| { | |
| "best_metric": 0.9428571428571428, | |
| "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-150", | |
| "epoch": 1.5, | |
| "eval_steps": 500, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.07965963333845139, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.005, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.08289032429456711, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 0.201, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.2235719859600067, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.5754, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.0644197091460228, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 1.1323, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 42.82566452026367, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.2046, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.10422177612781525, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.2749, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.051237553358078, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.0433, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_accuracy": 0.8857142857142857, | |
| "eval_loss": 0.356876015663147, | |
| "eval_runtime": 12.9523, | |
| "eval_samples_per_second": 5.404, | |
| "eval_steps_per_second": 1.39, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.14869919419288635, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.1337, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.21503913402557373, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.061, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.1664995551109314, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.226, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.13239744305610657, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.1242, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.032610807567834854, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.1647, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.5573199391365051, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.1372, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.3077225685119629, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.0047, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.22845041751861572, | |
| "learning_rate": 0.0, | |
| "loss": 0.0047, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_accuracy": 0.9428571428571428, | |
| "eval_loss": 0.3063296973705292, | |
| "eval_runtime": 12.6147, | |
| "eval_samples_per_second": 5.549, | |
| "eval_steps_per_second": 1.427, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "step": 150, | |
| "total_flos": 7.47692094062592e+17, | |
| "train_loss": 0.2195048264414072, | |
| "train_runtime": 232.8628, | |
| "train_samples_per_second": 2.577, | |
| "train_steps_per_second": 0.644 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_accuracy": 0.8205128205128205, | |
| "eval_loss": 0.6944047212600708, | |
| "eval_runtime": 28.6136, | |
| "eval_samples_per_second": 5.452, | |
| "eval_steps_per_second": 1.363, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_accuracy": 0.8193548387096774, | |
| "eval_loss": 0.6981304287910461, | |
| "eval_runtime": 28.8048, | |
| "eval_samples_per_second": 5.381, | |
| "eval_steps_per_second": 1.354, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 150, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 7.47692094062592e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |