| { | |
| "best_metric": 0.9346666666666666, | |
| "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0459/checkpoints/checkpoint-999", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 999, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.7006739377975464, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.697, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8954666666666666, | |
| "eval_loss": 0.3224630355834961, | |
| "eval_runtime": 5.6473, | |
| "eval_samples_per_second": 664.03, | |
| "eval_steps_per_second": 10.447, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5140449404716492, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.1543, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9226666666666666, | |
| "eval_loss": 0.2362789362668991, | |
| "eval_runtime": 5.7908, | |
| "eval_samples_per_second": 647.577, | |
| "eval_steps_per_second": 10.189, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.07932070642709732, | |
| "learning_rate": 0.0, | |
| "loss": 0.0544, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9346666666666666, | |
| "eval_loss": 0.20461750030517578, | |
| "eval_runtime": 6.1284, | |
| "eval_samples_per_second": 611.904, | |
| "eval_steps_per_second": 9.627, | |
| "step": 999 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 999, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.94223961867776e+18, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |