| { | |
| "best_metric": 0.9026666666666666, | |
| "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0287/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": 45.985721588134766, | |
| "learning_rate": 7.513604728193609e-05, | |
| "loss": 0.8834, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8336, | |
| "eval_loss": 0.5429711937904358, | |
| "eval_runtime": 11.7068, | |
| "eval_samples_per_second": 320.326, | |
| "eval_steps_per_second": 5.04, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 17.94068717956543, | |
| "learning_rate": 2.5272834918389072e-05, | |
| "loss": 0.2216, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8882666666666666, | |
| "eval_loss": 0.36174556612968445, | |
| "eval_runtime": 12.0345, | |
| "eval_samples_per_second": 311.605, | |
| "eval_steps_per_second": 4.903, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.9484797716140747, | |
| "learning_rate": 9.889340658375545e-10, | |
| "loss": 0.0297, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9026666666666666, | |
| "eval_loss": 0.33555924892425537, | |
| "eval_runtime": 11.9673, | |
| "eval_samples_per_second": 313.353, | |
| "eval_steps_per_second": 4.93, | |
| "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 | |
| } | |