1279, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 26.360898971557617, | |
| "learning_rate": 3.629803186504218e-05, | |
| "loss": 0.2128, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 5.624836444854736, | |
| "learning_rate": 3.625117150890347e-05, | |
| "loss": 0.2319, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "grad_norm": 9.12028980255127, | |
| "learning_rate": 3.620431115276476e-05, | |
| "loss": 0.1358, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 11.657383918762207, | |
| "learning_rate": 3.615745079662606e-05, | |
| "loss": 0.0597, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 9.352867126464844, | |
| "learning_rate": 3.6110590440487353e-05, | |
| "loss": 0.1819, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 33.10483932495117, | |
| "learning_rate": 3.606373008434864e-05, | |
| "loss": 0.2577, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 7.759758472442627, | |
| "learning_rate": 3.601686972820993e-05, | |
| "loss": 0.1209, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 1.226152777671814, | |
| "learning_rate": 3.597000937207123e-05, | |
| "loss": 0.2366, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 4.4371185302734375, | |
| "learning_rate": 3.592314901593252e-05, | |
| "loss": 0.1248, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 29.921714782714844, | |
| "learning_rate": 3.5876288659793815e-05, | |
| "loss": 0.3106, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 12.740315437316895, | |
| "learning_rate": 3.582942830365511e-05, | |
| "loss": 0.1804, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 23.39232635498047, | |
| "learning_rate": 3.5782567947516406e-05, | |
| "loss": 0.134, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 5.175128936767578, | |
| "learning_rate": 3.57357075913777e-05, | |
| "loss": 0.2728, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 18.326560974121094, | |
| "learning_rate": 3.5688847235238984e-05, | |
| "loss": 0.1816, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 0.42216312885284424, | |
| "learning_rate": 3.5641986879100283e-05, | |
| "loss": 0.0879, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 12.03, | |
| "grad_norm": 26.220693588256836, | |
| "learning_rate": 3.5595126522961576e-05, | |
| "loss": 0.094, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 12.03, | |
| "grad_norm": 2.731444835662842, | |
| "learning_rate": 3.554826616682287e-05, | |
| "loss": 0.0832, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 12.03, | |
| "grad_norm": 0.9680123925209045, | |
| "learning_rate": 3.550140581068416e-05, | |
| "loss": 0.0811, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 12.03, | |
| "grad_norm": 69.98845672607422, | |
| "learning_rate": 3.545454545454546e-05, | |
| "loss": 0.1903, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 12.03, | |
| "eval_accuracy": 0.7722063037249284, | |
| "eval_loss": 1.034330129623413, | |
| "eval_runtime": 34.5588, | |
| "eval_samples_per_second": 20.197, | |
| "eval_steps_per_second": 1.707, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 25.59967041015625, | |
| "learning_rate": 3.540768509840675e-05, | |
| "loss": 0.0995, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 24.92293357849121, | |
| "learning_rate": 3.5360824742268044e-05, | |
| "loss": 0.3308, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 30.706470489501953, | |
| "learning_rate": 3.5313964386129336e-05, | |
| "loss": 0.1114, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 1.1878864765167236, | |
| "learning_rate": 3.526710402999063e-05, | |
| "loss": 0.0588, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 31.14521026611328, | |
| "learning_rate": 3.522024367385192e-05, | |
| "loss": 0.1238, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 5.118488311767578, | |
| "learning_rate": 3.517338331771321e-05, | |
| "loss": 0.1793, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 30.64320182800293, | |
| "learning_rate": 3.512652296157451e-05, | |
| "loss": 0.1733, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 0.029796045273542404, | |
| "learning_rate": 3.5079662605435805e-05, | |
| "loss": 0.2219, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 12.55972671508789, | |
| "learning_rate": 3.50328022492971e-05, | |
| "loss": 0.1043, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 6.064866065979004, | |
| "learning_rate": 3.498594189315839e-05, | |
| "loss": 0.1679, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 0.1219923123717308, | |
| "learning_rate": 3.493908153701969e-05, | |
| "loss": 0.1371, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 7.214325904846191, | |
| "learning_rate": 3.4892221180880974e-05, | |
| "loss": 0.05, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 54.42354965209961, | |
| "learning_rate": 3.4845360824742266e-05, | |
| "loss": 0.1519, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 0.27510225772857666, | |
| "learning_rate": 3.4798500468603565e-05, | |
| "loss": 0.1608, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 0.35241690278053284, | |
| "learning_rate": 3.475164011246486e-05, | |
| "loss": 0.1467, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 17.937379837036133, | |
| "learning_rate": 3.470477975632615e-05, | |
| "loss": 0.1619, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 13.159626960754395, | |
| "learning_rate": 3.465791940018744e-05, | |
| "loss": 0.1981, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 19.368940353393555, | |
| "learning_rate": 3.461105904404874e-05, | |
| "loss": 0.3174, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 0.7653603553771973, | |
| "learning_rate": 3.4564198687910034e-05, | |
| "loss": 0.2421, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 2.392252206802368, | |
| "learning_rate": 3.451733833177132e-05, | |
| "loss": 0.248, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 1.3327248096466064, | |
| "learning_rate": 3.447047797563261e-05, | |
| "loss": 0.0903, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 8.378661155700684, | |
| "learning_rate": 3.442361761949391e-05, | |
| "loss": 0.1196, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 17.592124938964844, | |
| "learning_rate": 3.43767572633552e-05, | |
| "loss": 0.1543, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 22.954425811767578, | |
| "learning_rate": 3.4329896907216495e-05, | |
| "loss": 0.1726, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 11.798051834106445, | |
| "learning_rate": 3.428303655107779e-05, | |
| "loss": 0.3255, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 20.77471923828125, | |
| "learning_rate": 3.423617619493909e-05, | |
| "loss": 0.2043, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 5.149378299713135, | |
| "learning_rate": 3.418931583880038e-05, | |
| "loss": 0.0589, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 8.951346397399902, | |
| "learning_rate": 3.4142455482661665e-05, | |
| "loss": 0.207, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "grad_norm": 0.9169898629188538, | |
| "learning_rate": 3.4095595126522964e-05, | |
| "loss": 0.1265, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 13.03, | |
| "grad_norm": 13.90592098236084, | |
| "learning_rate": 3.4048734770384256e-05, | |
| "loss": 0.2189, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 13.03, | |
| "grad_norm": 23.327816009521484, | |
| "learning_rate": 3.400187441424555e-05, | |
| "loss": 0.184, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 13.03, | |
| "grad_norm": 14.074636459350586, | |
| "learning_rate": 3.395501405810684e-05, | |
| "loss": 0.1103, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 13.03, | |
| "grad_norm": 9.187418937683105, | |
| "learning_rate": 3.390815370196814e-05, | |
| "loss": 0.1169, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 13.03, | |
| "eval_accuracy": 0.7865329512893983, | |
| "eval_loss": 0.8527703285217285, | |
| "eval_runtime": 34.2781, | |
| "eval_samples_per_second": 20.363, | |
| "eval_steps_per_second": 1.721, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 11.900031089782715, | |
| "learning_rate": 3.386129334582943e-05, | |
| "loss": 0.074, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 11.011795043945312, | |
| "learning_rate": 3.3814432989690724e-05, | |
| "loss": 0.126, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 23.668609619140625, | |
| "learning_rate": 3.376757263355202e-05, | |
| "loss": 0.0747, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.17650875449180603, | |
| "learning_rate": 3.372071227741331e-05, | |
| "loss": 0.0751, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 5.125329971313477, | |
| "learning_rate": 3.36738519212746e-05, | |
| "loss": 0.2435, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 21.372135162353516, | |
| "learning_rate": 3.3626991565135894e-05, | |
| "loss": 0.1759, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 0.13896770775318146, | |
| "learning_rate": 3.358013120899719e-05, | |
| "loss": 0.0661, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 0.020779293030500412, | |
| "learning_rate": 3.3533270852858485e-05, | |
| "loss": 0.1733, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 14.163589477539062, | |
| "learning_rate": 3.348641049671978e-05, | |
| "loss": 0.1317, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 1.3063100576400757, | |
| "learning_rate": 3.343955014058107e-05, | |
| "loss": 0.1116, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 2.3075990676879883, | |
| "learning_rate": 3.339268978444237e-05, | |
| "loss": 0.0674, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 10.278874397277832, | |
| "learning_rate": 3.3345829428303654e-05, | |
| "loss": 0.2285, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 28.758888244628906, | |
| "learning_rate": 3.3298969072164947e-05, | |
| "loss": 0.0825, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 10.479151725769043, | |
| "learning_rate": 3.3252108716026246e-05, | |
| "loss": 0.1451, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 0.03889832645654678, | |
| "learning_rate": 3.320524835988754e-05, | |
| "loss": 0.0653, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 1.4351927042007446, | |
| "learning_rate": 3.315838800374883e-05, | |
| "loss": 0.1306, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "grad_norm": 28.180889129638672, | |
| "learning_rate": 3.311152764761012e-05, | |
| "loss": 0.2123, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 17.614906311035156, | |
| "learning_rate": 3.306466729147142e-05, | |
| "loss": 0.0832, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 0.06731338798999786, | |
| "learning_rate": 3.3017806935332714e-05, | |
| "loss": 0.0583, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 1.4568157196044922, | |
| "learning_rate": 3.2970946579194e-05, | |
| "loss": 0.2127, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 7.513918876647949, | |
| "learning_rate": 3.292408622305529e-05, | |
| "loss": 0.1357, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 0.8259130120277405, | |
| "learning_rate": 3.287722586691659e-05, | |
| "loss": 0.04, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 0.1765613704919815, | |
| "learning_rate": 3.283036551077788e-05, | |
| "loss": 0.086, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 15.376742362976074, | |
| "learning_rate": 3.2783505154639176e-05, | |
| "loss": 0.2212, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 0.4169343411922455, | |
| "learning_rate": 3.273664479850047e-05, | |
| "loss": 0.0478, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 43.11495590209961, | |
| "learning_rate": 3.268978444236177e-05, | |
| "loss": 0.096, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 15.717057228088379, | |
| "learning_rate": 3.264292408622306e-05, | |
| "loss": 0.154, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 24.132003784179688, | |
| "learning_rate": 3.2596063730084345e-05, | |
| "loss": 0.1402, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 9.180221557617188, | |
| "learning_rate": 3.2549203373945644e-05, | |
| "loss": 0.1438, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 14.03, | |
| "grad_norm": 17.912752151489258, | |
| "learning_rate": 3.2502343017806936e-05, | |
| "loss": 0.3277, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 14.03, | |
| "grad_norm": 23.05970573425293, | |
| "learning_rate": 3.245548266166823e-05, | |
| "loss": 0.1899, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 14.03, | |
| "grad_norm": 5.295262813568115, | |
| "learning_rate": 3.240862230552952e-05, | |
| "loss": 0.2962, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 14.03, | |
| "grad_norm": 0.08439239114522934, | |
| "learning_rate": 3.236176194939082e-05, | |
| "loss": 0.3502, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 14.03, | |
| "eval_accuracy": 0.7965616045845272, | |
| "eval_loss": 0.9265322089195251, | |
| "eval_runtime": 33.6194, | |
| "eval_samples_per_second": 20.762, | |
| "eval_steps_per_second": 1.755, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 27.462636947631836, | |
| "learning_rate": 3.231490159325211e-05, | |
| "loss": 0.1507, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.03183213621377945, | |
| "learning_rate": 3.2268041237113405e-05, | |
| "loss": 0.0413, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 24.36420440673828, | |
| "learning_rate": 3.22211808809747e-05, | |
| "loss": 0.1499, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 2.348003387451172, | |
| "learning_rate": 3.217432052483599e-05, | |
| "loss": 0.2091, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 14.074554443359375, | |
| "learning_rate": 3.212746016869728e-05, | |
| "loss": 0.1966, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 13.325398445129395, | |
| "learning_rate": 3.2080599812558574e-05, | |
| "loss": 0.0982, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 0.06735904514789581, | |
| "learning_rate": 3.203373945641987e-05, | |
| "loss": 0.0819, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 7.472801685333252, | |
| "learning_rate": 3.1986879100281165e-05, | |
| "loss": 0.1468, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 4.233068943023682, | |
| "learning_rate": 3.194001874414246e-05, | |
| "loss": 0.03, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 9.645890235900879, | |
| "learning_rate": 3.189315838800375e-05, | |
| "loss": 0.1579, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 18.024099349975586, | |
| "learning_rate": 3.184629803186505e-05, | |
| "loss": 0.1401, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 38.932167053222656, | |
| "learning_rate": 3.1799437675726335e-05, | |
| "loss": 0.1337, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 0.6836027503013611, | |
| "learning_rate": 3.175257731958763e-05, | |
| "loss": 0.1067, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 13.472026824951172, | |
| "learning_rate": 3.1705716963448926e-05, | |
| "loss": 0.137, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 49.68353271484375, | |
| "learning_rate": 3.165885660731022e-05, | |
| "loss": 0.2201, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 0.10768305510282516, | |
| "learning_rate": 3.161199625117151e-05, | |
| "loss": 0.0503, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 0.02888442948460579, | |
| "learning_rate": 3.15651358950328e-05, | |
| "loss": 0.2246, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 11.260727882385254, | |
| "learning_rate": 3.15182755388941e-05, | |
| "loss": 0.1655, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 0.16832049190998077, | |
| "learning_rate": 3.1471415182755394e-05, | |
| "loss": 0.2107, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 0.5747199058532715, | |
| "learning_rate": 3.142455482661668e-05, | |
| "loss": 0.1302, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 6.570047378540039, | |
| "learning_rate": 3.137769447047797e-05, | |
| "loss": 0.1102, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 0.02458810992538929, | |
| "learning_rate": 3.133083411433927e-05, | |
| "loss": 0.0711, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 28.233722686767578, | |
| "learning_rate": 3.1283973758200564e-05, | |
| "loss": 0.1243, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 34.339900970458984, | |
| "learning_rate": 3.1237113402061856e-05, | |
| "loss": 0.2122, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 0.042488761246204376, | |
| "learning_rate": 3.119025304592315e-05, | |
| "loss": 0.0696, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 10.747623443603516, | |
| "learning_rate": 3.114339268978445e-05, | |
| "loss": 0.0881, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 0.14769670367240906, | |
| "learning_rate": 3.109653233364574e-05, | |
| "loss": 0.2615, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 6.627384662628174, | |
| "learning_rate": 3.104967197750703e-05, | |
| "loss": 0.105, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "grad_norm": 0.06998647749423981, | |
| "learning_rate": 3.1002811621368324e-05, | |
| "loss": 0.1129, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 15.03, | |
| "grad_norm": 14.90441608428955, | |
| "learning_rate": 3.0955951265229617e-05, | |
| "loss": 0.1303, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 15.03, | |
| "grad_norm": 0.6457139849662781, | |
| "learning_rate": 3.090909090909091e-05, | |
| "loss": 0.0176, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 15.03, | |
| "grad_norm": 28.25990867614746, | |
| "learning_rate": 3.08622305529522e-05, | |
| "loss": 0.1868, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 15.03, | |
| "grad_norm": 18.335277557373047, | |
| "learning_rate": 3.08153701968135e-05, | |
| "loss": 0.1728, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 15.03, | |
| "eval_accuracy": 0.8209169054441261, | |
| "eval_loss": 0.8522208333015442, | |
| "eval_runtime": 33.7106, | |
| "eval_samples_per_second": 20.706, | |
| "eval_steps_per_second": 1.75, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.026595309376716614, | |
| "learning_rate": 3.076850984067479e-05, | |
| "loss": 0.0047, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 16.674278259277344, | |
| "learning_rate": 3.0721649484536085e-05, | |
| "loss": 0.1096, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.7340277433395386, | |
| "learning_rate": 3.067478912839738e-05, | |
| "loss": 0.1344, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.22693070769309998, | |
| "learning_rate": 3.062792877225867e-05, | |
| "loss": 0.1013, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.07706239074468613, | |
| "learning_rate": 3.058106841611996e-05, | |
| "loss": 0.1392, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 18.134891510009766, | |
| "learning_rate": 3.0534208059981254e-05, | |
| "loss": 0.06, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 1.382936716079712, | |
| "learning_rate": 3.048734770384255e-05, | |
| "loss": 0.1596, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 20.703399658203125, | |
| "learning_rate": 3.0440487347703846e-05, | |
| "loss": 0.0695, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 1.053179144859314, | |
| "learning_rate": 3.0393626991565138e-05, | |
| "loss": 0.1434, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 0.017900103703141212, | |
| "learning_rate": 3.0346766635426434e-05, | |
| "loss": 0.0591, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 0.056633152067661285, | |
| "learning_rate": 3.0299906279287726e-05, | |
| "loss": 0.1733, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 0.11656715720891953, | |
| "learning_rate": 3.0253045923149015e-05, | |
| "loss": 0.1159, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 2.425400972366333, | |
| "learning_rate": 3.020618556701031e-05, | |
| "loss": 0.1666, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 4.859916687011719, | |
| "learning_rate": 3.0159325210871603e-05, | |
| "loss": 0.0466, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 29.941722869873047, | |
| "learning_rate": 3.01124648547329e-05, | |
| "loss": 0.1313, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 1.4672462940216064, | |
| "learning_rate": 3.006560449859419e-05, | |
| "loss": 0.066, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "grad_norm": 20.58746910095215, | |
| "learning_rate": 3.0018744142455487e-05, | |
| "loss": 0.0488, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 19.143203735351562, | |
| "learning_rate": 2.997188378631678e-05, | |
| "loss": 0.1403, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 21.062692642211914, | |
| "learning_rate": 2.9925023430178075e-05, | |
| "loss": 0.0366, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 1.4022785425186157, | |
| "learning_rate": 2.987816307403936e-05, | |
| "loss": 0.0344, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 0.7755998373031616, | |
| "learning_rate": 2.9831302717900656e-05, | |
| "loss": 0.1005, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 0.02092009223997593, | |
| "learning_rate": 2.9784442361761948e-05, | |
| "loss": 0.1796, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 27.31386375427246, | |
| "learning_rate": 2.9737582005623244e-05, | |
| "loss": 0.1263, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 3.5492141246795654, | |
| "learning_rate": 2.9690721649484536e-05, | |
| "loss": 0.2096, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 23.727418899536133, | |
| "learning_rate": 2.9643861293345832e-05, | |
| "loss": 0.1176, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 2.065145492553711, | |
| "learning_rate": 2.9597000937207124e-05, | |
| "loss": 0.0873, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 1.5445265769958496, | |
| "learning_rate": 2.955014058106842e-05, | |
| "loss": 0.1622, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 3.4327638149261475, | |
| "learning_rate": 2.9503280224929712e-05, | |
| "loss": 0.1375, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 0.018016502261161804, | |
| "learning_rate": 2.9456419868791e-05, | |
| "loss": 0.0712, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "grad_norm": 11.649871826171875, | |
| "learning_rate": 2.9409559512652297e-05, | |
| "loss": 0.1606, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "grad_norm": 17.827857971191406, | |
| "learning_rate": 2.936269915651359e-05, | |
| "loss": 0.1209, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "grad_norm": 0.01631985232234001, | |
| "learning_rate": 2.9315838800374885e-05, | |
| "loss": 0.0295, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "grad_norm": 48.93087387084961, | |
| "learning_rate": 2.9268978444236177e-05, | |
| "loss": 0.0542, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "eval_accuracy": 0.8051575931232091, | |
| "eval_loss": 1.0106927156448364, | |
| "eval_runtime": 33.4281, | |
| "eval_samples_per_second": 20.881, | |
| "eval_steps_per_second": 1.765, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 3.7330472469329834, | |
| "learning_rate": 2.9222118088097473e-05, | |
| "loss": 0.1299, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 1.1770386695861816, | |
| "learning_rate": 2.9175257731958765e-05, | |
| "loss": 0.1494, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 6.161425590515137, | |
| "learning_rate": 2.912839737582006e-05, | |
| "loss": 0.0633, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.5936900973320007, | |
| "learning_rate": 2.908153701968135e-05, | |
| "loss": 0.0759, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.04893672466278076, | |
| "learning_rate": 2.9034676663542642e-05, | |
| "loss": 0.0792, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 6.019880294799805, | |
| "learning_rate": 2.8987816307403938e-05, | |
| "loss": 0.1079, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 45.10232162475586, | |
| "learning_rate": 2.894095595126523e-05, | |
| "loss": 0.0538, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 52.10890197753906, | |
| "learning_rate": 2.8894095595126526e-05, | |
| "loss": 0.0725, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 0.03488897159695625, | |
| "learning_rate": 2.8847235238987818e-05, | |
| "loss": 0.1523, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 0.07146196067333221, | |
| "learning_rate": 2.8800374882849114e-05, | |
| "loss": 0.0272, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 9.103580474853516, | |
| "learning_rate": 2.8753514526710406e-05, | |
| "loss": 0.0956, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 0.010441714897751808, | |
| "learning_rate": 2.8706654170571695e-05, | |
| "loss": 0.1186, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 51.32048797607422, | |
| "learning_rate": 2.865979381443299e-05, | |
| "loss": 0.0991, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 17.720178604125977, | |
| "learning_rate": 2.8612933458294283e-05, | |
| "loss": 0.0702, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 31.01833724975586, | |
| "learning_rate": 2.856607310215558e-05, | |
| "loss": 0.175, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 0.02792373113334179, | |
| "learning_rate": 2.851921274601687e-05, | |
| "loss": 0.0858, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 48.3105583190918, | |
| "learning_rate": 2.8472352389878167e-05, | |
| "loss": 0.2053, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 9.974466323852539, | |
| "learning_rate": 2.842549203373946e-05, | |
| "loss": 0.0544, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 0.7237229347229004, | |
| "learning_rate": 2.8378631677600755e-05, | |
| "loss": 0.1306, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 14.930671691894531, | |
| "learning_rate": 2.833177132146204e-05, | |
| "loss": 0.1952, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 0.019687309861183167, | |
| "learning_rate": 2.8284910965323336e-05, | |
| "loss": 0.1069, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 5.928595066070557, | |
| "learning_rate": 2.823805060918463e-05, | |
| "loss": 0.0161, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 0.03599075973033905, | |
| "learning_rate": 2.8191190253045924e-05, | |
| "loss": 0.09, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 0.05025621876120567, | |
| "learning_rate": 2.8144329896907216e-05, | |
| "loss": 0.0732, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 40.76129150390625, | |
| "learning_rate": 2.8097469540768512e-05, | |
| "loss": 0.0456, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 1.6205233335494995, | |
| "learning_rate": 2.8050609184629804e-05, | |
| "loss": 0.0213, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 56.61774444580078, | |
| "learning_rate": 2.80037488284911e-05, | |
| "loss": 0.2101, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 0.4398052394390106, | |
| "learning_rate": 2.7956888472352392e-05, | |
| "loss": 0.1504, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 3.440358877182007, | |
| "learning_rate": 2.791002811621368e-05, | |
| "loss": 0.0786, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 17.03, | |
| "grad_norm": 0.02337775193154812, | |
| "learning_rate": 2.7863167760074977e-05, | |
| "loss": 0.1404, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 17.03, | |
| "grad_norm": 36.56870651245117, | |
| "learning_rate": 2.781630740393627e-05, | |
| "loss": 0.0849, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 17.03, | |
| "grad_norm": 0.021380068734288216, | |
| "learning_rate": 2.7769447047797565e-05, | |
| "loss": 0.0257, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 17.03, | |
| "grad_norm": 0.030316907912492752, | |
| "learning_rate": 2.7722586691658857e-05, | |
| "loss": 0.0711, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 17.03, | |
| "eval_accuracy": 0.8080229226361032, | |
| "eval_loss": 0.9795148968696594, | |
| "eval_runtime": 34.1521, | |
| "eval_samples_per_second": 20.438, | |
| "eval_steps_per_second": 1.728, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.007565322797745466, | |
| "learning_rate": 2.7675726335520153e-05, | |
| "loss": 0.0015, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.006239545065909624, | |
| "learning_rate": 2.7628865979381445e-05, | |
| "loss": 0.0589, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.024090183898806572, | |
| "learning_rate": 2.758200562324274e-05, | |
| "loss": 0.0892, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 8.415689468383789, | |
| "learning_rate": 2.753514526710403e-05, | |
| "loss": 0.1953, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.46405917406082153, | |
| "learning_rate": 2.7488284910965322e-05, | |
| "loss": 0.0365, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 0.320950984954834, | |
| "learning_rate": 2.7441424554826618e-05, | |
| "loss": 0.1011, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 25.849971771240234, | |
| "learning_rate": 2.739456419868791e-05, | |
| "loss": 0.0787, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 0.018421674147248268, | |
| "learning_rate": 2.7347703842549206e-05, | |
| "loss": 0.0032, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 0.02502196654677391, | |
| "learning_rate": 2.73008434864105e-05, | |
| "loss": 0.173, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 0.05541960895061493, | |
| "learning_rate": 2.7253983130271794e-05, | |
| "loss": 0.1086, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 1.3273464441299438, | |
| "learning_rate": 2.7207122774133086e-05, | |
| "loss": 0.0214, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 27.578462600708008, | |
| "learning_rate": 2.7160262417994375e-05, | |
| "loss": 0.0767, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 1.713114857673645, | |
| "learning_rate": 2.711340206185567e-05, | |
| "loss": 0.1884, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 0.0781659409403801, | |
| "learning_rate": 2.7066541705716963e-05, | |
| "loss": 0.1062, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 61.0845832824707, | |
| "learning_rate": 2.701968134957826e-05, | |
| "loss": 0.0516, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 9.772195816040039, | |
| "learning_rate": 2.697282099343955e-05, | |
| "loss": 0.0589, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "grad_norm": 0.18929840624332428, | |
| "learning_rate": 2.6925960637300847e-05, | |
| "loss": 0.1057, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 0.020669307559728622, | |
| "learning_rate": 2.687910028116214e-05, | |
| "loss": 0.0687, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 5.4998016357421875, | |
| "learning_rate": 2.6832239925023435e-05, | |
| "loss": 0.2466, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 2.288482427597046, | |
| "learning_rate": 2.6785379568884727e-05, | |
| "loss": 0.0868, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 0.02619881182909012, | |
| "learning_rate": 2.6738519212746016e-05, | |
| "loss": 0.0071, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 0.011477050371468067, | |
| "learning_rate": 2.669165885660731e-05, | |
| "loss": 0.0421, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 0.03269781917333603, | |
| "learning_rate": 2.6644798500468604e-05, | |
| "loss": 0.0465, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 0.03204688802361488, | |
| "learning_rate": 2.6597938144329897e-05, | |
| "loss": 0.0769, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 28.57611656188965, | |
| "learning_rate": 2.6551077788191192e-05, | |
| "loss": 0.0488, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 27.48614501953125, | |
| "learning_rate": 2.6504217432052485e-05, | |
| "loss": 0.0817, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 37.647117614746094, | |
| "learning_rate": 2.645735707591378e-05, | |
| "loss": 0.0928, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 0.01060717087239027, | |
| "learning_rate": 2.6410496719775073e-05, | |
| "loss": 0.034, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 0.24268437922000885, | |
| "learning_rate": 2.636363636363636e-05, | |
| "loss": 0.108, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 18.03, | |
| "grad_norm": 0.00554469833150506, | |
| "learning_rate": 2.6316776007497657e-05, | |
| "loss": 0.1068, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 18.03, | |
| "grad_norm": 10.710355758666992, | |
| "learning_rate": 2.626991565135895e-05, | |
| "loss": 0.0637, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 18.03, | |
| "grad_norm": 23.989526748657227, | |
| "learning_rate": 2.6223055295220245e-05, | |
| "loss": 0.0926, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 18.03, | |
| "grad_norm": 0.2239922732114792, | |
| "learning_rate": 2.6176194939081538e-05, | |
| "loss": 0.0287, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 18.03, | |
| "eval_accuracy": 0.8094555873925502, | |
| "eval_loss": 1.1470834016799927, | |
| "eval_runtime": 34.1919, | |
| "eval_samples_per_second": 20.414, | |
| "eval_steps_per_second": 1.726, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.3733896315097809, | |
| "learning_rate": 2.6129334582942833e-05, | |
| "loss": 0.0541, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.011687755584716797, | |
| "learning_rate": 2.6082474226804126e-05, | |
| "loss": 0.0725, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 15.612820625305176, | |
| "learning_rate": 2.603561387066542e-05, | |
| "loss": 0.0287, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.0473637618124485, | |
| "learning_rate": 2.598875351452671e-05, | |
| "loss": 0.128, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 14.158834457397461, | |
| "learning_rate": 2.5941893158388003e-05, | |
| "loss": 0.0102, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 0.03251667320728302, | |
| "learning_rate": 2.58950328022493e-05, | |
| "loss": 0.0536, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 2.301870822906494, | |
| "learning_rate": 2.584817244611059e-05, | |
| "loss": 0.1364, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 36.87099075317383, | |
| "learning_rate": 2.5801312089971886e-05, | |
| "loss": 0.1997, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 0.012598090805113316, | |
| "learning_rate": 2.575445173383318e-05, | |
| "loss": 0.1696, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 0.016358235850930214, | |
| "learning_rate": 2.5707591377694474e-05, | |
| "loss": 0.0671, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 3.8293418884277344, | |
| "learning_rate": 2.5660731021555767e-05, | |
| "loss": 0.0023, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 32.61289978027344, | |
| "learning_rate": 2.5613870665417056e-05, | |
| "loss": 0.0762, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 0.12384048849344254, | |
| "learning_rate": 2.556701030927835e-05, | |
| "loss": 0.1048, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 34.71640396118164, | |
| "learning_rate": 2.5520149953139644e-05, | |
| "loss": 0.075, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 0.02384701929986477, | |
| "learning_rate": 2.547328959700094e-05, | |
| "loss": 0.0328, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 0.028293780982494354, | |
| "learning_rate": 2.542642924086223e-05, | |
| "loss": 0.0533, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 0.06875205785036087, | |
| "learning_rate": 2.5379568884723527e-05, | |
| "loss": 0.0736, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 0.020702671259641647, | |
| "learning_rate": 2.533270852858482e-05, | |
| "loss": 0.1326, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 27.13255500793457, | |
| "learning_rate": 2.5285848172446115e-05, | |
| "loss": 0.1638, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 0.04475679248571396, | |
| "learning_rate": 2.5238987816307408e-05, | |
| "loss": 0.0483, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 0.3056570291519165, | |
| "learning_rate": 2.5192127460168697e-05, | |
| "loss": 0.0657, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 0.01589689962565899, | |
| "learning_rate": 2.514526710402999e-05, | |
| "loss": 0.033, | |
| "step": 6490 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 0.02305714786052704, | |
| "learning_rate": 2.5098406747891285e-05, | |
| "loss": 0.1051, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 0.0253700353205204, | |
| "learning_rate": 2.5051546391752577e-05, | |
| "loss": 0.0328, | |
| "step": 6510 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 46.60805892944336, | |
| "learning_rate": 2.5004686035613873e-05, | |
| "loss": 0.0829, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 0.07593350857496262, | |
| "learning_rate": 2.4957825679475165e-05, | |
| "loss": 0.0278, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 0.011305336840450764, | |
| "learning_rate": 2.4910965323336457e-05, | |
| "loss": 0.1844, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 0.02104657143354416, | |
| "learning_rate": 2.4864104967197753e-05, | |
| "loss": 0.0412, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 0.05700293555855751, | |
| "learning_rate": 2.4817244611059045e-05, | |
| "loss": 0.004, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 19.03, | |
| "grad_norm": 5.577692985534668, | |
| "learning_rate": 2.477038425492034e-05, | |
| "loss": 0.1151, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 19.03, | |
| "grad_norm": 0.020554441958665848, | |
| "learning_rate": 2.472352389878163e-05, | |
| "loss": 0.0564, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 19.03, | |
| "grad_norm": 0.04367370158433914, | |
| "learning_rate": 2.4676663542642926e-05, | |
| "loss": 0.1458, | |
| "step": 6590 | |
| }, | |
| { | |
| "epoch": 19.03, | |
| "grad_norm": 0.026535367593169212, | |
| "learning_rate": 2.4629803186504218e-05, | |
| "loss": 0.1011, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 19.03, | |
| "eval_accuracy": 0.7851002865329513, | |
| "eval_loss": 1.0893527269363403, | |
| "eval_runtime": 33.9177, | |
| "eval_samples_per_second": 20.579, | |
| "eval_steps_per_second": 1.74, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 10.764158248901367, | |
| "learning_rate": 2.4582942830365514e-05, | |
| "loss": 0.159, | |
| "step": 6610 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.06043768674135208, | |
| "learning_rate": 2.4536082474226803e-05, | |
| "loss": 0.003, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.006598788313567638, | |
| "learning_rate": 2.4489222118088098e-05, | |
| "loss": 0.0761, | |
| "step": 6630 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.02938709408044815, | |
| "learning_rate": 2.444236176194939e-05, | |
| "loss": 0.069, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.7408674955368042, | |
| "learning_rate": 2.4395501405810686e-05, | |
| "loss": 0.0754, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 14.34585952758789, | |
| "learning_rate": 2.434864104967198e-05, | |
| "loss": 0.0301, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 0.12531670928001404, | |
| "learning_rate": 2.430178069353327e-05, | |
| "loss": 0.0021, | |
| "step": 6670 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 0.11880137771368027, | |
| "learning_rate": 2.4254920337394567e-05, | |
| "loss": 0.0008, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 15.903475761413574, | |
| "learning_rate": 2.420805998125586e-05, | |
| "loss": 0.0479, | |
| "step": 6690 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 29.952171325683594, | |
| "learning_rate": 2.416119962511715e-05, | |
| "loss": 0.1002, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 0.30753955245018005, | |
| "learning_rate": 2.4114339268978444e-05, | |
| "loss": 0.0304, | |
| "step": 6710 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 0.014064520597457886, | |
| "learning_rate": 2.406747891283974e-05, | |
| "loss": 0.1304, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 0.4187520146369934, | |
| "learning_rate": 2.402061855670103e-05, | |
| "loss": 0.1015, | |
| "step": 6730 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 0.02327924221754074, | |
| "learning_rate": 2.3973758200562327e-05, | |
| "loss": 0.0053, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 16.745342254638672, | |
| "learning_rate": 2.392689784442362e-05, | |
| "loss": 0.2538, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 0.24011173844337463, | |
| "learning_rate": 2.3880037488284912e-05, | |
| "loss": 0.0485, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 20.01, | |
| "grad_norm": 29.788806915283203, | |
| "learning_rate": 2.3833177132146208e-05, | |
| "loss": 0.0324, | |
| "step": 6770 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 0.7102867960929871, | |
| "learning_rate": 2.37863167760075e-05, | |
| "loss": 0.1429, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 0.08376511186361313, | |
| "learning_rate": 2.3739456419868792e-05, | |
| "loss": 0.0972, | |
| "step": 6790 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 0.335637629032135, | |
| "learning_rate": 2.3692596063730085e-05, | |
| "loss": 0.062, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 28.6505069732666, | |
| "learning_rate": 2.364573570759138e-05, | |
| "loss": 0.2001, | |
| "step": 6810 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 29.769094467163086, | |
| "learning_rate": 2.3598875351452673e-05, | |
| "loss": 0.1555, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 52.66693878173828, | |
| "learning_rate": 2.3552014995313965e-05, | |
| "loss": 0.0873, | |
| "step": 6830 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 0.011765277944505215, | |
| "learning_rate": 2.3505154639175257e-05, | |
| "loss": 0.0721, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 0.08442965894937515, | |
| "learning_rate": 2.3458294283036553e-05, | |
| "loss": 0.0497, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 0.017878804355859756, | |
| "learning_rate": 2.3411433926897845e-05, | |
| "loss": 0.0389, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 0.017131801694631577, | |
| "learning_rate": 2.3364573570759138e-05, | |
| "loss": 0.1382, | |
| "step": 6870 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 2.3948495388031006, | |
| "learning_rate": 2.3317713214620433e-05, | |
| "loss": 0.0425, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "grad_norm": 9.163774490356445, | |
| "learning_rate": 2.3270852858481726e-05, | |
| "loss": 0.0738, | |
| "step": 6890 | |
| }, | |
| { | |
| "epoch": 20.03, | |
| "grad_norm": 0.009526137262582779, | |
| "learning_rate": 2.322399250234302e-05, | |
| "loss": 0.1145, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 20.03, | |
| "grad_norm": 0.11400933563709259, | |
| "learning_rate": 2.317713214620431e-05, | |
| "loss": 0.0527, | |
| "step": 6910 | |
| }, | |
| { | |
| "epoch": 20.03, | |
| "grad_norm": 1.8810824155807495, | |
| "learning_rate": 2.3130271790065606e-05, | |
| "loss": 0.0034, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 20.03, | |
| "grad_norm": 0.02618522383272648, | |
| "learning_rate": 2.3083411433926898e-05, | |
| "loss": 0.0424, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 20.03, | |
| "eval_accuracy": 0.7822349570200573, | |
| "eval_loss": 1.144364356994629, | |
| "eval_runtime": 34.682, | |
| "eval_samples_per_second": 20.126, | |
| "eval_steps_per_second": 1.701, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 0.017015540972352028, | |
| "learning_rate": 2.3036551077788194e-05, | |
| "loss": 0.0144, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 1.5228683948516846, | |
| "learning_rate": 2.2989690721649483e-05, | |
| "loss": 0.049, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 0.00678257504478097, | |
| "learning_rate": 2.294283036551078e-05, | |
| "loss": 0.0779, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 0.011534970253705978, | |
| "learning_rate": 2.289597000937207e-05, | |
| "loss": 0.1179, | |
| "step": 6970 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 22.936405181884766, | |
| "learning_rate": 2.2849109653233367e-05, | |
| "loss": 0.0972, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 0.06261157989501953, | |
| "learning_rate": 2.280224929709466e-05, | |
| "loss": 0.0625, | |
| "step": 6990 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 9.488935470581055, | |
| "learning_rate": 2.275538894095595e-05, | |
| "loss": 0.0421, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 11.307842254638672, | |
| "learning_rate": 2.2708528584817247e-05, | |
| "loss": 0.043, | |
| "step": 7010 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 0.03459528461098671, | |
| "learning_rate": 2.266166822867854e-05, | |
| "loss": 0.0242, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 9.111845016479492, | |
| "learning_rate": 2.261480787253983e-05, | |
| "loss": 0.0573, | |
| "step": 7030 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 0.01025596633553505, | |
| "learning_rate": 2.2567947516401124e-05, | |
| "loss": 0.0159, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 2.14906907081604, | |
| "learning_rate": 2.252108716026242e-05, | |
| "loss": 0.0856, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 11.889404296875, | |
| "learning_rate": 2.2474226804123712e-05, | |
| "loss": 0.0287, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 0.046489182859659195, | |
| "learning_rate": 2.2427366447985008e-05, | |
| "loss": 0.0729, | |
| "step": 7070 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 0.009697903878986835, | |
| "learning_rate": 2.23805060918463e-05, | |
| "loss": 0.0428, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 0.03072419762611389, | |
| "learning_rate": 2.2333645735707592e-05, | |
| "loss": 0.0601, | |
| "step": 7090 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "grad_norm": 10.731389045715332, | |
| "learning_rate": 2.2286785379568888e-05, | |
| "loss": 0.0125, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 0.04367499053478241, | |
| "learning_rate": 2.223992502343018e-05, | |
| "loss": 0.0223, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 0.013104673475027084, | |
| "learning_rate": 2.2193064667291473e-05, | |
| "loss": 0.1196, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 0.5789304375648499, | |
| "learning_rate": 2.2146204311152765e-05, | |
| "loss": 0.0593, | |
| "step": 7130 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 0.005018654279410839, | |
| "learning_rate": 2.209934395501406e-05, | |
| "loss": 0.0508, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 2.954718828201294, | |
| "learning_rate": 2.2052483598875353e-05, | |
| "loss": 0.0327, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 9.058338165283203, | |
| "learning_rate": 2.2005623242736645e-05, | |
| "loss": 0.0686, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 0.009537008590996265, | |
| "learning_rate": 2.1958762886597937e-05, | |
| "loss": 0.0273, | |
| "step": 7170 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 0.009294161573052406, | |
| "learning_rate": 2.1911902530459233e-05, | |
| "loss": 0.0853, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 0.31812822818756104, | |
