| { | |
| "best_global_step": 392, | |
| "best_metric": 0.9479166666666666, | |
| "best_model_checkpoint": "BxSD_RGBCROP_Aug16F-8B16F-GACWDlr/checkpoint-392", | |
| "epoch": 11.025112107623318, | |
| "eval_steps": 500, | |
| "global_step": 672, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004484304932735426, | |
| "grad_norm": 3.1429996490478516, | |
| "learning_rate": 4.0358744394618834e-07, | |
| "loss": 0.7032, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008968609865470852, | |
| "grad_norm": 3.9541077613830566, | |
| "learning_rate": 8.520179372197309e-07, | |
| "loss": 0.6956, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013452914798206279, | |
| "grad_norm": 3.382608413696289, | |
| "learning_rate": 1.3004484304932734e-06, | |
| "loss": 0.7003, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.017937219730941704, | |
| "grad_norm": 3.3949358463287354, | |
| "learning_rate": 1.7488789237668164e-06, | |
| "loss": 0.6888, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02242152466367713, | |
| "grad_norm": 2.802041530609131, | |
| "learning_rate": 2.197309417040359e-06, | |
| "loss": 0.663, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.025112107623318385, | |
| "eval_accuracy": 0.6354166666666666, | |
| "eval_loss": 0.6602393984794617, | |
| "eval_runtime": 30.0703, | |
| "eval_samples_per_second": 6.385, | |
| "eval_steps_per_second": 0.798, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0017937219730941, | |
| "grad_norm": 3.6131398677825928, | |
| "learning_rate": 2.6457399103139016e-06, | |
| "loss": 0.6442, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0062780269058296, | |
| "grad_norm": 3.0706064701080322, | |
| "learning_rate": 3.0941704035874443e-06, | |
| "loss": 0.5936, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.010762331838565, | |
| "grad_norm": 3.5030055046081543, | |
| "learning_rate": 3.542600896860987e-06, | |
| "loss": 0.5627, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0152466367713004, | |
| "grad_norm": 4.468491077423096, | |
| "learning_rate": 3.991031390134529e-06, | |
| "loss": 0.5351, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.0197309417040359, | |
| "grad_norm": 3.892331600189209, | |
| "learning_rate": 4.439461883408072e-06, | |
| "loss": 0.4513, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0242152466367713, | |
| "grad_norm": 4.7978291511535645, | |
| "learning_rate": 4.887892376681615e-06, | |
| "loss": 0.4419, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0251121076233183, | |
| "eval_accuracy": 0.8125, | |
| "eval_loss": 0.4702857434749603, | |
| "eval_runtime": 29.1485, | |
| "eval_samples_per_second": 6.587, | |
| "eval_steps_per_second": 0.823, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.0035874439461883, | |
| "grad_norm": 2.970027208328247, | |
| "learning_rate": 5.3363228699551574e-06, | |
| "loss": 0.3586, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.008071748878924, | |
| "grad_norm": 4.3814697265625, | |
| "learning_rate": 5.7847533632287e-06, | |
| "loss": 0.2856, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.012556053811659, | |
| "grad_norm": 3.017439603805542, | |
| "learning_rate": 6.233183856502243e-06, | |
| "loss": 0.2565, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.017040358744395, | |
| "grad_norm": 7.53488302230835, | |
| "learning_rate": 6.681614349775786e-06, | |
| "loss": 0.2532, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.02152466367713, | |
| "grad_norm": 1.6428829431533813, | |
| "learning_rate": 7.1300448430493275e-06, | |
| "loss": 0.1752, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0251121076233183, | |
| "eval_accuracy": 0.890625, | |
| "eval_loss": 0.3134443759918213, | |
| "eval_runtime": 28.705, | |
| "eval_samples_per_second": 6.689, | |
| "eval_steps_per_second": 0.836, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.000896860986547, | |
| "grad_norm": 1.969995379447937, | |
| "learning_rate": 7.57847533632287e-06, | |
| "loss": 0.1334, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.0053811659192826, | |
| "grad_norm": 1.8544121980667114, | |
| "learning_rate": 8.026905829596413e-06, | |
| "loss": 0.0963, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.009865470852018, | |
| "grad_norm": 3.637763023376465, | |
| "learning_rate": 8.475336322869956e-06, | |
| "loss": 0.062, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.0143497757847535, | |
| "grad_norm": 1.3532997369766235, | |
| "learning_rate": 8.923766816143498e-06, | |
| "loss": 0.0643, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.0188340807174887, | |
| "grad_norm": 0.3547247052192688, | |
| "learning_rate": 9.372197309417041e-06, | |
| "loss": 0.0566, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.0233183856502244, | |
| "grad_norm": 6.522655487060547, | |
| "learning_rate": 9.820627802690584e-06, | |
| "loss": 0.0656, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.0251121076233183, | |
| "eval_accuracy": 0.9375, | |
| "eval_loss": 0.2310737818479538, | |
| "eval_runtime": 29.4697, | |
| "eval_samples_per_second": 6.515, | |
| "eval_steps_per_second": 0.814, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 4.002690582959641, | |
| "grad_norm": 0.4656350910663605, | |
| "learning_rate": 9.970104633781765e-06, | |
| "loss": 0.0137, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.0071748878923765, | |
| "grad_norm": 0.16051346063613892, | |
| "learning_rate": 9.920279023418038e-06, | |
| "loss": 0.0085, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.011659192825112, | |
| "grad_norm": 0.23789650201797485, | |
| "learning_rate": 9.87045341305431e-06, | |
| "loss": 0.0135, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.016143497757848, | |
| "grad_norm": 0.908409595489502, | |
| "learning_rate": 9.820627802690584e-06, | |
| "loss": 0.0069, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.020627802690583, | |
| "grad_norm": 0.11391621083021164, | |
| "learning_rate": 9.770802192326857e-06, | |
| "loss": 0.0106, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.025112107623318, | |
| "grad_norm": 0.08297699689865112, | |
| "learning_rate": 9.72097658196313e-06, | |
| "loss": 0.0046, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.025112107623318, | |
| "eval_accuracy": 0.9427083333333334, | |
| "eval_loss": 0.28078916668891907, | |
| "eval_runtime": 29.5695, | |
| "eval_samples_per_second": 6.493, | |
| "eval_steps_per_second": 0.812, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.004484304932736, | |
| "grad_norm": 0.06622310727834702, | |
| "learning_rate": 9.671150971599403e-06, | |
| "loss": 0.0044, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.00896860986547, | |
| "grad_norm": 0.0407695509493351, | |
| "learning_rate": 9.621325361235676e-06, | |
| "loss": 0.0022, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.013452914798206, | |
| "grad_norm": 0.039162054657936096, | |
| "learning_rate": 9.571499750871948e-06, | |
| "loss": 0.0021, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 5.017937219730942, | |
| "grad_norm": 0.03691912814974785, | |
| "learning_rate": 9.521674140508222e-06, | |
| "loss": 0.0016, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.022421524663677, | |
| "grad_norm": 0.02548583783209324, | |
| "learning_rate": 9.471848530144496e-06, | |
| "loss": 0.0012, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 5.025112107623318, | |
| "eval_accuracy": 0.9427083333333334, | |
| "eval_loss": 0.2592454254627228, | |
| "eval_runtime": 29.1301, | |
| "eval_samples_per_second": 6.591, | |
| "eval_steps_per_second": 0.824, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 6.001793721973094, | |
| "grad_norm": 0.022539280354976654, | |
| "learning_rate": 9.422022919780767e-06, | |
| "loss": 0.0009, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 6.00627802690583, | |
| "grad_norm": 0.024405544623732567, | |
| "learning_rate": 9.372197309417041e-06, | |
| "loss": 0.001, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 6.010762331838565, | |
| "grad_norm": 0.020192157477140427, | |
| "learning_rate": 9.322371699053315e-06, | |
| "loss": 0.0008, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 6.0152466367713, | |
| "grad_norm": 0.01488098967820406, | |
| "learning_rate": 9.272546088689587e-06, | |
| "loss": 0.0008, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 6.019730941704036, | |
| "grad_norm": 0.033892132341861725, | |
| "learning_rate": 9.22272047832586e-06, | |
| "loss": 0.0009, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 6.024215246636771, | |
| "grad_norm": 0.01757296919822693, | |
| "learning_rate": 9.172894867962134e-06, | |
| "loss": 0.0007, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 6.025112107623318, | |
| "eval_accuracy": 0.9479166666666666, | |
| "eval_loss": 0.2576257288455963, | |
| "eval_runtime": 29.5146, | |
| "eval_samples_per_second": 6.505, | |
| "eval_steps_per_second": 0.813, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 7.003587443946189, | |
| "grad_norm": 0.013150567188858986, | |
| "learning_rate": 9.123069257598406e-06, | |
| "loss": 0.0007, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.0080717488789235, | |
| "grad_norm": 0.015239037573337555, | |
| "learning_rate": 9.07324364723468e-06, | |
| "loss": 0.0006, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 7.012556053811659, | |
| "grad_norm": 0.018098143860697746, | |
| "learning_rate": 9.023418036870953e-06, | |
| "loss": 0.0006, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 7.017040358744395, | |
| "grad_norm": 0.010625138878822327, | |
| "learning_rate": 8.973592426507225e-06, | |
| "loss": 0.0005, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 7.0215246636771305, | |
| "grad_norm": 0.014908026903867722, | |
| "learning_rate": 8.923766816143498e-06, | |
| "loss": 0.0005, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 7.025112107623318, | |
| "eval_accuracy": 0.9479166666666666, | |
| "eval_loss": 0.2630920112133026, | |
| "eval_runtime": 28.1464, | |
| "eval_samples_per_second": 6.821, | |
| "eval_steps_per_second": 0.853, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 8.000896860986547, | |
| "grad_norm": 0.014149523340165615, | |
| "learning_rate": 8.873941205779772e-06, | |
| "loss": 0.0005, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 8.005381165919282, | |
| "grad_norm": 0.013675632886588573, | |
| "learning_rate": 8.824115595416045e-06, | |
| "loss": 0.0005, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 8.009865470852018, | |
| "grad_norm": 0.021204130724072456, | |
| "learning_rate": 8.774289985052317e-06, | |
| "loss": 0.0005, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 8.014349775784753, | |
| "grad_norm": 0.01068932469934225, | |
| "learning_rate": 8.724464374688591e-06, | |
| "loss": 0.0004, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 8.01883408071749, | |
| "grad_norm": 0.011664162389934063, | |
| "learning_rate": 8.674638764324865e-06, | |
| "loss": 0.0004, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 8.023318385650224, | |
| "grad_norm": 0.015364346094429493, | |
| "learning_rate": 8.624813153961136e-06, | |
| "loss": 0.0004, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.025112107623318, | |
| "eval_accuracy": 0.9479166666666666, | |
| "eval_loss": 0.2744747996330261, | |
| "eval_runtime": 30.499, | |
| "eval_samples_per_second": 6.295, | |
| "eval_steps_per_second": 0.787, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 9.00269058295964, | |
| "grad_norm": 0.009640182368457317, | |
| "learning_rate": 8.57498754359741e-06, | |
| "loss": 0.0004, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 9.007174887892377, | |
| "grad_norm": 0.008297263644635677, | |
| "learning_rate": 8.525161933233684e-06, | |
| "loss": 0.0004, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 9.011659192825112, | |
| "grad_norm": 0.00965182390064001, | |
| "learning_rate": 8.475336322869956e-06, | |
| "loss": 0.0004, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 9.016143497757847, | |
| "grad_norm": 0.008539954200387001, | |
| "learning_rate": 8.42551071250623e-06, | |
| "loss": 0.0003, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 9.020627802690584, | |
| "grad_norm": 0.01318803709000349, | |
| "learning_rate": 8.375685102142503e-06, | |
| "loss": 0.0003, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 9.025112107623318, | |
| "grad_norm": 0.013389434665441513, | |
| "learning_rate": 8.325859491778775e-06, | |
| "loss": 0.0004, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 9.025112107623318, | |
| "eval_accuracy": 0.9479166666666666, | |
| "eval_loss": 0.28136080503463745, | |
| "eval_runtime": 28.6966, | |
| "eval_samples_per_second": 6.691, | |
| "eval_steps_per_second": 0.836, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 10.004484304932735, | |
| "grad_norm": 0.0122822942212224, | |
| "learning_rate": 8.276033881415048e-06, | |
| "loss": 0.0003, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 10.008968609865471, | |
| "grad_norm": 0.030896110460162163, | |
| "learning_rate": 8.226208271051322e-06, | |
| "loss": 0.0003, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 10.013452914798206, | |
| "grad_norm": 0.012824727222323418, | |
| "learning_rate": 8.176382660687594e-06, | |
| "loss": 0.0003, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 10.01793721973094, | |
| "grad_norm": 0.00790180079638958, | |
| "learning_rate": 8.126557050323867e-06, | |
| "loss": 0.0003, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.022421524663677, | |
| "grad_norm": 0.011255297809839249, | |
| "learning_rate": 8.076731439960141e-06, | |
| "loss": 0.0003, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 10.025112107623318, | |
| "eval_accuracy": 0.9479166666666666, | |
| "eval_loss": 0.286301851272583, | |
| "eval_runtime": 30.4742, | |
| "eval_samples_per_second": 6.3, | |
| "eval_steps_per_second": 0.788, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 11.001793721973094, | |
| "grad_norm": 0.009772414341568947, | |
| "learning_rate": 8.026905829596413e-06, | |
| "loss": 0.0003, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 11.00627802690583, | |
| "grad_norm": 0.006887427996844053, | |
| "learning_rate": 7.977080219232686e-06, | |
| "loss": 0.0003, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 11.010762331838565, | |
| "grad_norm": 0.0069274078123271465, | |
| "learning_rate": 7.92725460886896e-06, | |
| "loss": 0.0003, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 11.0152466367713, | |
| "grad_norm": 0.010375886224210262, | |
| "learning_rate": 7.877428998505232e-06, | |
| "loss": 0.0003, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 11.019730941704037, | |
| "grad_norm": 0.006271818652749062, | |
| "learning_rate": 7.827603388141506e-06, | |
| "loss": 0.0002, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 11.024215246636771, | |
| "grad_norm": 0.006681330036371946, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 0.0002, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 11.025112107623318, | |
| "eval_accuracy": 0.9479166666666666, | |
| "eval_loss": 0.28504976630210876, | |
| "eval_runtime": 30.651, | |
| "eval_samples_per_second": 6.264, | |
| "eval_steps_per_second": 0.783, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 11.025112107623318, | |
| "step": 672, | |
| "total_flos": 1.336778091585287e+19, | |
| "train_loss": 0.12756322894849637, | |
| "train_runtime": 3434.3534, | |
| "train_samples_per_second": 10.389, | |
| "train_steps_per_second": 0.649 | |
| }, | |
| { | |
| "epoch": 11.025112107623318, | |
| "eval_accuracy": 0.958974358974359, | |
| "eval_loss": 0.1425883024930954, | |
| "eval_runtime": 29.0317, | |
| "eval_samples_per_second": 6.717, | |
| "eval_steps_per_second": 0.861, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 11.025112107623318, | |
| "eval_accuracy": 0.958974358974359, | |
| "eval_loss": 0.14259304106235504, | |
| "eval_runtime": 28.0432, | |
| "eval_samples_per_second": 6.954, | |
| "eval_steps_per_second": 0.891, | |
| "step": 672 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2230, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 5 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.336778091585287e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |