{ "best_metric": 0.9857142857142858, "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-525", "epoch": 7.125, "global_step": 600, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.02, "learning_rate": 8.333333333333334e-06, "loss": 2.4273, "step": 10 }, { "epoch": 0.03, "learning_rate": 1.6666666666666667e-05, "loss": 2.3496, "step": 20 }, { "epoch": 0.05, "learning_rate": 2.5e-05, "loss": 2.291, "step": 30 }, { "epoch": 0.07, "learning_rate": 3.3333333333333335e-05, "loss": 2.193, "step": 40 }, { "epoch": 0.08, "learning_rate": 4.166666666666667e-05, "loss": 2.2776, "step": 50 }, { "epoch": 0.1, "learning_rate": 5e-05, "loss": 2.0892, "step": 60 }, { "epoch": 0.12, "learning_rate": 4.9074074074074075e-05, "loss": 2.0934, "step": 70 }, { "epoch": 0.12, "eval_accuracy": 0.5428571428571428, "eval_loss": 1.7595646381378174, "eval_runtime": 17.4275, "eval_samples_per_second": 4.017, "eval_steps_per_second": 1.033, "step": 75 }, { "epoch": 1.01, "learning_rate": 4.814814814814815e-05, "loss": 1.6538, "step": 80 }, { "epoch": 1.02, "learning_rate": 4.722222222222222e-05, "loss": 1.3968, "step": 90 }, { "epoch": 1.04, "learning_rate": 4.62962962962963e-05, "loss": 1.3605, "step": 100 }, { "epoch": 1.06, "learning_rate": 4.5370370370370374e-05, "loss": 1.0257, "step": 110 }, { "epoch": 1.07, "learning_rate": 4.4444444444444447e-05, "loss": 1.0096, "step": 120 }, { "epoch": 1.09, "learning_rate": 4.351851851851852e-05, "loss": 0.7081, "step": 130 }, { "epoch": 1.11, "learning_rate": 4.259259259259259e-05, "loss": 0.6225, "step": 140 }, { "epoch": 1.12, "learning_rate": 4.166666666666667e-05, "loss": 0.9556, "step": 150 }, { "epoch": 1.12, "eval_accuracy": 0.6428571428571429, "eval_loss": 1.1258111000061035, "eval_runtime": 17.2592, "eval_samples_per_second": 4.056, "eval_steps_per_second": 1.043, "step": 150 }, { "epoch": 2.02, "learning_rate": 4.074074074074074e-05, "loss": 0.6893, "step": 160 }, { "epoch": 2.03, "learning_rate": 3.981481481481482e-05, "loss": 0.5757, "step": 170 }, { "epoch": 2.05, "learning_rate": 3.888888888888889e-05, "loss": 0.4064, "step": 180 }, { "epoch": 2.07, "learning_rate": 3.7962962962962964e-05, "loss": 0.5554, "step": 190 }, { "epoch": 2.08, "learning_rate": 3.7037037037037037e-05, "loss": 0.3856, "step": 200 }, { "epoch": 2.1, "learning_rate": 3.611111111111111e-05, "loss": 0.3292, "step": 210 }, { "epoch": 2.12, "learning_rate": 3.518518518518519e-05, "loss": 0.4154, "step": 220 }, { "epoch": 2.12, "eval_accuracy": 0.8571428571428571, "eval_loss": 0.4632211923599243, "eval_runtime": 14.7059, "eval_samples_per_second": 4.76, "eval_steps_per_second": 1.224, "step": 225 }, { "epoch": 3.01, "learning_rate": 3.425925925925926e-05, "loss": 0.3805, "step": 230 }, { "epoch": 3.02, "learning_rate": 3.3333333333333335e-05, "loss": 0.1538, "step": 240 }, { "epoch": 3.04, "learning_rate": 3.240740740740741e-05, "loss": 0.1232, "step": 250 }, { "epoch": 3.06, "learning_rate": 3.148148148148148e-05, "loss": 0.2196, "step": 260 }, { "epoch": 3.08, "learning_rate": 3.055555555555556e-05, "loss": 0.3353, "step": 270 }, { "epoch": 3.09, "learning_rate": 2.962962962962963e-05, "loss": 0.2452, "step": 280 }, { "epoch": 3.11, "learning_rate": 2.8703703703703706e-05, "loss": 0.1609, "step": 290 }, { "epoch": 3.12, "learning_rate": 2.777777777777778e-05, "loss": 0.4116, "step": 300 }, { "epoch": 3.12, "eval_accuracy": 0.8571428571428571, "eval_loss": 0.5566033124923706, "eval_runtime": 16.6659, "eval_samples_per_second": 4.2, "eval_steps_per_second": 1.08, "step": 300 }, { "epoch": 4.02, "learning_rate": 2.6851851851851855e-05, "loss": 0.2113, "step": 310 }, { "epoch": 4.03, "learning_rate": 2.5925925925925925e-05, "loss": 0.3154, "step": 320 }, { "epoch": 4.05, "learning_rate": 2.5e-05, "loss": 0.0308, "step": 330 }, { "epoch": 4.07, "learning_rate": 2.4074074074074074e-05, "loss": 0.026, "step": 340 }, { "epoch": 4.08, "learning_rate": 2.314814814814815e-05, "loss": 0.0414, "step": 350 }, { "epoch": 4.1, "learning_rate": 2.2222222222222223e-05, "loss": 0.0476, "step": 360 }, { "epoch": 4.12, "learning_rate": 2.1296296296296296e-05, "loss": 0.021, "step": 370 }, { "epoch": 4.12, "eval_accuracy": 0.9714285714285714, "eval_loss": 0.17257067561149597, "eval_runtime": 17.1058, "eval_samples_per_second": 4.092, "eval_steps_per_second": 1.052, "step": 375 }, { "epoch": 5.01, "learning_rate": 2.037037037037037e-05, "loss": 0.189, "step": 380 }, { "epoch": 5.03, "learning_rate": 1.9444444444444445e-05, "loss": 0.2123, "step": 390 }, { "epoch": 5.04, "learning_rate": 1.8518518518518518e-05, "loss": 0.0096, "step": 400 }, { "epoch": 5.06, "learning_rate": 1.7592592592592595e-05, "loss": 0.0385, "step": 410 }, { "epoch": 5.08, "learning_rate": 1.6666666666666667e-05, "loss": 0.1287, "step": 420 }, { "epoch": 5.09, "learning_rate": 1.574074074074074e-05, "loss": 0.015, "step": 430 }, { "epoch": 5.11, "learning_rate": 1.4814814814814815e-05, "loss": 0.0473, "step": 440 }, { "epoch": 5.12, "learning_rate": 1.388888888888889e-05, "loss": 0.0084, "step": 450 }, { "epoch": 5.12, "eval_accuracy": 0.9714285714285714, "eval_loss": 0.10275492072105408, "eval_runtime": 17.0163, "eval_samples_per_second": 4.114, "eval_steps_per_second": 1.058, "step": 450 }, { "epoch": 6.02, "learning_rate": 1.2962962962962962e-05, "loss": 0.108, "step": 460 }, { "epoch": 6.03, "learning_rate": 1.2037037037037037e-05, "loss": 0.0063, "step": 470 }, { "epoch": 6.05, "learning_rate": 1.1111111111111112e-05, "loss": 0.0816, "step": 480 }, { "epoch": 6.07, "learning_rate": 1.0185185185185185e-05, "loss": 0.1057, "step": 490 }, { "epoch": 6.08, "learning_rate": 9.259259259259259e-06, "loss": 0.0057, "step": 500 }, { "epoch": 6.1, "learning_rate": 8.333333333333334e-06, "loss": 0.006, "step": 510 }, { "epoch": 6.12, "learning_rate": 7.4074074074074075e-06, "loss": 0.0056, "step": 520 }, { "epoch": 6.12, "eval_accuracy": 0.9857142857142858, "eval_loss": 0.035175032913684845, "eval_runtime": 16.5696, "eval_samples_per_second": 4.225, "eval_steps_per_second": 1.086, "step": 525 }, { "epoch": 7.01, "learning_rate": 6.481481481481481e-06, "loss": 0.0052, "step": 530 }, { "epoch": 7.03, "learning_rate": 5.555555555555556e-06, "loss": 0.0054, "step": 540 }, { "epoch": 7.04, "learning_rate": 4.6296296296296296e-06, "loss": 0.0049, "step": 550 }, { "epoch": 7.06, "learning_rate": 3.7037037037037037e-06, "loss": 0.0369, "step": 560 }, { "epoch": 7.08, "learning_rate": 2.777777777777778e-06, "loss": 0.139, "step": 570 }, { "epoch": 7.09, "learning_rate": 1.8518518518518519e-06, "loss": 0.0054, "step": 580 }, { "epoch": 7.11, "learning_rate": 9.259259259259259e-07, "loss": 0.008, "step": 590 }, { "epoch": 7.12, "learning_rate": 0.0, "loss": 0.0051, "step": 600 }, { "epoch": 7.12, "eval_accuracy": 0.9714285714285714, "eval_loss": 0.04347994923591614, "eval_runtime": 16.6367, "eval_samples_per_second": 4.208, "eval_steps_per_second": 1.082, "step": 600 }, { "epoch": 7.12, "step": 600, "total_flos": 2.990768376250368e+18, "train_loss": 0.5285309781382481, "train_runtime": 1415.9208, "train_samples_per_second": 1.695, "train_steps_per_second": 0.424 }, { "epoch": 7.12, "eval_accuracy": 0.9548387096774194, "eval_loss": 0.16580888628959656, "eval_runtime": 38.7083, "eval_samples_per_second": 4.004, "eval_steps_per_second": 1.008, "step": 600 }, { "epoch": 7.12, "eval_accuracy": 0.9548387096774194, "eval_loss": 0.16578371822834015, "eval_runtime": 38.5267, "eval_samples_per_second": 4.023, "eval_steps_per_second": 1.012, "step": 600 } ], "max_steps": 600, "num_train_epochs": 9223372036854775807, "total_flos": 2.990768376250368e+18, "trial_name": null, "trial_params": null }