| { | |
| "best_metric": 1.0, | |
| "best_model_checkpoint": "vivit-b-16x2-kinetics400-finetuned-ucf101-subset/checkpoint-38", | |
| "epoch": 3.22972972972973, | |
| "eval_steps": 500, | |
| "global_step": 148, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 11.20502758026123, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 2.4387, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 7.179543495178223, | |
| "learning_rate": 4.81203007518797e-05, | |
| "loss": 1.4999, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 5.833919525146484, | |
| "learning_rate": 4.43609022556391e-05, | |
| "loss": 0.7382, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.2207001894712448, | |
| "eval_runtime": 85.8262, | |
| "eval_samples_per_second": 0.431, | |
| "eval_steps_per_second": 0.058, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 2.8914296627044678, | |
| "learning_rate": 4.0601503759398494e-05, | |
| "loss": 0.4191, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.2957464456558228, | |
| "learning_rate": 3.6842105263157895e-05, | |
| "loss": 0.1263, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 7.1389689445495605, | |
| "learning_rate": 3.3082706766917295e-05, | |
| "loss": 0.1227, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.3911808133125305, | |
| "learning_rate": 2.9323308270676693e-05, | |
| "loss": 0.0404, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "eval_accuracy": 0.972972972972973, | |
| "eval_loss": 0.054861173033714294, | |
| "eval_runtime": 86.2035, | |
| "eval_samples_per_second": 0.429, | |
| "eval_steps_per_second": 0.058, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 0.20817312598228455, | |
| "learning_rate": 2.556390977443609e-05, | |
| "loss": 0.0473, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 0.41731318831443787, | |
| "learning_rate": 2.1804511278195487e-05, | |
| "loss": 0.0191, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.8074580430984497, | |
| "learning_rate": 1.8045112781954888e-05, | |
| "loss": 0.0182, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 0.2408752292394638, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.0115, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.028172966092824936, | |
| "eval_runtime": 85.2593, | |
| "eval_samples_per_second": 0.434, | |
| "eval_steps_per_second": 0.059, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 0.14038172364234924, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 0.0097, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 0.15160618722438812, | |
| "learning_rate": 6.766917293233083e-06, | |
| "loss": 0.0085, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 0.0902809351682663, | |
| "learning_rate": 3.007518796992481e-06, | |
| "loss": 0.0172, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.0235783401876688, | |
| "eval_runtime": 86.3082, | |
| "eval_samples_per_second": 0.429, | |
| "eval_steps_per_second": 0.058, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "step": 148, | |
| "total_flos": 3.0029279215406285e+18, | |
| "train_loss": 0.3731363031107026, | |
| "train_runtime": 6285.1813, | |
| "train_samples_per_second": 0.188, | |
| "train_steps_per_second": 0.024 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.2172325849533081, | |
| "eval_runtime": 200.1151, | |
| "eval_samples_per_second": 0.435, | |
| "eval_steps_per_second": 0.055, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.2172325998544693, | |
| "eval_runtime": 199.0471, | |
| "eval_samples_per_second": 0.437, | |
| "eval_steps_per_second": 0.055, | |
| "step": 148 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 148, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 3.0029279215406285e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |