| { | |
| "best_global_step": 288, | |
| "best_metric": 0.7926829268292683, | |
| "best_model_checkpoint": "ALL_RGBCROP_ori16F-8B16F-GACWD5lrDO/checkpoint-288", | |
| "epoch": 10.025, | |
| "eval_steps": 500, | |
| "global_step": 528, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005208333333333333, | |
| "grad_norm": 6.546933174133301, | |
| "learning_rate": 4.6875000000000006e-07, | |
| "loss": 0.7185, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.010416666666666666, | |
| "grad_norm": 6.9191694259643555, | |
| "learning_rate": 9.895833333333333e-07, | |
| "loss": 0.6892, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.015625, | |
| "grad_norm": 6.326386451721191, | |
| "learning_rate": 1.5104166666666667e-06, | |
| "loss": 0.7244, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.020833333333333332, | |
| "grad_norm": 8.42184829711914, | |
| "learning_rate": 2.0312500000000002e-06, | |
| "loss": 0.694, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "eval_accuracy": 0.49390243902439024, | |
| "eval_loss": 0.7127410769462585, | |
| "eval_runtime": 25.0192, | |
| "eval_samples_per_second": 6.555, | |
| "eval_steps_per_second": 0.839, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.0010416666666666, | |
| "grad_norm": 6.573472499847412, | |
| "learning_rate": 2.5520833333333334e-06, | |
| "loss": 0.6915, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.00625, | |
| "grad_norm": 6.270707130432129, | |
| "learning_rate": 3.072916666666667e-06, | |
| "loss": 0.6531, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0114583333333333, | |
| "grad_norm": 6.025201797485352, | |
| "learning_rate": 3.59375e-06, | |
| "loss": 0.632, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.0166666666666666, | |
| "grad_norm": 6.816848278045654, | |
| "learning_rate": 4.114583333333334e-06, | |
| "loss": 0.6395, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.021875, | |
| "grad_norm": 6.194951057434082, | |
| "learning_rate": 4.635416666666667e-06, | |
| "loss": 0.6395, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.025, | |
| "eval_accuracy": 0.5975609756097561, | |
| "eval_loss": 0.6772211790084839, | |
| "eval_runtime": 25.1799, | |
| "eval_samples_per_second": 6.513, | |
| "eval_steps_per_second": 0.834, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.002083333333333, | |
| "grad_norm": 6.220993995666504, | |
| "learning_rate": 5.156250000000001e-06, | |
| "loss": 0.6222, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.0072916666666667, | |
| "grad_norm": 6.413397312164307, | |
| "learning_rate": 5.677083333333334e-06, | |
| "loss": 0.6061, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.0125, | |
| "grad_norm": 7.089827537536621, | |
| "learning_rate": 6.197916666666667e-06, | |
| "loss": 0.5924, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.017708333333333, | |
| "grad_norm": 6.679469108581543, | |
| "learning_rate": 6.718750000000001e-06, | |
| "loss": 0.5323, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.0229166666666667, | |
| "grad_norm": 5.224033832550049, | |
| "learning_rate": 7.239583333333334e-06, | |
| "loss": 0.4871, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.025, | |
| "eval_accuracy": 0.6707317073170732, | |
| "eval_loss": 0.6142382025718689, | |
| "eval_runtime": 25.3499, | |
| "eval_samples_per_second": 6.469, | |
| "eval_steps_per_second": 0.828, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 3.003125, | |
| "grad_norm": 6.379062652587891, | |
| "learning_rate": 7.760416666666666e-06, | |
| "loss": 0.4493, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.0083333333333333, | |
| "grad_norm": 5.42802619934082, | |
| "learning_rate": 8.281250000000001e-06, | |
| "loss": 0.4371, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.013541666666667, | |
| "grad_norm": 6.1493096351623535, | |
| "learning_rate": 8.802083333333335e-06, | |
| "loss": 0.402, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.01875, | |
| "grad_norm": 5.070779323577881, | |
| "learning_rate": 9.322916666666667e-06, | |
| "loss": 0.4197, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.0239583333333333, | |
| "grad_norm": 6.595272064208984, | |
| "learning_rate": 9.84375e-06, | |
| "loss": 0.4162, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.025, | |
| "eval_accuracy": 0.725609756097561, | |
| "eval_loss": 0.5417684316635132, | |
| "eval_runtime": 24.8265, | |
| "eval_samples_per_second": 6.606, | |
| "eval_steps_per_second": 0.846, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 4.004166666666666, | |
| "grad_norm": 12.621065139770508, | |
| "learning_rate": 9.959490740740742e-06, | |
| "loss": 0.3567, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.009375, | |
| "grad_norm": 7.9822893142700195, | |
| "learning_rate": 9.901620370370372e-06, | |
| "loss": 0.319, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.014583333333333, | |
| "grad_norm": 5.883790969848633, | |
| "learning_rate": 9.84375e-06, | |
| "loss": 0.2885, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.019791666666666, | |
| "grad_norm": 10.191251754760742, | |
| "learning_rate": 9.78587962962963e-06, | |
| "loss": 0.2845, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.025, | |
| "grad_norm": 11.520015716552734, | |
| "learning_rate": 9.72800925925926e-06, | |
| "loss": 0.2336, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.025, | |
| "eval_accuracy": 0.7621951219512195, | |
| "eval_loss": 0.5030018091201782, | |
| "eval_runtime": 25.1022, | |
| "eval_samples_per_second": 6.533, | |
| "eval_steps_per_second": 0.837, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.005208333333333, | |
| "grad_norm": 9.863582611083984, | |
| "learning_rate": 9.670138888888889e-06, | |
| "loss": 0.2117, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.010416666666667, | |
| "grad_norm": 9.696749687194824, | |
| "learning_rate": 9.612268518518519e-06, | |
| "loss": 0.2162, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 5.015625, | |
| "grad_norm": 5.5694580078125, | |
| "learning_rate": 9.554398148148149e-06, | |
| "loss": 0.224, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.020833333333333, | |
| "grad_norm": 2.6210453510284424, | |
| "learning_rate": 9.496527777777779e-06, | |
| "loss": 0.159, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.025, | |
| "eval_accuracy": 0.7926829268292683, | |
| "eval_loss": 0.5044757723808289, | |
| "eval_runtime": 23.3535, | |
| "eval_samples_per_second": 7.023, | |
| "eval_steps_per_second": 0.899, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 6.001041666666667, | |
| "grad_norm": 7.435839653015137, | |
| "learning_rate": 9.438657407407409e-06, | |
| "loss": 0.1674, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 6.00625, | |
| "grad_norm": 39.46736526489258, | |
| "learning_rate": 9.380787037037039e-06, | |
| "loss": 0.1278, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.011458333333334, | |
| "grad_norm": 1.9832502603530884, | |
| "learning_rate": 9.322916666666667e-06, | |
| "loss": 0.1017, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 6.016666666666667, | |
| "grad_norm": 7.218878269195557, | |
| "learning_rate": 9.265046296296297e-06, | |
| "loss": 0.1574, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 6.021875, | |
| "grad_norm": 6.560127258300781, | |
| "learning_rate": 9.207175925925927e-06, | |
| "loss": 0.1486, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 6.025, | |
| "eval_accuracy": 0.7804878048780488, | |
| "eval_loss": 0.5186034440994263, | |
| "eval_runtime": 25.0796, | |
| "eval_samples_per_second": 6.539, | |
| "eval_steps_per_second": 0.837, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 7.002083333333333, | |
| "grad_norm": 1.7817628383636475, | |
| "learning_rate": 9.149305555555557e-06, | |
| "loss": 0.0987, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 7.007291666666666, | |
| "grad_norm": 10.60669994354248, | |
| "learning_rate": 9.091435185185185e-06, | |
| "loss": 0.1023, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 7.0125, | |
| "grad_norm": 0.7290067076683044, | |
| "learning_rate": 9.033564814814815e-06, | |
| "loss": 0.097, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 7.017708333333333, | |
| "grad_norm": 5.742852687835693, | |
| "learning_rate": 8.975694444444445e-06, | |
| "loss": 0.1295, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 7.022916666666666, | |
| "grad_norm": 1.4199665784835815, | |
| "learning_rate": 8.917824074074075e-06, | |
| "loss": 0.0997, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 7.025, | |
| "eval_accuracy": 0.7865853658536586, | |
| "eval_loss": 0.5648921728134155, | |
| "eval_runtime": 25.5274, | |
| "eval_samples_per_second": 6.424, | |
| "eval_steps_per_second": 0.823, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 8.003125, | |
| "grad_norm": 1.0618127584457397, | |
| "learning_rate": 8.859953703703705e-06, | |
| "loss": 0.0975, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 8.008333333333333, | |
| "grad_norm": 0.4267812967300415, | |
| "learning_rate": 8.802083333333335e-06, | |