| "learning_rate": 2.1865042174320525e-05, | |
| "loss": 0.0038, | |
| "step": 7190 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 21.410701751708984, | |
| "learning_rate": 2.1818181818181818e-05, | |
| "loss": 0.053, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 0.6843286752700806, | |
| "learning_rate": 2.1771321462043114e-05, | |
| "loss": 0.0013, | |
| "step": 7210 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "grad_norm": 0.013939457014203072, | |
| "learning_rate": 2.1724461105904406e-05, | |
| "loss": 0.0543, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 21.03, | |
| "grad_norm": 0.0053044771775603294, | |
| "learning_rate": 2.16776007497657e-05, | |
| "loss": 0.0194, | |
| "step": 7230 | |
| }, | |
| { | |
| "epoch": 21.03, | |
| "grad_norm": 73.98345947265625, | |
| "learning_rate": 2.163074039362699e-05, | |
| "loss": 0.0579, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 21.03, | |
| "grad_norm": 0.03918803855776787, | |
| "learning_rate": 2.1583880037488286e-05, | |
| "loss": 0.0161, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 21.03, | |
| "grad_norm": 0.0033793123438954353, | |
| "learning_rate": 2.153701968134958e-05, | |
| "loss": 0.0229, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 21.03, | |
| "eval_accuracy": 0.7822349570200573, | |
| "eval_loss": 1.3766086101531982, | |
| "eval_runtime": 34.2388, | |
| "eval_samples_per_second": 20.386, | |
| "eval_steps_per_second": 1.723, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 0.02534145675599575, | |
| "learning_rate": 2.1490159325210874e-05, | |
| "loss": 0.0177, | |
| "step": 7270 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 27.288881301879883, | |
| "learning_rate": 2.1443298969072163e-05, | |
| "loss": 0.0594, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 3.099937677383423, | |
| "learning_rate": 2.139643861293346e-05, | |
| "loss": 0.07, | |
| "step": 7290 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 0.0036865780130028725, | |
| "learning_rate": 2.134957825679475e-05, | |
| "loss": 0.006, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 59.88230514526367, | |
| "learning_rate": 2.1302717900656047e-05, | |
| "loss": 0.1675, | |
| "step": 7310 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.12187007069587708, | |
| "learning_rate": 2.125585754451734e-05, | |
| "loss": 0.0348, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.011614816263318062, | |
| "learning_rate": 2.120899718837863e-05, | |
| "loss": 0.0227, | |
| "step": 7330 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.013845368288457394, | |
| "learning_rate": 2.1162136832239927e-05, | |
| "loss": 0.0653, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.009227645583450794, | |
| "learning_rate": 2.111527647610122e-05, | |
| "loss": 0.141, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.5920379161834717, | |
| "learning_rate": 2.1068416119962515e-05, | |
| "loss": 0.0401, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.006525806616991758, | |
| "learning_rate": 2.1021555763823804e-05, | |
| "loss": 0.0885, | |
| "step": 7370 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.006711127702146769, | |
| "learning_rate": 2.09746954076851e-05, | |
| "loss": 0.0618, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.004596900660544634, | |
| "learning_rate": 2.0927835051546392e-05, | |
| "loss": 0.0039, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.22587594389915466, | |
| "learning_rate": 2.0880974695407688e-05, | |
| "loss": 0.0897, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.007293607573956251, | |
| "learning_rate": 2.083411433926898e-05, | |
| "loss": 0.1129, | |
| "step": 7410 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.02033209055662155, | |
| "learning_rate": 2.0787253983130272e-05, | |
| "loss": 0.0583, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "grad_norm": 0.006020266562700272, | |
| "learning_rate": 2.0740393626991568e-05, | |
| "loss": 0.0519, | |
| "step": 7430 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 0.050974782556295395, | |
| "learning_rate": 2.069353327085286e-05, | |
| "loss": 0.0972, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 15.67349624633789, | |
| "learning_rate": 2.0646672914714153e-05, | |
| "loss": 0.0047, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 54.079017639160156, | |
| "learning_rate": 2.0599812558575445e-05, | |
| "loss": 0.0765, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 0.28876176476478577, | |
| "learning_rate": 2.055295220243674e-05, | |
| "loss": 0.0463, | |
| "step": 7470 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 0.01849912479519844, | |
| "learning_rate": 2.0506091846298033e-05, | |
| "loss": 0.0064, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 0.014657631516456604, | |
| "learning_rate": 2.0459231490159325e-05, | |
| "loss": 0.0517, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 0.023860439658164978, | |
| "learning_rate": 2.0412371134020618e-05, | |
| "loss": 0.0165, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 0.274640828371048, | |
| "learning_rate": 2.0365510777881913e-05, | |
| "loss": 0.0707, | |
| "step": 7510 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 8.938105583190918, | |
| "learning_rate": 2.0318650421743206e-05, | |
| "loss": 0.0653, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 0.023853203281760216, | |
| "learning_rate": 2.0271790065604498e-05, | |
| "loss": 0.1298, | |
| "step": 7530 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 0.01518749725073576, | |
| "learning_rate": 2.0224929709465794e-05, | |
| "loss": 0.054, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "grad_norm": 0.2764262855052948, | |
| "learning_rate": 2.0178069353327086e-05, | |
| "loss": 0.1653, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 22.03, | |
| "grad_norm": 0.01938408799469471, | |
| "learning_rate": 2.0131208997188382e-05, | |
| "loss": 0.0305, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 22.03, | |
| "grad_norm": 0.06406080722808838, | |
| "learning_rate": 2.008434864104967e-05, | |
| "loss": 0.0038, | |
| "step": 7570 | |
| }, | |
| { | |
| "epoch": 22.03, | |
| "grad_norm": 0.08752384036779404, | |
| "learning_rate": 2.0037488284910966e-05, | |
| "loss": 0.0272, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 22.03, | |
| "grad_norm": 0.00483914278447628, | |
| "learning_rate": 1.999062792877226e-05, | |
| "loss": 0.058, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 22.03, | |
| "eval_accuracy": 0.7893982808022922, | |
| "eval_loss": 1.279589056968689, | |
| "eval_runtime": 34.2577, | |
| "eval_samples_per_second": 20.375, | |
| "eval_steps_per_second": 1.722, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 1.573994755744934, | |
| "learning_rate": 1.9943767572633554e-05, | |
| "loss": 0.0124, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.014693894423544407, | |
| "learning_rate": 1.9896907216494843e-05, | |
| "loss": 0.0532, | |
| "step": 7610 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.06175214797258377, | |
| "learning_rate": 1.985004686035614e-05, | |
| "loss": 0.0272, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.1148892492055893, | |
| "learning_rate": 1.980318650421743e-05, | |
| "loss": 0.0026, | |
| "step": 7630 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.019487930461764336, | |
| "learning_rate": 1.9756326148078727e-05, | |
| "loss": 0.0537, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 34.296722412109375, | |
| "learning_rate": 1.970946579194002e-05, | |
| "loss": 0.0903, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 0.38383758068084717, | |
| "learning_rate": 1.9662605435801312e-05, | |
| "loss": 0.0016, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 7.190293788909912, | |
| "learning_rate": 1.9615745079662607e-05, | |
| "loss": 0.1071, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 0.01017380878329277, | |
| "learning_rate": 1.95688847235239e-05, | |
| "loss": 0.0267, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 18.401382446289062, | |
| "learning_rate": 1.9522024367385195e-05, | |
| "loss": 0.0738, | |
| "step": 7690 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 0.0242378581315279, | |
| "learning_rate": 1.9475164011246484e-05, | |
| "loss": 0.1412, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 0.023785755038261414, | |
| "learning_rate": 1.942830365510778e-05, | |
| "loss": 0.0955, | |
| "step": 7710 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 0.02424936555325985, | |
| "learning_rate": 1.9381443298969072e-05, | |
| "loss": 0.0043, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 0.38176319003105164, | |
| "learning_rate": 1.9334582942830368e-05, | |
| "loss": 0.0326, | |
| "step": 7730 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 0.011367129161953926, | |
| "learning_rate": 1.928772258669166e-05, | |
| "loss": 0.0404, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 0.5925320386886597, | |
| "learning_rate": 1.9240862230552953e-05, | |
| "loss": 0.0306, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "grad_norm": 0.07246335595846176, | |
| "learning_rate": 1.919400187441425e-05, | |
| "loss": 0.069, | |
| "step": 7760 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.29946988821029663, | |
| "learning_rate": 1.914714151827554e-05, | |
| "loss": 0.0996, | |
| "step": 7770 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.035218510776758194, | |
| "learning_rate": 1.9100281162136833e-05, | |
| "loss": 0.0348, | |
| "step": 7780 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.035237617790699005, | |
| "learning_rate": 1.9053420805998125e-05, | |
| "loss": 0.0059, | |
| "step": 7790 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 28.824886322021484, | |
| "learning_rate": 1.900656044985942e-05, | |
| "loss": 0.1736, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.12462026625871658, | |
| "learning_rate": 1.8959700093720713e-05, | |
| "loss": 0.003, | |
| "step": 7810 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.6549420952796936, | |
| "learning_rate": 1.8912839737582006e-05, | |
| "loss": 0.0087, | |
| "step": 7820 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.05357836186885834, | |
| "learning_rate": 1.8865979381443298e-05, | |
| "loss": 0.0214, | |
| "step": 7830 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.006000145338475704, | |
| "learning_rate": 1.8819119025304594e-05, | |
| "loss": 0.0458, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.006179989781230688, | |
| "learning_rate": 1.8772258669165886e-05, | |
| "loss": 0.1173, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.336616575717926, | |
| "learning_rate": 1.872539831302718e-05, | |
| "loss": 0.1676, | |
| "step": 7860 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.0050712330266833305, | |
| "learning_rate": 1.8678537956888474e-05, | |
| "loss": 0.0059, | |
| "step": 7870 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "grad_norm": 0.005236570257693529, | |
| "learning_rate": 1.8631677600749766e-05, | |
| "loss": 0.1685, | |
| "step": 7880 | |
| }, | |
| { | |
| "epoch": 23.03, | |
| "grad_norm": 0.06752946227788925, | |
| "learning_rate": 1.8584817244611062e-05, | |
| "loss": 0.0035, | |
| "step": 7890 | |
| }, | |
| { | |
| "epoch": 23.03, | |
| "grad_norm": 0.4543127119541168, | |
| "learning_rate": 1.853795688847235e-05, | |
| "loss": 0.1258, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 23.03, | |
| "grad_norm": 0.022716745734214783, | |
| "learning_rate": 1.8491096532333647e-05, | |
| "loss": 0.1408, | |
| "step": 7910 | |
| }, | |
| { | |
| "epoch": 23.03, | |
| "grad_norm": 0.02292322926223278, | |
| "learning_rate": 1.844423617619494e-05, | |
| "loss": 0.1045, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 23.03, | |
| "eval_accuracy": 0.7750716332378224, | |
| "eval_loss": 1.3584957122802734, | |
| "eval_runtime": 34.3543, | |
| "eval_samples_per_second": 20.318, | |
| "eval_steps_per_second": 1.717, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.01888859085738659, | |
| "learning_rate": 1.8397375820056235e-05, | |
| "loss": 0.0033, | |
| "step": 7930 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 2.7452802658081055, | |
| "learning_rate": 1.8350515463917524e-05, | |
| "loss": 0.0672, | |
| "step": 7940 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 17.95488166809082, | |
| "learning_rate": 1.830365510777882e-05, | |
| "loss": 0.0579, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 28.065513610839844, | |
| "learning_rate": 1.825679475164011e-05, | |
| "loss": 0.0383, | |
| "step": 7960 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.6805478930473328, | |
| "learning_rate": 1.8209934395501407e-05, | |
| "loss": 0.0037, | |
| "step": 7970 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 0.018596457317471504, | |
| "learning_rate": 1.81630740393627e-05, | |
| "loss": 0.0333, | |
| "step": 7980 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 0.8254786133766174, | |
| "learning_rate": 1.8116213683223992e-05, | |
| "loss": 0.0006, | |
| "step": 7990 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 0.014021596871316433, | |
| "learning_rate": 1.8069353327085288e-05, | |
| "loss": 0.0004, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 0.009306724183261395, | |
| "learning_rate": 1.802249297094658e-05, | |
| "loss": 0.0018, | |
| "step": 8010 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 0.005670236889272928, | |
| "learning_rate": 1.7975632614807876e-05, | |
| "loss": 0.1225, | |
| "step": 8020 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 0.01667461171746254, | |
| "learning_rate": 1.7928772258669165e-05, | |
| "loss": 0.0701, | |
| "step": 8030 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 0.0773952454328537, | |
| "learning_rate": 1.788191190253046e-05, | |
| "loss": 0.0048, | |
| "step": 8040 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 57.535404205322266, | |
| "learning_rate": 1.7835051546391753e-05, | |
| "loss": 0.027, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 0.01095606479793787, | |
| "learning_rate": 1.778819119025305e-05, | |
| "loss": 0.0294, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 0.006405522581189871, | |
| "learning_rate": 1.774133083411434e-05, | |
| "loss": 0.0722, | |
| "step": 8070 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 0.003576676594093442, | |
| "learning_rate": 1.7694470477975633e-05, | |
| "loss": 0.0024, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "grad_norm": 43.07078552246094, | |
| "learning_rate": 1.764761012183693e-05, | |
| "loss": 0.0385, | |
| "step": 8090 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 47.99566650390625, | |
| "learning_rate": 1.760074976569822e-05, | |
| "loss": 0.1164, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 0.029569357633590698, | |
| "learning_rate": 1.7553889409559513e-05, | |
| "loss": 0.0441, | |
| "step": 8110 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 0.020374421030282974, | |
| "learning_rate": 1.7507029053420806e-05, | |
| "loss": 0.0203, | |
| "step": 8120 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 0.2725122570991516, | |
| "learning_rate": 1.74601686972821e-05, | |
| "loss": 0.0015, | |
| "step": 8130 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 25.096572875976562, | |
| "learning_rate": 1.7413308341143394e-05, | |
| "loss": 0.0571, | |
| "step": 8140 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 0.2794142961502075, | |
| "learning_rate": 1.7366447985004686e-05, | |
| "loss": 0.0592, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 0.007156948558986187, | |
| "learning_rate": 1.7319587628865978e-05, | |
| "loss": 0.0116, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 0.06009415537118912, | |
| "learning_rate": 1.7272727272727274e-05, | |
| "loss": 0.0774, | |
| "step": 8170 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 0.0062246439047157764, | |
| "learning_rate": 1.7225866916588566e-05, | |
| "loss": 0.007, | |
| "step": 8180 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 0.0417780727148056, | |
| "learning_rate": 1.717900656044986e-05, | |
| "loss": 0.1236, | |
| "step": 8190 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 53.34578323364258, | |
| "learning_rate": 1.7132146204311154e-05, | |
| "loss": 0.0527, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "grad_norm": 0.20870938897132874, | |
| "learning_rate": 1.7085285848172447e-05, | |
| "loss": 0.0519, | |
| "step": 8210 | |
| }, | |
| { | |
| "epoch": 24.03, | |
| "grad_norm": 0.011034387163817883, | |
| "learning_rate": 1.7038425492033742e-05, | |
| "loss": 0.0005, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 24.03, | |
| "grad_norm": 0.004511510953307152, | |
| "learning_rate": 1.699156513589503e-05, | |
| "loss": 0.0125, | |
| "step": 8230 | |
| }, | |
| { | |
| "epoch": 24.03, | |
| "grad_norm": 0.4468887746334076, | |
| "learning_rate": 1.6944704779756327e-05, | |
| "loss": 0.0015, | |
| "step": 8240 | |
| }, | |
| { | |
| "epoch": 24.03, | |
| "grad_norm": 0.002701952587813139, | |
| "learning_rate": 1.689784442361762e-05, | |
| "loss": 0.0379, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 24.03, | |
| "eval_accuracy": 0.7836676217765043, | |
| "eval_loss": 1.315584421157837, | |
| "eval_runtime": 33.9983, | |
| "eval_samples_per_second": 20.53, | |
| "eval_steps_per_second": 1.735, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.0075071449391543865, | |
| "learning_rate": 1.6850984067478915e-05, | |
| "loss": 0.0626, | |
| "step": 8260 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.23369351029396057, | |
| "learning_rate": 1.6804123711340207e-05, | |
| "loss": 0.0197, | |
| "step": 8270 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 11.089585304260254, | |
| "learning_rate": 1.67572633552015e-05, | |
| "loss": 0.0019, | |
| "step": 8280 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.0038492237217724323, | |
| "learning_rate": 1.6710402999062792e-05, | |
| "loss": 0.0002, | |
| "step": 8290 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.003863664111122489, | |
| "learning_rate": 1.6663542642924088e-05, | |
| "loss": 0.0012, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 0.013411330990493298, | |
| "learning_rate": 1.661668228678538e-05, | |
| "loss": 0.0003, | |
| "step": 8310 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 0.028628764674067497, | |
| "learning_rate": 1.6569821930646672e-05, | |
| "loss": 0.0005, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 0.010359793901443481, | |
| "learning_rate": 1.6522961574507968e-05, | |
| "loss": 0.0662, | |
| "step": 8330 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 4.15015983581543, | |
| "learning_rate": 1.647610121836926e-05, | |
| "loss": 0.0023, | |
| "step": 8340 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 34.17851257324219, | |
| "learning_rate": 1.6429240862230556e-05, | |
| "loss": 0.0411, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 0.0031898904126137495, | |
| "learning_rate": 1.6382380506091845e-05, | |
| "loss": 0.0003, | |
| "step": 8360 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 0.002583961235359311, | |
| "learning_rate": 1.633552014995314e-05, | |
| "loss": 0.0485, | |
| "step": 8370 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 18.96380615234375, | |
| "learning_rate": 1.6288659793814433e-05, | |
| "loss": 0.1036, | |
| "step": 8380 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 45.696075439453125, | |
| "learning_rate": 1.624179943767573e-05, | |
| "loss": 0.0844, | |
| "step": 8390 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 0.00760071724653244, | |
| "learning_rate": 1.619493908153702e-05, | |
| "loss": 0.0098, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 0.019091518595814705, | |
| "learning_rate": 1.6148078725398313e-05, | |
| "loss": 0.0273, | |
| "step": 8410 | |
| }, | |
| { | |
| "epoch": 25.01, | |
| "grad_norm": 0.004600458778440952, | |
| "learning_rate": 1.610121836925961e-05, | |
| "loss": 0.0008, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 22.48053741455078, | |
| "learning_rate": 1.60543580131209e-05, | |
| "loss": 0.0469, | |
| "step": 8430 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 0.011721034534275532, | |
| "learning_rate": 1.6007497656982194e-05, | |
| "loss": 0.046, | |
| "step": 8440 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 0.004451930057257414, | |
| "learning_rate": 1.5960637300843486e-05, | |
| "loss": 0.0038, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 0.010986747220158577, | |
| "learning_rate": 1.591377694470478e-05, | |
| "loss": 0.125, | |
| "step": 8460 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 41.15037155151367, | |
| "learning_rate": 1.5866916588566074e-05, | |
| "loss": 0.0238, | |
| "step": 8470 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 0.006322337780147791, | |
| "learning_rate": 1.5820056232427366e-05, | |
| "loss": 0.0012, | |
| "step": 8480 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 0.007438218221068382, | |
| "learning_rate": 1.577319587628866e-05, | |
| "loss": 0.0558, | |
| "step": 8490 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 0.007493423763662577, | |
| "learning_rate": 1.5726335520149954e-05, | |
| "loss": 0.0244, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 0.005765008274465799, | |
| "learning_rate": 1.5679475164011247e-05, | |
| "loss": 0.0097, | |
| "step": 8510 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 0.3964434266090393, | |
| "learning_rate": 1.563261480787254e-05, | |
| "loss": 0.0018, | |
| "step": 8520 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 2.117600440979004, | |
| "learning_rate": 1.5585754451733835e-05, | |
| "loss": 0.0519, | |
| "step": 8530 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "grad_norm": 0.10996006429195404, | |
| "learning_rate": 1.5538894095595127e-05, | |
| "loss": 0.0021, | |
| "step": 8540 | |
| }, | |
| { | |
| "epoch": 25.03, | |
| "grad_norm": 0.0033693662844598293, | |
| "learning_rate": 1.5492033739456423e-05, | |
| "loss": 0.0811, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 25.03, | |
| "grad_norm": 10.247265815734863, | |
| "learning_rate": 1.544517338331771e-05, | |
| "loss": 0.0058, | |
| "step": 8560 | |
| }, | |
| { | |
| "epoch": 25.03, | |
| "grad_norm": 0.2538563013076782, | |
| "learning_rate": 1.5398313027179007e-05, | |
| "loss": 0.03, | |
| "step": 8570 | |
| }, | |
| { | |
| "epoch": 25.03, | |
| "grad_norm": 0.00820246897637844, | |
| "learning_rate": 1.53514526710403e-05, | |
| "loss": 0.0945, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 25.03, | |
| "eval_accuracy": 0.7922636103151862, | |
| "eval_loss": 1.3046531677246094, | |
| "eval_runtime": 33.5684, | |
| "eval_samples_per_second": 20.793, | |
| "eval_steps_per_second": 1.758, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 0.0038312424439936876, | |
| "learning_rate": 1.5304592314901595e-05, | |
| "loss": 0.0008, | |
| "step": 8590 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 0.06059965118765831, | |
| "learning_rate": 1.525773195876289e-05, | |
| "loss": 0.0115, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 0.004089924972504377, | |
| "learning_rate": 1.521087160262418e-05, | |
| "loss": 0.0002, | |
| "step": 8610 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 0.0043753357604146, | |
| "learning_rate": 1.5164011246485474e-05, | |
| "loss": 0.0727, | |
| "step": 8620 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 5.026497840881348, | |
| "learning_rate": 1.5117150890346768e-05, | |
| "loss": 0.0639, | |
| "step": 8630 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.014356585219502449, | |
| "learning_rate": 1.5070290534208062e-05, | |
| "loss": 0.036, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.24706010520458221, | |
| "learning_rate": 1.5023430178069353e-05, | |
| "loss": 0.001, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.06222040206193924, | |
| "learning_rate": 1.4976569821930647e-05, | |
| "loss": 0.0436, | |
| "step": 8660 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.0071958452463150024, | |
| "learning_rate": 1.492970946579194e-05, | |
| "loss": 0.0006, | |
| "step": 8670 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.011005638167262077, | |
| "learning_rate": 1.4882849109653235e-05, | |
| "loss": 0.0528, | |
| "step": 8680 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.09897544980049133, | |
| "learning_rate": 1.4835988753514527e-05, | |
| "loss": 0.0099, | |
| "step": 8690 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 1.022418737411499, | |
| "learning_rate": 1.4789128397375821e-05, | |
| "loss": 0.006, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.0073012434877455235, | |
| "learning_rate": 1.4742268041237115e-05, | |
| "loss": 0.0008, | |
| "step": 8710 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.07601054012775421, | |
| "learning_rate": 1.4695407685098409e-05, | |
| "loss": 0.0256, | |
| "step": 8720 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.010327538475394249, | |
| "learning_rate": 1.46485473289597e-05, | |
| "loss": 0.0105, | |
| "step": 8730 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.4567221999168396, | |
| "learning_rate": 1.4601686972820994e-05, | |
| "loss": 0.0338, | |
| "step": 8740 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "grad_norm": 0.03632340207695961, | |
| "learning_rate": 1.4554826616682288e-05, | |
| "loss": 0.0004, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.005994404200464487, | |
| "learning_rate": 1.4507966260543582e-05, | |
| "loss": 0.0004, | |
| "step": 8760 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.004464009311050177, | |
| "learning_rate": 1.4461105904404872e-05, | |
| "loss": 0.0654, | |
| "step": 8770 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.5976660251617432, | |
| "learning_rate": 1.4414245548266168e-05, | |
| "loss": 0.0374, | |
| "step": 8780 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.003634576452895999, | |
| "learning_rate": 1.4367385192127462e-05, | |
| "loss": 0.0921, | |
| "step": 8790 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.005951224360615015, | |
| "learning_rate": 1.4320524835988756e-05, | |
| "loss": 0.0096, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.0037890600506216288, | |
| "learning_rate": 1.4273664479850047e-05, | |
| "loss": 0.0005, | |
| "step": 8810 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.00492624519392848, | |
| "learning_rate": 1.422680412371134e-05, | |
| "loss": 0.0106, | |
| "step": 8820 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.0032419913914054632, | |
| "learning_rate": 1.4179943767572635e-05, | |
| "loss": 0.0068, | |
| "step": 8830 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.006293057929724455, | |
| "learning_rate": 1.4133083411433929e-05, | |
| "loss": 0.0646, | |
| "step": 8840 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.00998301524668932, | |
| "learning_rate": 1.4086223055295219e-05, | |
| "loss": 0.0163, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.08308320492506027, | |
| "learning_rate": 1.4039362699156513e-05, | |
| "loss": 0.0003, | |
| "step": 8860 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "grad_norm": 0.0030508143827319145, | |
| "learning_rate": 1.3992502343017807e-05, | |
| "loss": 0.064, | |
| "step": 8870 | |
| }, | |
| { | |
| "epoch": 26.03, | |
| "grad_norm": 0.3473857045173645, | |
| "learning_rate": 1.3945641986879101e-05, | |
| "loss": 0.0006, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 26.03, | |
| "grad_norm": 0.002809175057336688, | |
| "learning_rate": 1.3898781630740395e-05, | |
| "loss": 0.0004, | |
| "step": 8890 | |
| }, | |
| { | |
| "epoch": 26.03, | |
| "grad_norm": 1.4648064374923706, | |
| "learning_rate": 1.3851921274601688e-05, | |
| "loss": 0.0478, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 26.03, | |
| "grad_norm": 0.008449913933873177, | |
| "learning_rate": 1.3805060918462982e-05, | |
| "loss": 0.0528, | |
| "step": 8910 | |
| }, | |
| { | |
| "epoch": 26.03, | |
| "eval_accuracy": 0.7893982808022922, | |
| "eval_loss": 1.3669614791870117, | |
| "eval_runtime": 34.4841, | |
| "eval_samples_per_second": 20.241, | |
| "eval_steps_per_second": 1.711, | |
| "step": 8910 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.009650280699133873, | |
| "learning_rate": 1.3758200562324276e-05, | |
| "loss": 0.0131, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.00797184742987156, | |
| "learning_rate": 1.371134020618557e-05, | |
| "loss": 0.0228, | |
| "step": 8930 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.004884254653006792, | |
| "learning_rate": 1.366447985004686e-05, | |
| "loss": 0.0002, | |
| "step": 8940 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.0034428162034600973, | |
| "learning_rate": 1.3617619493908154e-05, | |
| "loss": 0.0017, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.004850251600146294, | |
| "learning_rate": 1.3570759137769448e-05, | |
| "loss": 0.0611, | |
| "step": 8960 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.028377506881952286, | |
| "learning_rate": 1.3523898781630742e-05, | |
| "loss": 0.0002, | |
| "step": 8970 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.0027832870837301016, | |
| "learning_rate": 1.3477038425492033e-05, | |
| "loss": 0.0475, | |
| "step": 8980 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.11803574860095978, | |
| "learning_rate": 1.3430178069353327e-05, | |
| "loss": 0.0005, | |
| "step": 8990 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.23726528882980347, | |
| "learning_rate": 1.338331771321462e-05, | |
| "loss": 0.0003, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.00947210006415844, | |
| "learning_rate": 1.3336457357075915e-05, | |
| "loss": 0.0004, | |
| "step": 9010 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.006413722410798073, | |
| "learning_rate": 1.3289597000937207e-05, | |
| "loss": 0.0002, | |
| "step": 9020 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.0032037473283708096, | |
| "learning_rate": 1.3242736644798501e-05, | |
| "loss": 0.0022, | |
| "step": 9030 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.0032357927411794662, | |
| "learning_rate": 1.3195876288659795e-05, | |
| "loss": 0.0336, | |
| "step": 9040 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 1.7443645000457764, | |
| "learning_rate": 1.314901593252109e-05, | |
| "loss": 0.0008, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.002807668410241604, | |
| "learning_rate": 1.310215557638238e-05, | |
| "loss": 0.0002, | |
| "step": 9060 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.0025303384754806757, | |
| "learning_rate": 1.3055295220243674e-05, | |
| "loss": 0.0005, | |
| "step": 9070 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "grad_norm": 0.37589800357818604, | |
| "learning_rate": 1.3008434864104968e-05, | |
| "loss": 0.0713, | |
| "step": 9080 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.002870983211323619, | |
| "learning_rate": 1.2961574507966262e-05, | |
| "loss": 0.0204, | |
| "step": 9090 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.017298812046647072, | |
| "learning_rate": 1.2914714151827554e-05, | |
