{ "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": 2.4741, "step": 10 }, { "epoch": 0.07, "learning_rate": 3.3333333333333335e-05, "loss": 2.2495, "step": 20 }, { "epoch": 0.1, "learning_rate": 5e-05, "loss": 2.2511, "step": 30 }, { "epoch": 0.13, "learning_rate": 4.814814814814815e-05, "loss": 2.1369, "step": 40 }, { "epoch": 0.17, "learning_rate": 4.62962962962963e-05, "loss": 2.0541, "step": 50 }, { "epoch": 0.2, "learning_rate": 4.4444444444444447e-05, "loss": 1.947, "step": 60 }, { "epoch": 0.23, "learning_rate": 4.259259259259259e-05, "loss": 1.6659, "step": 70 }, { "epoch": 0.25, "eval_accuracy": 0.5428571428571428, "eval_loss": 1.3182615041732788, "eval_runtime": 7.1048, "eval_samples_per_second": 9.852, "eval_steps_per_second": 2.533, "step": 75 }, { "epoch": 1.02, "learning_rate": 4.074074074074074e-05, "loss": 1.4359, "step": 80 }, { "epoch": 1.05, "learning_rate": 3.888888888888889e-05, "loss": 1.1439, "step": 90 }, { "epoch": 1.08, "learning_rate": 3.7037037037037037e-05, "loss": 1.1886, "step": 100 }, { "epoch": 1.12, "learning_rate": 3.518518518518519e-05, "loss": 1.005, "step": 110 }, { "epoch": 1.15, "learning_rate": 3.3333333333333335e-05, "loss": 1.0358, "step": 120 }, { "epoch": 1.18, "learning_rate": 3.148148148148148e-05, "loss": 0.725, "step": 130 }, { "epoch": 1.22, "learning_rate": 2.962962962962963e-05, "loss": 0.7268, "step": 140 }, { "epoch": 1.25, "learning_rate": 2.777777777777778e-05, "loss": 0.5699, "step": 150 }, { "epoch": 1.25, "eval_accuracy": 0.8142857142857143, "eval_loss": 0.5787243843078613, "eval_runtime": 7.3698, "eval_samples_per_second": 9.498, "eval_steps_per_second": 2.442, "step": 150 }, { "epoch": 2.03, "learning_rate": 2.5925925925925925e-05, "loss": 0.5818, "step": 160 }, { "epoch": 2.07, "learning_rate": 2.4074074074074074e-05, "loss": 0.5371, "step": 170 }, { "epoch": 2.1, "learning_rate": 2.2222222222222223e-05, "loss": 0.3346, "step": 180 }, { "epoch": 2.13, "learning_rate": 2.037037037037037e-05, "loss": 0.504, "step": 190 }, { "epoch": 2.17, "learning_rate": 1.8518518518518518e-05, "loss": 0.4449, "step": 200 }, { "epoch": 2.2, "learning_rate": 1.6666666666666667e-05, "loss": 0.3952, "step": 210 }, { "epoch": 2.23, "learning_rate": 1.4814814814814815e-05, "loss": 0.3734, "step": 220 }, { "epoch": 2.25, "eval_accuracy": 0.8857142857142857, "eval_loss": 0.4036785364151001, "eval_runtime": 7.7162, "eval_samples_per_second": 9.072, "eval_steps_per_second": 2.333, "step": 225 }, { "epoch": 3.02, "learning_rate": 1.2962962962962962e-05, "loss": 0.2228, "step": 230 }, { "epoch": 3.05, "learning_rate": 1.1111111111111112e-05, "loss": 0.2178, "step": 240 }, { "epoch": 3.08, "learning_rate": 9.259259259259259e-06, "loss": 0.3622, "step": 250 }, { "epoch": 3.12, "learning_rate": 7.4074074074074075e-06, "loss": 0.2362, "step": 260 }, { "epoch": 3.15, "learning_rate": 5.555555555555556e-06, "loss": 0.281, "step": 270 }, { "epoch": 3.18, "learning_rate": 3.7037037037037037e-06, "loss": 0.1473, "step": 280 }, { "epoch": 3.22, "learning_rate": 1.8518518518518519e-06, "loss": 0.287, "step": 290 }, { "epoch": 3.25, "learning_rate": 0.0, "loss": 0.108, "step": 300 }, { "epoch": 3.25, "eval_accuracy": 0.9571428571428572, "eval_loss": 0.285332590341568, "eval_runtime": 6.9201, "eval_samples_per_second": 10.115, "eval_steps_per_second": 2.601, "step": 300 }, { "epoch": 3.25, "step": 300, "total_flos": 1.495384188125184e+18, "train_loss": 0.9214260558287303, "train_runtime": 354.6805, "train_samples_per_second": 3.383, "train_steps_per_second": 0.846 }, { "epoch": 3.25, "eval_accuracy": 0.8580645161290322, "eval_loss": 0.3962918817996979, "eval_runtime": 15.3392, "eval_samples_per_second": 10.105, "eval_steps_per_second": 2.543, "step": 300 }, { "epoch": 3.25, "eval_accuracy": 0.8580645161290322, "eval_loss": 0.3962918221950531, "eval_runtime": 15.5306, "eval_samples_per_second": 9.98, "eval_steps_per_second": 2.511, "step": 300 } ], "max_steps": 300, "num_train_epochs": 9223372036854775807, "total_flos": 1.495384188125184e+18, "trial_name": null, "trial_params": null }