| "loss": 0.0738, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 8.013541666666667, | |
| "grad_norm": 16.63909339904785, | |
| "learning_rate": 8.744212962962963e-06, | |
| "loss": 0.0395, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 8.01875, | |
| "grad_norm": 16.013010025024414, | |
| "learning_rate": 8.686342592592593e-06, | |
| "loss": 0.0566, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 8.023958333333333, | |
| "grad_norm": 3.0400283336639404, | |
| "learning_rate": 8.628472222222223e-06, | |
| "loss": 0.07, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 8.025, | |
| "eval_accuracy": 0.7804878048780488, | |
| "eval_loss": 0.6179517507553101, | |
| "eval_runtime": 24.5292, | |
| "eval_samples_per_second": 6.686, | |
| "eval_steps_per_second": 0.856, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 9.004166666666666, | |
| "grad_norm": 14.060826301574707, | |
| "learning_rate": 8.570601851851853e-06, | |
| "loss": 0.0566, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 9.009375, | |
| "grad_norm": 5.963779926300049, | |
| "learning_rate": 8.512731481481481e-06, | |
| "loss": 0.0421, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 9.014583333333333, | |
| "grad_norm": 1.6425102949142456, | |
| "learning_rate": 8.454861111111111e-06, | |
| "loss": 0.0395, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 9.019791666666666, | |
| "grad_norm": 7.537419319152832, | |
| "learning_rate": 8.396990740740741e-06, | |
| "loss": 0.0402, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 9.025, | |
| "grad_norm": 17.922319412231445, | |
| "learning_rate": 8.339120370370371e-06, | |
| "loss": 0.0377, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 9.025, | |
| "eval_accuracy": 0.7926829268292683, | |
| "eval_loss": 0.6363745927810669, | |
| "eval_runtime": 24.3896, | |
| "eval_samples_per_second": 6.724, | |
| "eval_steps_per_second": 0.861, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 10.005208333333334, | |
| "grad_norm": 0.31738758087158203, | |
| "learning_rate": 8.281250000000001e-06, | |
| "loss": 0.0276, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 10.010416666666666, | |
| "grad_norm": 2.9940338134765625, | |
| "learning_rate": 8.22337962962963e-06, | |
| "loss": 0.0255, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 10.015625, | |
| "grad_norm": 24.363996505737305, | |
| "learning_rate": 8.16550925925926e-06, | |
| "loss": 0.0269, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 10.020833333333334, | |
| "grad_norm": 0.5430907011032104, | |
| "learning_rate": 8.10763888888889e-06, | |
| "loss": 0.0103, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 10.025, | |
| "eval_accuracy": 0.7865853658536586, | |
| "eval_loss": 0.7102433443069458, | |
| "eval_runtime": 24.5626, | |
| "eval_samples_per_second": 6.677, | |
| "eval_steps_per_second": 0.855, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 10.025, | |
| "step": 528, | |
| "total_flos": 1.0526753651857293e+19, | |
| "train_loss": 0.2883184655625938, | |
| "train_runtime": 2976.1505, | |
| "train_samples_per_second": 10.322, | |
| "train_steps_per_second": 0.645 | |
| }, | |
| { | |
| "epoch": 10.025, | |
| "eval_accuracy": 0.8143712574850299, | |
| "eval_loss": 0.40913674235343933, | |
| "eval_runtime": 27.9503, | |
| "eval_samples_per_second": 5.975, | |
| "eval_steps_per_second": 0.751, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 10.025, | |
| "eval_accuracy": 0.8143712574850299, | |
| "eval_loss": 0.4080381691455841, | |
| "eval_runtime": 26.7248, | |
| "eval_samples_per_second": 6.249, | |
| "eval_steps_per_second": 0.786, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 10.025, | |
| "eval_accuracy": 0.8143712574850299, | |
| "eval_loss": 0.4105586111545563, | |
| "eval_runtime": 25.8358, | |
| "eval_samples_per_second": 6.464, | |
| "eval_steps_per_second": 0.813, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 10.025, | |
| "eval_accuracy": 0.8143712574850299, | |
| "eval_loss": 0.40935227274894714, | |
| "eval_runtime": 27.3841, | |
| "eval_samples_per_second": 6.098, | |
| "eval_steps_per_second": 0.767, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 10.025, | |
| "eval_accuracy": 0.8143712574850299, | |
| "eval_loss": 0.40921783447265625, | |
| "eval_runtime": 26.5963, | |
| "eval_samples_per_second": 6.279, | |
| "eval_steps_per_second": 0.79, | |
| "step": 528 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1920, | |
| "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": 1.0526753651857293e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |