{ "best_metric": 0.9571428571428572, "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-300", "epoch": 3.25, "global_step": 300, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03, "learning_rate": 1.6666666666666667e-05, "loss": 2.3844, "step": 10 }, { "epoch": 0.07, "learning_rate": 3.3333333333333335e-05, "loss": 2.3207, "step": 20 }, { "epoch": 0.1, "learning_rate": 5e-05, "loss": 2.2188, "step": 30 }, { "epoch": 0.13, "learning_rate": 4.814814814814815e-05, "loss": 2.0927, "step": 40 }, { "epoch": 0.17, "learning_rate": 4.62962962962963e-05, "loss": 2.0254, "step": 50 }, { "epoch": 0.2, "learning_rate": 4.4444444444444447e-05, "loss": 1.8138, "step": 60 }, { "epoch": 0.23, "learning_rate": 4.259259259259259e-05, "loss": 1.8145, "step": 70 }, { "epoch": 0.25, "eval_accuracy": 0.5142857142857142, "eval_loss": 1.5607458353042603, "eval_runtime": 20.48, "eval_samples_per_second": 3.418, "eval_steps_per_second": 0.879, "step": 75 }, { "epoch": 1.02, "learning_rate": 4.074074074074074e-05, "loss": 1.5163, "step": 80 }, { "epoch": 1.05, "learning_rate": 3.888888888888889e-05, "loss": 1.128, "step": 90 }, { "epoch": 1.08, "learning_rate": 3.7037037037037037e-05, "loss": 1.0866, "step": 100 }, { "epoch": 1.12, "learning_rate": 3.518518518518519e-05, "loss": 0.7664, "step": 110 }, { "epoch": 1.15, "learning_rate": 3.3333333333333335e-05, "loss": 0.8309, "step": 120 }, { "epoch": 1.18, "learning_rate": 3.148148148148148e-05, "loss": 0.578, "step": 130 }, { "epoch": 1.22, "learning_rate": 2.962962962962963e-05, "loss": 0.4809, "step": 140 }, { "epoch": 1.25, "learning_rate": 2.777777777777778e-05, "loss": 0.9552, "step": 150 }, { "epoch": 1.25, "eval_accuracy": 0.8, "eval_loss": 0.5637401342391968, "eval_runtime": 21.3373, "eval_samples_per_second": 3.281, "eval_steps_per_second": 0.844, "step": 150 }, { "epoch": 2.03, "learning_rate": 2.5925925925925925e-05, "loss": 0.3284, "step": 160 }, { "epoch": 2.07, "learning_rate": 2.4074074074074074e-05, "loss": 0.4275, "step": 170 }, { "epoch": 2.1, "learning_rate": 2.2222222222222223e-05, "loss": 0.3179, "step": 180 }, { "epoch": 2.13, "learning_rate": 2.037037037037037e-05, "loss": 0.4524, "step": 190 }, { "epoch": 2.17, "learning_rate": 1.8518518518518518e-05, "loss": 0.5079, "step": 200 }, { "epoch": 2.2, "learning_rate": 1.6666666666666667e-05, "loss": 0.1765, "step": 210 }, { "epoch": 2.23, "learning_rate": 1.4814814814814815e-05, "loss": 0.3072, "step": 220 }, { "epoch": 2.25, "eval_accuracy": 0.9285714285714286, "eval_loss": 0.2845524251461029, "eval_runtime": 20.9274, "eval_samples_per_second": 3.345, "eval_steps_per_second": 0.86, "step": 225 }, { "epoch": 3.02, "learning_rate": 1.2962962962962962e-05, "loss": 0.2264, "step": 230 }, { "epoch": 3.05, "learning_rate": 1.1111111111111112e-05, "loss": 0.1006, "step": 240 }, { "epoch": 3.08, "learning_rate": 9.259259259259259e-06, "loss": 0.1124, "step": 250 }, { "epoch": 3.12, "learning_rate": 7.4074074074074075e-06, "loss": 0.1111, "step": 260 }, { "epoch": 3.15, "learning_rate": 5.555555555555556e-06, "loss": 0.1377, "step": 270 }, { "epoch": 3.18, "learning_rate": 3.7037037037037037e-06, "loss": 0.196, "step": 280 }, { "epoch": 3.22, "learning_rate": 1.8518518518518519e-06, "loss": 0.2782, "step": 290 }, { "epoch": 3.25, "learning_rate": 0.0, "loss": 0.2755, "step": 300 }, { "epoch": 3.25, "eval_accuracy": 0.9571428571428572, "eval_loss": 0.18497656285762787, "eval_runtime": 21.3698, "eval_samples_per_second": 3.276, "eval_steps_per_second": 0.842, "step": 300 }, { "epoch": 3.25, "step": 300, "total_flos": 1.495384188125184e+18, "train_loss": 0.8656054441134134, "train_runtime": 938.1828, "train_samples_per_second": 1.279, "train_steps_per_second": 0.32 }, { "epoch": 3.25, "eval_accuracy": 0.9225806451612903, "eval_loss": 0.2732711136341095, "eval_runtime": 47.353, "eval_samples_per_second": 3.273, "eval_steps_per_second": 0.824, "step": 300 }, { "epoch": 3.25, "eval_accuracy": 0.9225806451612903, "eval_loss": 0.2732711136341095, "eval_runtime": 47.0808, "eval_samples_per_second": 3.292, "eval_steps_per_second": 0.828, "step": 300 } ], "max_steps": 300, "num_train_epochs": 9223372036854775807, "total_flos": 1.495384188125184e+18, "trial_name": null, "trial_params": null }