| "loss": 0.0268, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.00800339411944151, | |
| "learning_rate": 1.2867853795688848e-05, | |
| "loss": 0.1058, | |
| "step": 9110 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.0037049567326903343, | |
| "learning_rate": 1.2820993439550142e-05, | |
| "loss": 0.0002, | |
| "step": 9120 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.0456349216401577, | |
| "learning_rate": 1.2774133083411436e-05, | |
| "loss": 0.013, | |
| "step": 9130 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.003435475053265691, | |
| "learning_rate": 1.2727272727272727e-05, | |
| "loss": 0.0392, | |
| "step": 9140 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.0036417213268578053, | |
| "learning_rate": 1.268041237113402e-05, | |
| "loss": 0.0003, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.01712891273200512, | |
| "learning_rate": 1.2633552014995315e-05, | |
| "loss": 0.0151, | |
| "step": 9160 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.22050367295742035, | |
| "learning_rate": 1.2586691658856609e-05, | |
| "loss": 0.0004, | |
| "step": 9170 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.07363563030958176, | |
| "learning_rate": 1.2539831302717903e-05, | |
| "loss": 0.0084, | |
| "step": 9180 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.0359061174094677, | |
| "learning_rate": 1.2492970946579195e-05, | |
| "loss": 0.0008, | |
| "step": 9190 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "grad_norm": 0.007665242068469524, | |
| "learning_rate": 1.2446110590440487e-05, | |
| "loss": 0.0623, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 27.03, | |
| "grad_norm": 0.0135026965290308, | |
| "learning_rate": 1.2399250234301781e-05, | |
| "loss": 0.011, | |
| "step": 9210 | |
| }, | |
| { | |
| "epoch": 27.03, | |
| "grad_norm": 3.0960659980773926, | |
| "learning_rate": 1.2352389878163074e-05, | |
| "loss": 0.05, | |
| "step": 9220 | |
| }, | |
| { | |
| "epoch": 27.03, | |
| "grad_norm": 0.01721220277249813, | |
| "learning_rate": 1.2305529522024368e-05, | |
| "loss": 0.0707, | |
| "step": 9230 | |
| }, | |
| { | |
| "epoch": 27.03, | |
| "grad_norm": 0.004766266793012619, | |
| "learning_rate": 1.2258669165885662e-05, | |
| "loss": 0.0002, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 27.03, | |
| "eval_accuracy": 0.7965616045845272, | |
| "eval_loss": 1.3398090600967407, | |
| "eval_runtime": 33.8835, | |
| "eval_samples_per_second": 20.6, | |
| "eval_steps_per_second": 1.741, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.0029590607155114412, | |
| "learning_rate": 1.2211808809746956e-05, | |
| "loss": 0.0831, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.0406717024743557, | |
| "learning_rate": 1.2164948453608248e-05, | |
| "loss": 0.0004, | |
| "step": 9260 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.03818991780281067, | |
| "learning_rate": 1.2118088097469542e-05, | |
| "loss": 0.0003, | |
| "step": 9270 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.007692787330597639, | |
| "learning_rate": 1.2071227741330834e-05, | |
| "loss": 0.0096, | |
| "step": 9280 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.08763420581817627, | |
| "learning_rate": 1.2024367385192128e-05, | |
| "loss": 0.0663, | |
| "step": 9290 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 0.00407880125567317, | |
| "learning_rate": 1.197750702905342e-05, | |
| "loss": 0.0031, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 0.006175840273499489, | |
| "learning_rate": 1.1930646672914715e-05, | |
| "loss": 0.0865, | |
| "step": 9310 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 6.280770301818848, | |
| "learning_rate": 1.1883786316776007e-05, | |
| "loss": 0.0024, | |
| "step": 9320 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 12.87985610961914, | |
| "learning_rate": 1.1836925960637301e-05, | |
| "loss": 0.0468, | |
| "step": 9330 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 0.04478368163108826, | |
| "learning_rate": 1.1790065604498595e-05, | |
| "loss": 0.0002, | |
| "step": 9340 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 0.005779411643743515, | |
| "learning_rate": 1.1743205248359889e-05, | |
| "loss": 0.0002, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 0.0021354747004806995, | |
| "learning_rate": 1.1696344892221181e-05, | |
| "loss": 0.0002, | |
| "step": 9360 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 5.800276756286621, | |
| "learning_rate": 1.1649484536082475e-05, | |
| "loss": 0.01, | |
| "step": 9370 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 0.0021325971465557814, | |
| "learning_rate": 1.1602624179943768e-05, | |
| "loss": 0.0133, | |
| "step": 9380 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 0.003749684663489461, | |
| "learning_rate": 1.1555763823805062e-05, | |
| "loss": 0.0003, | |
| "step": 9390 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 0.006699393503367901, | |
| "learning_rate": 1.1508903467666354e-05, | |
| "loss": 0.0058, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 28.01, | |
| "grad_norm": 0.0032480955123901367, | |
| "learning_rate": 1.1462043111527648e-05, | |
| "loss": 0.0184, | |
| "step": 9410 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.0017323597567155957, | |
| "learning_rate": 1.141518275538894e-05, | |
| "loss": 0.0002, | |
| "step": 9420 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.0023274635896086693, | |
| "learning_rate": 1.1368322399250234e-05, | |
| "loss": 0.0002, | |
| "step": 9430 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.003413414815440774, | |
| "learning_rate": 1.1321462043111528e-05, | |
| "loss": 0.0014, | |
| "step": 9440 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.01558777131140232, | |
| "learning_rate": 1.1274601686972822e-05, | |
| "loss": 0.042, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.003739135107025504, | |
| "learning_rate": 1.1227741330834115e-05, | |
| "loss": 0.0136, | |
| "step": 9460 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.043951474130153656, | |
| "learning_rate": 1.1180880974695409e-05, | |
| "loss": 0.0804, | |
| "step": 9470 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.0028257304802536964, | |
| "learning_rate": 1.1134020618556703e-05, | |
| "loss": 0.0002, | |
| "step": 9480 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.0037733283825218678, | |
| "learning_rate": 1.1087160262417995e-05, | |
| "loss": 0.0004, | |
| "step": 9490 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.056008391082286835, | |
| "learning_rate": 1.1040299906279289e-05, | |
| "loss": 0.0319, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.0031692145857959986, | |
| "learning_rate": 1.0993439550140581e-05, | |
| "loss": 0.0685, | |
| "step": 9510 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.004492649342864752, | |
| "learning_rate": 1.0946579194001875e-05, | |
| "loss": 0.0137, | |
| "step": 9520 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "grad_norm": 0.0059276544488966465, | |
| "learning_rate": 1.0899718837863168e-05, | |
| "loss": 0.0007, | |
| "step": 9530 | |
| }, | |
| { | |
| "epoch": 28.03, | |
| "grad_norm": 0.004752593580633402, | |
| "learning_rate": 1.0852858481724462e-05, | |
| "loss": 0.0193, | |
| "step": 9540 | |
| }, | |
| { | |
| "epoch": 28.03, | |
| "grad_norm": 0.006295239552855492, | |
| "learning_rate": 1.0805998125585754e-05, | |
| "loss": 0.0037, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 28.03, | |
| "grad_norm": 41.76048278808594, | |
| "learning_rate": 1.0759137769447048e-05, | |
| "loss": 0.0485, | |
| "step": 9560 | |
| }, | |
| { | |
| "epoch": 28.03, | |
| "grad_norm": 0.0014590555801987648, | |
| "learning_rate": 1.0712277413308342e-05, | |
| "loss": 0.0562, | |
| "step": 9570 | |
| }, | |
| { | |
| "epoch": 28.03, | |
| "eval_accuracy": 0.7979942693409742, | |
| "eval_loss": 1.3444451093673706, | |
| "eval_runtime": 34.4815, | |
| "eval_samples_per_second": 20.243, | |
| "eval_steps_per_second": 1.711, | |
| "step": 9570 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 0.00453572254627943, | |
| "learning_rate": 1.0665417057169636e-05, | |
| "loss": 0.0002, | |
| "step": 9580 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 0.04278896749019623, | |
| "learning_rate": 1.0618556701030928e-05, | |
| "loss": 0.0332, | |
| "step": 9590 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 0.0018639545887708664, | |
| "learning_rate": 1.0571696344892222e-05, | |
| "loss": 0.0002, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 0.005231024231761694, | |
| "learning_rate": 1.0524835988753515e-05, | |
| "loss": 0.0009, | |
| "step": 9610 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 0.0017461972311139107, | |
| "learning_rate": 1.0477975632614809e-05, | |
| "loss": 0.0001, | |
| "step": 9620 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 7.924661636352539, | |
| "learning_rate": 1.0431115276476101e-05, | |
| "loss": 0.0308, | |
| "step": 9630 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 0.012182795442640781, | |
| "learning_rate": 1.0384254920337395e-05, | |
| "loss": 0.0043, | |
| "step": 9640 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 0.0014665969647467136, | |
| "learning_rate": 1.0337394564198687e-05, | |
| "loss": 0.0002, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 0.0027665491215884686, | |
| "learning_rate": 1.0290534208059981e-05, | |
| "loss": 0.0097, | |
| "step": 9660 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 0.0026171233039349318, | |
| "learning_rate": 1.0243673851921275e-05, | |
| "loss": 0.0543, | |
| "step": 9670 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 0.005574346520006657, | |
| "learning_rate": 1.019681349578257e-05, | |
| "loss": 0.0001, | |
| "step": 9680 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 0.07008225470781326, | |
| "learning_rate": 1.0149953139643862e-05, | |
| "loss": 0.0321, | |
| "step": 9690 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 58.78215408325195, | |
| "learning_rate": 1.0103092783505156e-05, | |
| "loss": 0.106, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 0.010860403068363667, | |
| "learning_rate": 1.0056232427366448e-05, | |
| "loss": 0.0625, | |
| "step": 9710 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 0.09613881260156631, | |
| "learning_rate": 1.0009372071227742e-05, | |
| "loss": 0.007, | |
| "step": 9720 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 0.01900539919734001, | |
| "learning_rate": 9.962511715089034e-06, | |
| "loss": 0.0109, | |
| "step": 9730 | |
| }, | |
| { | |
| "epoch": 29.01, | |
| "grad_norm": 0.009411687031388283, | |
| "learning_rate": 9.915651358950328e-06, | |
| "loss": 0.0018, | |
| "step": 9740 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 16.337764739990234, | |
| "learning_rate": 9.86879100281162e-06, | |
| "loss": 0.0593, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.028839366510510445, | |
| "learning_rate": 9.821930646672915e-06, | |
| "loss": 0.0604, | |
| "step": 9760 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.017779843881726265, | |
| "learning_rate": 9.775070290534209e-06, | |
| "loss": 0.0365, | |
| "step": 9770 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.003166783368214965, | |
| "learning_rate": 9.728209934395503e-06, | |
| "loss": 0.0053, | |
| "step": 9780 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.0021175253205001354, | |
| "learning_rate": 9.681349578256797e-06, | |
| "loss": 0.0001, | |
| "step": 9790 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.20540370047092438, | |
| "learning_rate": 9.634489222118089e-06, | |
| "loss": 0.0003, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.002469886327162385, | |
| "learning_rate": 9.587628865979383e-06, | |
| "loss": 0.0002, | |
| "step": 9810 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.012953460216522217, | |
| "learning_rate": 9.540768509840675e-06, | |
| "loss": 0.0271, | |
| "step": 9820 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.02899201773107052, | |
| "learning_rate": 9.49390815370197e-06, | |
| "loss": 0.0414, | |
| "step": 9830 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.0029837109614163637, | |
| "learning_rate": 9.447047797563262e-06, | |
| "loss": 0.0002, | |
| "step": 9840 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 61.894805908203125, | |
| "learning_rate": 9.400187441424556e-06, | |
| "loss": 0.0608, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.003072848543524742, | |
| "learning_rate": 9.353327085285848e-06, | |
| "loss": 0.0014, | |
| "step": 9860 | |
| }, | |
| { | |
| "epoch": 29.03, | |
| "grad_norm": 0.0036488294135779142, | |
| "learning_rate": 9.306466729147142e-06, | |
| "loss": 0.001, | |
| "step": 9870 | |
| }, | |
| { | |
| "epoch": 29.03, | |
| "grad_norm": 0.002994240028783679, | |
| "learning_rate": 9.259606373008434e-06, | |
| "loss": 0.0007, | |
| "step": 9880 | |
| }, | |
| { | |
| "epoch": 29.03, | |
| "grad_norm": 0.01359494123607874, | |
| "learning_rate": 9.212746016869728e-06, | |
| "loss": 0.0007, | |
| "step": 9890 | |
| }, | |
| { | |
| "epoch": 29.03, | |
| "grad_norm": 0.0029486478306353092, | |
| "learning_rate": 9.165885660731022e-06, | |
| "loss": 0.0002, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 29.03, | |
| "eval_accuracy": 0.8166189111747851, | |
| "eval_loss": 1.2678812742233276, | |
| "eval_runtime": 33.8703, | |
| "eval_samples_per_second": 20.608, | |
| "eval_steps_per_second": 1.742, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.0020161494612693787, | |
| "learning_rate": 9.119025304592316e-06, | |
| "loss": 0.0002, | |
| "step": 9910 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.001892567495815456, | |
| "learning_rate": 9.072164948453609e-06, | |
| "loss": 0.0069, | |
| "step": 9920 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.003333768341690302, | |
| "learning_rate": 9.025304592314903e-06, | |
| "loss": 0.0017, | |
| "step": 9930 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 5.158819675445557, | |
| "learning_rate": 8.978444236176195e-06, | |
| "loss": 0.0472, | |
| "step": 9940 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.010453186929225922, | |
| "learning_rate": 8.931583880037489e-06, | |
| "loss": 0.0017, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.006460112985223532, | |
| "learning_rate": 8.884723523898781e-06, | |
| "loss": 0.0028, | |
| "step": 9960 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 1.868633508682251, | |
| "learning_rate": 8.837863167760075e-06, | |
| "loss": 0.0004, | |
| "step": 9970 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.0062409755773842335, | |
| "learning_rate": 8.791002811621368e-06, | |
| "loss": 0.0182, | |
| "step": 9980 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.002611867617815733, | |
| "learning_rate": 8.744142455482662e-06, | |
| "loss": 0.0002, | |
| "step": 9990 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.003704243106767535, | |
| "learning_rate": 8.697282099343956e-06, | |
| "loss": 0.0001, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.0019702170975506306, | |
| "learning_rate": 8.65042174320525e-06, | |
| "loss": 0.0574, | |
| "step": 10010 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.006285363808274269, | |
| "learning_rate": 8.603561387066542e-06, | |
| "loss": 0.0001, | |
| "step": 10020 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.0028792780358344316, | |
| "learning_rate": 8.556701030927836e-06, | |
| "loss": 0.0059, | |
| "step": 10030 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.0020247281063348055, | |
| "learning_rate": 8.509840674789128e-06, | |
| "loss": 0.0005, | |
| "step": 10040 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.0075753917917609215, | |
| "learning_rate": 8.462980318650422e-06, | |
| "loss": 0.0039, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.0014275303110480309, | |
| "learning_rate": 8.416119962511715e-06, | |
| "loss": 0.0304, | |
| "step": 10060 | |
| }, | |
| { | |
| "epoch": 30.01, | |
| "grad_norm": 0.0023228314239531755, | |
| "learning_rate": 8.369259606373009e-06, | |
| "loss": 0.0002, | |
| "step": 10070 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.2118721902370453, | |
| "learning_rate": 8.322399250234301e-06, | |
| "loss": 0.0763, | |
| "step": 10080 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.12885122001171112, | |
| "learning_rate": 8.275538894095595e-06, | |
| "loss": 0.0003, | |
| "step": 10090 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.01231481321156025, | |
| "learning_rate": 8.228678537956889e-06, | |
| "loss": 0.0004, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.003690751502290368, | |
| "learning_rate": 8.181818181818183e-06, | |
| "loss": 0.0019, | |
| "step": 10110 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.0029746764339506626, | |
| "learning_rate": 8.134957825679477e-06, | |
| "loss": 0.058, | |
| "step": 10120 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.003089478937909007, | |
| "learning_rate": 8.08809746954077e-06, | |
| "loss": 0.0789, | |
| "step": 10130 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.0020090111065655947, | |
| "learning_rate": 8.041237113402063e-06, | |
| "loss": 0.0001, | |
| "step": 10140 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.23149579763412476, | |
| "learning_rate": 7.994376757263356e-06, | |
| "loss": 0.0285, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.006695437245070934, | |
| "learning_rate": 7.94751640112465e-06, | |
| "loss": 0.0811, | |
| "step": 10160 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.03278821334242821, | |
| "learning_rate": 7.900656044985942e-06, | |
| "loss": 0.0118, | |
| "step": 10170 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.002686940599232912, | |
| "learning_rate": 7.853795688847236e-06, | |
| "loss": 0.0013, | |
| "step": 10180 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "grad_norm": 0.004242202267050743, | |
| "learning_rate": 7.806935332708528e-06, | |
| "loss": 0.0229, | |
| "step": 10190 | |
| }, | |
| { | |
| "epoch": 30.03, | |
| "grad_norm": 0.12063409388065338, | |
| "learning_rate": 7.760074976569822e-06, | |
| "loss": 0.0447, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 30.03, | |
| "grad_norm": 0.0015561155742034316, | |
| "learning_rate": 7.713214620431115e-06, | |
| "loss": 0.0445, | |
| "step": 10210 | |
| }, | |
| { | |
| "epoch": 30.03, | |
| "grad_norm": 0.003858069656416774, | |
| "learning_rate": 7.666354264292409e-06, | |
| "loss": 0.0108, | |
| "step": 10220 | |
| }, | |
| { | |
| "epoch": 30.03, | |
| "grad_norm": 0.001974466722458601, | |
| "learning_rate": 7.619493908153702e-06, | |
| "loss": 0.0018, | |
| "step": 10230 | |
| }, | |
| { | |
| "epoch": 30.03, | |
| "eval_accuracy": 0.7965616045845272, | |
| "eval_loss": 1.3747639656066895, | |
| "eval_runtime": 33.9072, | |
| "eval_samples_per_second": 20.586, | |
| "eval_steps_per_second": 1.74, | |
| "step": 10230 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 0.0030528667848557234, | |
| "learning_rate": 7.572633552014996e-06, | |
| "loss": 0.0006, | |
| "step": 10240 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 0.07026159763336182, | |
| "learning_rate": 7.525773195876289e-06, | |
| "loss": 0.0002, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 0.013108909130096436, | |
| "learning_rate": 7.478912839737583e-06, | |
| "loss": 0.0083, | |
| "step": 10260 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 0.003818312892690301, | |
| "learning_rate": 7.432052483598875e-06, | |
| "loss": 0.0001, | |
| "step": 10270 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 53.924110412597656, | |
| "learning_rate": 7.385192127460169e-06, | |
| "loss": 0.0049, | |
| "step": 10280 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 6.479648590087891, | |
| "learning_rate": 7.3383317713214616e-06, | |
| "loss": 0.066, | |
| "step": 10290 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.005377662368118763, | |
| "learning_rate": 7.2914714151827556e-06, | |
| "loss": 0.0042, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.0028942872304469347, | |
| "learning_rate": 7.244611059044049e-06, | |
| "loss": 0.0006, | |
| "step": 10310 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.0026112585328519344, | |
| "learning_rate": 7.197750702905343e-06, | |
| "loss": 0.0074, | |
| "step": 10320 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.00254084006883204, | |
| "learning_rate": 7.150890346766635e-06, | |
| "loss": 0.0001, | |
| "step": 10330 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.027123264968395233, | |
| "learning_rate": 7.104029990627929e-06, | |
| "loss": 0.0002, | |
| "step": 10340 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.0028091182466596365, | |
| "learning_rate": 7.057169634489222e-06, | |
| "loss": 0.0005, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.0014785215025767684, | |
| "learning_rate": 7.010309278350516e-06, | |
| "loss": 0.0001, | |
| "step": 10360 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.002326256362721324, | |
| "learning_rate": 6.9634489222118085e-06, | |
| "loss": 0.0312, | |
| "step": 10370 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.020408490672707558, | |
| "learning_rate": 6.9165885660731026e-06, | |
| "loss": 0.0289, | |
| "step": 10380 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.0020945239812135696, | |
| "learning_rate": 6.869728209934395e-06, | |
| "loss": 0.0003, | |
| "step": 10390 | |
| }, | |
| { | |
| "epoch": 31.01, | |
| "grad_norm": 0.050324421375989914, | |
| "learning_rate": 6.822867853795689e-06, | |
| "loss": 0.0001, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 0.09508836269378662, | |
| "learning_rate": 6.776007497656983e-06, | |
| "loss": 0.0003, | |
| "step": 10410 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 0.0019429631065577269, | |
| "learning_rate": 6.729147141518276e-06, | |
| "loss": 0.0076, | |
| "step": 10420 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 0.0016007090453058481, | |
| "learning_rate": 6.68228678537957e-06, | |
| "loss": 0.0132, | |
| "step": 10430 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 4.64786958694458, | |
| "learning_rate": 6.635426429240862e-06, | |
| "loss": 0.0006, | |
| "step": 10440 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 63.0573616027832, | |
| "learning_rate": 6.588566073102156e-06, | |
| "loss": 0.0605, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 0.001429658499546349, | |
| "learning_rate": 6.541705716963449e-06, | |
| "loss": 0.0001, | |
| "step": 10460 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 0.012436199001967907, | |
| "learning_rate": 6.494845360824743e-06, | |
| "loss": 0.0055, | |
| "step": 10470 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 0.002089001704007387, | |
| "learning_rate": 6.447985004686036e-06, | |
| "loss": 0.0004, | |
| "step": 10480 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 0.006997866556048393, | |
| "learning_rate": 6.40112464854733e-06, | |
| "loss": 0.1141, | |
| "step": 10490 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 0.018327118828892708, | |
| "learning_rate": 6.354264292408622e-06, | |
| "loss": 0.0071, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 0.30546697974205017, | |
| "learning_rate": 6.307403936269916e-06, | |
| "loss": 0.0002, | |
| "step": 10510 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "grad_norm": 64.27317810058594, | |
| "learning_rate": 6.260543580131209e-06, | |
| "loss": 0.0119, | |
| "step": 10520 | |
| }, | |
| { | |
| "epoch": 31.03, | |
| "grad_norm": 0.0014340607449412346, | |
| "learning_rate": 6.2136832239925025e-06, | |
| "loss": 0.0002, | |
| "step": 10530 | |
| }, | |
| { | |
| "epoch": 31.03, | |
| "grad_norm": 0.012228988111019135, | |
| "learning_rate": 6.166822867853796e-06, | |
| "loss": 0.0011, | |
| "step": 10540 | |
| }, | |
| { | |
| "epoch": 31.03, | |
| "grad_norm": 0.0018130154348909855, | |
| "learning_rate": 6.11996251171509e-06, | |
| "loss": 0.0012, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 31.03, | |
| "grad_norm": 0.0017277301521971822, | |
| "learning_rate": 6.073102155576383e-06, | |
| "loss": 0.0371, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 31.03, | |
| "eval_accuracy": 0.8080229226361032, | |
| "eval_loss": 1.40940523147583, | |
| "eval_runtime": 33.9429, | |
| "eval_samples_per_second": 20.564, | |
| "eval_steps_per_second": 1.738, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.0025587286800146103, | |
| "learning_rate": 6.026241799437676e-06, | |
| "loss": 0.0003, | |
| "step": 10570 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.002236501080915332, | |
| "learning_rate": 5.97938144329897e-06, | |
| "loss": 0.009, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.00986202247440815, | |
| "learning_rate": 5.932521087160263e-06, | |
| "loss": 0.0003, | |
| "step": 10590 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.001730900607071817, | |
| "learning_rate": 5.885660731021556e-06, | |
| "loss": 0.0176, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.004606322385370731, | |
| "learning_rate": 5.8388003748828495e-06, | |
| "loss": 0.0001, | |
| "step": 10610 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.00272651226259768, | |
| "learning_rate": 5.791940018744143e-06, | |
| "loss": 0.0002, | |
| "step": 10620 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.0064768255688250065, | |
| "learning_rate": 5.745079662605436e-06, | |
| "loss": 0.0012, | |
| "step": 10630 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.002017725259065628, | |
| "learning_rate": 5.69821930646673e-06, | |
| "loss": 0.0001, | |
| "step": 10640 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.0018518833676353097, | |
| "learning_rate": 5.651358950328023e-06, | |
| "loss": 0.0001, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.002825092989951372, | |
| "learning_rate": 5.604498594189316e-06, | |
| "loss": 0.0002, | |
| "step": 10660 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.0021178831811994314, | |
| "learning_rate": 5.557638238050609e-06, | |
| "loss": 0.0013, | |
| "step": 10670 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.0012423048028722405, | |
| "learning_rate": 5.5107778819119025e-06, | |
| "loss": 0.072, | |
| "step": 10680 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.002971925074234605, | |
| "learning_rate": 5.4639175257731965e-06, | |
| "loss": 0.0001, | |
| "step": 10690 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.001466022222302854, | |
| "learning_rate": 5.41705716963449e-06, | |
| "loss": 0.0001, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.002207581652328372, | |
| "learning_rate": 5.370196813495783e-06, | |
| "loss": 0.0001, | |
| "step": 10710 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.0013427763478830457, | |
| "learning_rate": 5.323336457357076e-06, | |
| "loss": 0.0543, | |
| "step": 10720 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "grad_norm": 0.0013656431110575795, | |
| "learning_rate": 5.276476101218369e-06, | |
| "loss": 0.0191, | |
| "step": 10730 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.5649217367172241, | |
| "learning_rate": 5.229615745079663e-06, | |
| "loss": 0.0178, | |
| "step": 10740 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.0019292469369247556, | |
| "learning_rate": 5.182755388940956e-06, | |
| "loss": 0.008, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.005805399268865585, | |
| "learning_rate": 5.1358950328022495e-06, | |
| "loss": 0.0001, | |
| "step": 10760 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 1.1440871953964233, | |
| "learning_rate": 5.089034676663543e-06, | |
| "loss": 0.067, | |
| "step": 10770 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.0036515570245683193, | |
| "learning_rate": 5.042174320524836e-06, | |
| "loss": 0.0027, | |
| "step": 10780 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.0021176172886043787, | |
| "learning_rate": 4.995313964386129e-06, | |
| "loss": 0.0175, | |
| "step": 10790 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.04823027923703194, | |
| "learning_rate": 4.948453608247423e-06, | |
| "loss": 0.0059, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.006527293939143419, | |
| "learning_rate": 4.901593252108716e-06, | |
| "loss": 0.0001, | |
| "step": 10810 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.02050858922302723, | |
| "learning_rate": 4.854732895970009e-06, | |
| "loss": 0.0013, | |
| "step": 10820 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.0015726288547739387, | |
| "learning_rate": 4.8078725398313025e-06, | |
| "loss": 0.0001, | |
| "step": 10830 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.00158556061796844, | |
| "learning_rate": 4.761012183692596e-06, | |
| "loss": 0.0394, | |
| "step": 10840 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "grad_norm": 0.0019515565363690257, | |
| "learning_rate": 4.71415182755389e-06, | |
| "loss": 0.0048, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 32.03, | |
| "grad_norm": 0.0021331189200282097, | |
| "learning_rate": 4.667291471415184e-06, | |
| "loss": 0.0001, | |
| "step": 10860 | |
| }, | |
| { | |
| "epoch": 32.03, | |
| "grad_norm": 0.002155827358365059, | |
| "learning_rate": 4.620431115276477e-06, | |
| "loss": 0.0003, | |
| "step": 10870 | |
| }, | |
| { | |
| "epoch": 32.03, | |
| "grad_norm": 0.0019439981551840901, | |
| "learning_rate": 4.57357075913777e-06, | |
| "loss": 0.0002, | |
| "step": 10880 | |
| }, | |
| { | |
| "epoch": 32.03, | |
| "grad_norm": 0.0034947495441883802, | |
| "learning_rate": 4.526710402999063e-06, | |
| "loss": 0.0157, | |
| "step": 10890 | |
| }, | |
| { | |
| "epoch": 32.03, | |
| "eval_accuracy": 0.8022922636103151, | |
| "eval_loss": 1.4391096830368042, | |
| "eval_runtime": 33.8979, | |
| "eval_samples_per_second": 20.591, | |
| "eval_steps_per_second": 1.741, | |
| "step": 10890 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 0.007783001288771629, | |
| "learning_rate": 4.479850046860356e-06, | |
| "loss": 0.0002, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 0.004935835022479296, | |
| "learning_rate": 4.43298969072165e-06, | |
| "loss": 0.0065, | |
| "step": 10910 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 0.0015104643534868956, | |
| "learning_rate": 4.3861293345829435e-06, | |
| "loss": 0.0765, | |
| "step": 10920 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 0.0014216667041182518, | |
| "learning_rate": 4.339268978444237e-06, | |
| "loss": 0.0018, | |
| "step": 10930 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 0.022745607420802116, | |
| "learning_rate": 4.29240862230553e-06, | |
| "loss": 0.0081, | |
| "step": 10940 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.0021193595603108406, | |
| "learning_rate": 4.245548266166823e-06, | |
| "loss": 0.0003, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.0011568386107683182, | |
| "learning_rate": 4.198687910028116e-06, | |
| "loss": 0.0001, | |
| "step": 10960 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.019418692216277122, | |
| "learning_rate": 4.15182755388941e-06, | |
| "loss": 0.1781, | |
| "step": 10970 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 34.7269401550293, | |
| "learning_rate": 4.104967197750703e-06, | |
| "loss": 0.0067, | |
| "step": 10980 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.0017960992408916354, | |
| "learning_rate": 4.0581068416119964e-06, | |
| "loss": 0.0001, | |
| "step": 10990 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.01256940234452486, | |
| "learning_rate": 4.01124648547329e-06, | |
| "loss": 0.0003, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.0063404180109500885, | |
| "learning_rate": 3.964386129334583e-06, | |
| "loss": 0.0001, | |
| "step": 11010 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.018559589982032776, | |
| "learning_rate": 3.917525773195877e-06, | |
| "loss": 0.0001, | |
| "step": 11020 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.0015321632381528616, | |
| "learning_rate": 3.87066541705717e-06, | |
| "loss": 0.0001, | |
| "step": 11030 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.020633699372410774, | |
| "learning_rate": 3.823805060918463e-06, | |
| "loss": 0.0001, | |
| "step": 11040 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.0021525041665881872, | |
| "learning_rate": 3.7769447047797563e-06, | |
| "loss": 0.0601, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.0019328080816194415, | |
| "learning_rate": 3.73008434864105e-06, | |
| "loss": 0.0002, | |
| "step": 11060 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.004199670627713203, | |
| "learning_rate": 3.683223992502343e-06, | |
| "loss": 0.0377, | |
| "step": 11070 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.0021045091561973095, | |
| "learning_rate": 3.636363636363636e-06, | |
| "loss": 0.0001, | |
| "step": 11080 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.0019059345358982682, | |
| "learning_rate": 3.5895032802249297e-06, | |
| "loss": 0.0001, | |
| "step": 11090 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.002579369815066457, | |
| "learning_rate": 3.542642924086223e-06, | |
| "loss": 0.0373, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.0014094141079112887, | |
| "learning_rate": 3.4957825679475165e-06, | |
| "loss": 0.0001, | |
| "step": 11110 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.0017881103558465838, | |
| "learning_rate": 3.4489222118088097e-06, | |
| "loss": 0.0093, | |
| "step": 11120 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.001685052178800106, | |
| "learning_rate": 3.402061855670103e-06, | |
| "loss": 0.0556, | |
| "step": 11130 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.004337169695645571, | |
| "learning_rate": 3.3552014995313964e-06, | |
| "loss": 0.0396, | |
| "step": 11140 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.011138683184981346, | |
| "learning_rate": 3.3083411433926896e-06, | |
| "loss": 0.0001, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.0024086865596473217, | |
| "learning_rate": 3.2614807872539836e-06, | |
| "loss": 0.0001, | |
| "step": 11160 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.0029931641183793545, | |
| "learning_rate": 3.2146204311152767e-06, | |
| "loss": 0.0001, | |
| "step": 11170 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "grad_norm": 0.42393574118614197, | |
| "learning_rate": 3.1677600749765703e-06, | |
| "loss": 0.0016, | |
| "step": 11180 | |
| }, | |
| { | |
| "epoch": 33.03, | |
| "grad_norm": 0.0025856320280581713, | |
| "learning_rate": 3.120899718837863e-06, | |
| "loss": 0.0001, | |
| "step": 11190 | |
| }, | |
| { | |
| "epoch": 33.03, | |
| "grad_norm": 0.0019572244491428137, | |
| "learning_rate": 3.0740393626991566e-06, | |
| "loss": 0.0001, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 33.03, | |
| "grad_norm": 0.3476734459400177, | |
| "learning_rate": 3.02717900656045e-06, | |
| "loss": 0.0044, | |
| "step": 11210 | |
| }, | |
| { | |
| "epoch": 33.03, | |
| "grad_norm": 0.0009709845762699842, | |
| "learning_rate": 2.9803186504217434e-06, | |
| "loss": 0.0001, | |
| "step": 11220 | |
| }, | |
| { | |
| "epoch": 33.03, | |
| "eval_accuracy": 0.8051575931232091, | |
| "eval_loss": 1.383094310760498, | |
| "eval_runtime": 33.8221, | |
| "eval_samples_per_second": 20.637, | |
| "eval_steps_per_second": 1.744, | |
| "step": 11220 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 0.0015264974208548665, | |
| "learning_rate": 2.9334582942830366e-06, | |
| "loss": 0.0001, | |
| "step": 11230 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 0.003399324370548129, | |
| "learning_rate": 2.88659793814433e-06, | |
| "loss": 0.0514, | |
| "step": 11240 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 0.0033612081315368414, | |
| "learning_rate": 2.8397375820056233e-06, | |
| "loss": 0.0001, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 0.0017946057487279177, | |
| "learning_rate": 2.792877225866917e-06, | |
| "loss": 0.0001, | |
| "step": 11260 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 0.001860388438217342, | |
| "learning_rate": 2.74601686972821e-06, | |
| "loss": 0.0001, | |
| "step": 11270 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.0021725373808294535, | |
| "learning_rate": 2.6991565135895036e-06, | |
| "loss": 0.0001, | |
| "step": 11280 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.030490437522530556, | |
| "learning_rate": 2.652296157450797e-06, | |
| "loss": 0.0001, | |
| "step": 11290 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.0011160429567098618, | |
| "learning_rate": 2.60543580131209e-06, | |
| "loss": 0.0686, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.002387237036600709, | |
| "learning_rate": 2.5585754451733835e-06, | |
| "loss": 0.0074, | |
| "step": 11310 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.002495302353054285, | |
| "learning_rate": 2.5117150890346767e-06, | |
| "loss": 0.0002, | |
| "step": 11320 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.022143444046378136, | |
| "learning_rate": 2.46485473289597e-06, | |
| "loss": 0.0002, | |
| "step": 11330 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.002796097891405225, | |
| "learning_rate": 2.4179943767572634e-06, | |
| "loss": 0.0001, | |
| "step": 11340 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.0035200004931539297, | |
| "learning_rate": 2.3711340206185566e-06, | |
| "loss": 0.0428, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 27.465776443481445, | |
| "learning_rate": 2.32427366447985e-06, | |
| "loss": 0.059, | |
| "step": 11360 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.00824726838618517, | |
| "learning_rate": 2.2774133083411434e-06, | |
| "loss": 0.0537, | |
| "step": 11370 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.0025804354809224606, | |
| "learning_rate": 2.2305529522024365e-06, | |
| "loss": 0.0001, | |
| "step": 11380 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "grad_norm": 0.01662873476743698, | |
| "learning_rate": 2.1836925960637305e-06, | |
| "loss": 0.0068, | |
| "step": 11390 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.0012878773268312216, | |
| "learning_rate": 2.1368322399250237e-06, | |
| "loss": 0.0001, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.0017231220845133066, | |
| "learning_rate": 2.089971883786317e-06, | |
| "loss": 0.0002, | |
| "step": 11410 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.007844923995435238, | |
| "learning_rate": 2.0431115276476104e-06, | |
| "loss": 0.0001, | |
| "step": 11420 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.001733616110868752, | |
| "learning_rate": 1.9962511715089036e-06, | |
| "loss": 0.0034, | |
| "step": 11430 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.002772507956251502, | |
| "learning_rate": 1.9493908153701968e-06, | |
| "loss": 0.0234, | |
| "step": 11440 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.001203456544317305, | |
| "learning_rate": 1.9025304592314903e-06, | |
| "loss": 0.0001, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.004119969438761473, | |
| "learning_rate": 1.8556701030927835e-06, | |
| "loss": 0.0399, | |
| "step": 11460 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.0023524747230112553, | |
| "learning_rate": 1.8088097469540769e-06, | |
| "loss": 0.0003, | |
| "step": 11470 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.0018690053839236498, | |
| "learning_rate": 1.7619493908153703e-06, | |
| "loss": 0.0001, | |
| "step": 11480 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.0028986113611608744, | |
| "learning_rate": 1.7150890346766636e-06, | |
| "loss": 0.0371, | |
| "step": 11490 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.0018392838537693024, | |
| "learning_rate": 1.6682286785379568e-06, | |
| "loss": 0.0001, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "grad_norm": 0.0016958696069195867, | |
| "learning_rate": 1.6213683223992502e-06, | |
| "loss": 0.0001, | |
| "step": 11510 | |
| }, | |
| { | |
| "epoch": 34.03, | |
| "grad_norm": 0.004104943014681339, | |
| "learning_rate": 1.5745079662605435e-06, | |
| "loss": 0.0001, | |
| "step": 11520 | |
| }, | |
| { | |
| "epoch": 34.03, | |
| "grad_norm": 0.008021462708711624, | |
| "learning_rate": 1.527647610121837e-06, | |
| "loss": 0.0001, | |
| "step": 11530 | |
| }, | |
| { | |
| "epoch": 34.03, | |
| "grad_norm": 5.87186861038208, | |
| "learning_rate": 1.4807872539831303e-06, | |
| "loss": 0.0109, | |
| "step": 11540 | |
| }, | |
| { | |
| "epoch": 34.03, | |
| "grad_norm": 0.0011679594172164798, | |
| "learning_rate": 1.4339268978444237e-06, | |
| "loss": 0.0001, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 34.03, | |
| "eval_accuracy": 0.8080229226361032, | |
| "eval_loss": 1.3971121311187744, | |
| "eval_runtime": 33.9156, | |
| "eval_samples_per_second": 20.581, | |
| "eval_steps_per_second": 1.74, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 0.0017861544620245695, | |
| "learning_rate": 1.387066541705717e-06, | |
| "loss": 0.0001, | |
| "step": 11560 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 0.0019135787151753902, | |
| "learning_rate": 1.3402061855670102e-06, | |
| "loss": 0.0002, | |
| "step": 11570 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 0.04832134768366814, | |
| "learning_rate": 1.2933458294283038e-06, | |
| "loss": 0.0001, | |
| "step": 11580 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 0.0034780928399413824, | |
| "learning_rate": 1.2464854732895972e-06, | |
| "loss": 0.0001, | |
| "step": 11590 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 0.001626980840228498, | |
| "learning_rate": 1.1996251171508905e-06, | |
| "loss": 0.0003, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.006935800425708294, | |
| "learning_rate": 1.1527647610121837e-06, | |
| "loss": 0.0001, | |
| "step": 11610 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.0010628902819007635, | |
| "learning_rate": 1.105904404873477e-06, | |
| "loss": 0.001, | |
| "step": 11620 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.0014588043559342623, | |
| "learning_rate": 1.0590440487347704e-06, | |
| "loss": 0.0001, | |
| "step": 11630 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.007299583870917559, | |
| "learning_rate": 1.0121836925960638e-06, | |
| "loss": 0.0001, | |
| "step": 11640 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.002258384833112359, | |
| "learning_rate": 9.65323336457357e-07, | |
| "loss": 0.0001, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 61.81095504760742, | |
| "learning_rate": 9.184629803186506e-07, | |
| "loss": 0.0222, | |
| "step": 11660 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.0012000646675005555, | |
| "learning_rate": 8.716026241799438e-07, | |
| "loss": 0.0001, | |
| "step": 11670 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.01990874670445919, | |
| "learning_rate": 8.247422680412372e-07, | |
| "loss": 0.0001, | |
| "step": 11680 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.01420762948691845, | |
| "learning_rate": 7.778819119025305e-07, | |
| "loss": 0.0001, | |
| "step": 11690 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.002269094344228506, | |
| "learning_rate": 7.310215557638238e-07, | |
| "loss": 0.0001, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.0038073186296969652, | |
| "learning_rate": 6.841611996251172e-07, | |
| "loss": 0.0181, | |
| "step": 11710 | |
| }, | |
| { | |
| "epoch": 35.01, | |
| "grad_norm": 0.03407540172338486, | |
| "learning_rate": 6.373008434864106e-07, | |
| "loss": 0.029, | |
| "step": 11720 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.030750174075365067, | |
| "learning_rate": 5.904404873477039e-07, | |
| "loss": 0.0093, | |
| "step": 11730 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.002124561695381999, | |
| "learning_rate": 5.435801312089972e-07, | |
| "loss": 0.0027, | |
| "step": 11740 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.0012157908640801907, | |
| "learning_rate": 4.967197750702906e-07, | |
| "loss": 0.0001, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.002194877015426755, | |
| "learning_rate": 4.498594189315839e-07, | |
| "loss": 0.0001, | |
| "step": 11760 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.003094845451414585, | |
| "learning_rate": 4.0299906279287724e-07, | |
| "loss": 0.0001, | |
| "step": 11770 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.001540425349958241, | |
| "learning_rate": 3.561387066541706e-07, | |
| "loss": 0.0001, | |
| "step": 11780 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.0017908586887642741, | |
| "learning_rate": 3.0927835051546394e-07, | |
| "loss": 0.0001, | |
| "step": 11790 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.007806339301168919, | |
| "learning_rate": 2.6241799437675726e-07, | |
| "loss": 0.0399, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.004247527569532394, | |
| "learning_rate": 2.155576382380506e-07, | |
| "loss": 0.034, | |
| "step": 11810 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.0017347713001072407, | |
| "learning_rate": 1.6869728209934398e-07, | |
| "loss": 0.0005, | |
| "step": 11820 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.001252649468369782, | |
| "learning_rate": 1.218369259606373e-07, | |
| "loss": 0.0001, | |
| "step": 11830 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "grad_norm": 0.0012991786934435368, | |
| "learning_rate": 7.497656982193066e-08, | |
| "loss": 0.0528, | |
| "step": 11840 | |
| }, | |
| { | |
| "epoch": 35.03, | |
| "grad_norm": 11.709962844848633, | |
| "learning_rate": 2.8116213683223995e-08, | |
| "loss": 0.0041, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 35.03, | |
| "eval_accuracy": 0.8080229226361032, | |
| "eval_loss": 1.3906208276748657, | |
| "eval_runtime": 33.666, | |
| "eval_samples_per_second": 20.733, | |
| "eval_steps_per_second": 1.753, | |
| "step": 11856 | |
| }, | |
| { | |
| "epoch": 35.03, | |
| "step": 11856, | |
| "total_flos": 1.770184277095349e+20, | |
| "train_loss": 0.18484977827027105, | |
| "train_runtime": 12835.1405, | |
| "train_samples_per_second": 11.085, | |
| "train_steps_per_second": 0.924 | |
| }, | |
| { | |
| "epoch": 35.03, | |
| "eval_accuracy": 0.8209169054441261, | |
| "eval_loss": 0.8522207736968994, | |
| "eval_runtime": 35.5618, | |
| "eval_samples_per_second": 19.628, | |
| "eval_steps_per_second": 1.659, | |
| "step": 11856 | |
| }, | |
| { | |
| "epoch": 35.03, | |
| "eval_accuracy": 0.8209169054441261, | |
| "eval_loss": 0.8522207736968994, | |
| "eval_runtime": 33.138, | |
| "eval_samples_per_second": 21.063, | |
| "eval_steps_per_second": 1.78, | |
| "step": 11856 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 11856, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 1.770184277095349e+20, | |
| "train_batch_size": 12, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |