{ "best_metric": 0.9285714285714286, "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-225", "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": 2.3532, "step": 10 }, { "epoch": 0.07, "learning_rate": 3.3333333333333335e-05, "loss": 2.3027, "step": 20 }, { "epoch": 0.1, "learning_rate": 5e-05, "loss": 2.2167, "step": 30 }, { "epoch": 0.13, "learning_rate": 4.814814814814815e-05, "loss": 2.019, "step": 40 }, { "epoch": 0.17, "learning_rate": 4.62962962962963e-05, "loss": 2.0514, "step": 50 }, { "epoch": 0.2, "learning_rate": 4.4444444444444447e-05, "loss": 1.7294, "step": 60 }, { "epoch": 0.23, "learning_rate": 4.259259259259259e-05, "loss": 1.6613, "step": 70 }, { "epoch": 0.25, "eval_accuracy": 0.5285714285714286, "eval_loss": 1.3506253957748413, "eval_runtime": 6.0005, "eval_samples_per_second": 11.666, "eval_steps_per_second": 3.0, "step": 75 }, { "epoch": 1.02, "learning_rate": 4.074074074074074e-05, "loss": 1.2301, "step": 80 }, { "epoch": 1.05, "learning_rate": 3.888888888888889e-05, "loss": 0.9771, "step": 90 }, { "epoch": 1.08, "learning_rate": 3.7037037037037037e-05, "loss": 1.0313, "step": 100 }, { "epoch": 1.12, "learning_rate": 3.518518518518519e-05, "loss": 0.6111, "step": 110 }, { "epoch": 1.15, "learning_rate": 3.3333333333333335e-05, "loss": 1.0471, "step": 120 }, { "epoch": 1.18, "learning_rate": 3.148148148148148e-05, "loss": 0.8325, "step": 130 }, { "epoch": 1.22, "learning_rate": 2.962962962962963e-05, "loss": 0.8005, "step": 140 }, { "epoch": 1.25, "learning_rate": 2.777777777777778e-05, "loss": 0.4175, "step": 150 }, { "epoch": 1.25, "eval_accuracy": 0.7285714285714285, "eval_loss": 0.5572353601455688, "eval_runtime": 5.7245, "eval_samples_per_second": 12.228, "eval_steps_per_second": 3.144, "step": 150 }, { "epoch": 2.03, "learning_rate": 2.5925925925925925e-05, "loss": 0.5881, "step": 160 }, { "epoch": 2.07, "learning_rate": 2.4074074074074074e-05, "loss": 0.402, "step": 170 }, { "epoch": 2.1, "learning_rate": 2.2222222222222223e-05, "loss": 0.2738, "step": 180 }, { "epoch": 2.13, "learning_rate": 2.037037037037037e-05, "loss": 0.3433, "step": 190 }, { "epoch": 2.17, "learning_rate": 1.8518518518518518e-05, "loss": 0.4863, "step": 200 }, { "epoch": 2.2, "learning_rate": 1.6666666666666667e-05, "loss": 0.2916, "step": 210 }, { "epoch": 2.23, "learning_rate": 1.4814814814814815e-05, "loss": 0.3592, "step": 220 }, { "epoch": 2.25, "eval_accuracy": 0.9285714285714286, "eval_loss": 0.24635276198387146, "eval_runtime": 5.6321, "eval_samples_per_second": 12.429, "eval_steps_per_second": 3.196, "step": 225 }, { "epoch": 3.02, "learning_rate": 1.2962962962962962e-05, "loss": 0.2532, "step": 230 }, { "epoch": 3.05, "learning_rate": 1.1111111111111112e-05, "loss": 0.3399, "step": 240 }, { "epoch": 3.08, "learning_rate": 9.259259259259259e-06, "loss": 0.1234, "step": 250 }, { "epoch": 3.12, "learning_rate": 7.4074074074074075e-06, "loss": 0.2255, "step": 260 }, { "epoch": 3.15, "learning_rate": 5.555555555555556e-06, "loss": 0.2442, "step": 270 }, { "epoch": 3.18, "learning_rate": 3.7037037037037037e-06, "loss": 0.087, "step": 280 }, { "epoch": 3.22, "learning_rate": 1.8518518518518519e-06, "loss": 0.1592, "step": 290 }, { "epoch": 3.25, "learning_rate": 0.0, "loss": 0.1935, "step": 300 }, { "epoch": 3.25, "eval_accuracy": 0.9, "eval_loss": 0.260299414396286, "eval_runtime": 6.0335, "eval_samples_per_second": 11.602, "eval_steps_per_second": 2.983, "step": 300 }, { "epoch": 3.25, "step": 300, "total_flos": 1.495384188125184e+18, "train_loss": 0.8550392069419225, "train_runtime": 221.509, "train_samples_per_second": 5.417, "train_steps_per_second": 1.354 }, { "epoch": 3.25, "eval_accuracy": 0.8387096774193549, "eval_loss": 0.4286609888076782, "eval_runtime": 13.9577, "eval_samples_per_second": 11.105, "eval_steps_per_second": 2.794, "step": 300 }, { "epoch": 3.25, "eval_accuracy": 0.8387096774193549, "eval_loss": 0.4286610186100006, "eval_runtime": 12.6984, "eval_samples_per_second": 12.206, "eval_steps_per_second": 3.071, "step": 300 } ], "max_steps": 300, "num_train_epochs": 9223372036854775807, "total_flos": 1.495384188125184e+18, "trial_name": null, "trial_params": null }