| { | |
| "best_metric": 1.3125017881393433, | |
| "best_model_checkpoint": "cat_breed_image_detection/checkpoint-7623", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 7623, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.19677292404565133, | |
| "grad_norm": 7.071441173553467, | |
| "learning_rate": 8.465205334741845e-07, | |
| "loss": 1.2994, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.39354584809130266, | |
| "grad_norm": 9.168732643127441, | |
| "learning_rate": 7.870989040010564e-07, | |
| "loss": 1.3024, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5903187721369539, | |
| "grad_norm": 8.262181282043457, | |
| "learning_rate": 7.276772745279281e-07, | |
| "loss": 1.2652, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7870916961826053, | |
| "grad_norm": 10.966161727905273, | |
| "learning_rate": 6.682556450547999e-07, | |
| "loss": 1.2789, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.9838646202282566, | |
| "grad_norm": 8.47767448425293, | |
| "learning_rate": 6.088340155816718e-07, | |
| "loss": 1.269, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6525409654561559, | |
| "eval_loss": 1.3324801921844482, | |
| "eval_model_preparation_time": 0.0054, | |
| "eval_runtime": 1508.0216, | |
| "eval_samples_per_second": 71.872, | |
| "eval_steps_per_second": 2.246, | |
| "step": 2541 | |
| }, | |
| { | |
| "epoch": 1.1806375442739079, | |
| "grad_norm": 14.689343452453613, | |
| "learning_rate": 5.494123861085435e-07, | |
| "loss": 1.2627, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.3774104683195592, | |
| "grad_norm": 7.966395854949951, | |
| "learning_rate": 4.899907566354153e-07, | |
| "loss": 1.2615, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.5741833923652107, | |
| "grad_norm": 6.865269184112549, | |
| "learning_rate": 4.3056912716228705e-07, | |
| "loss": 1.2636, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.770956316410862, | |
| "grad_norm": 9.544075012207031, | |
| "learning_rate": 3.7114749768915883e-07, | |
| "loss": 1.2484, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.9677292404565132, | |
| "grad_norm": 12.453811645507812, | |
| "learning_rate": 3.117258682160306e-07, | |
| "loss": 1.264, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6554196191319752, | |
| "eval_loss": 1.3183132410049438, | |
| "eval_model_preparation_time": 0.0054, | |
| "eval_runtime": 1205.014, | |
| "eval_samples_per_second": 89.944, | |
| "eval_steps_per_second": 2.811, | |
| "step": 5082 | |
| }, | |
| { | |
| "epoch": 2.1645021645021645, | |
| "grad_norm": 9.422158241271973, | |
| "learning_rate": 2.523042387429024e-07, | |
| "loss": 1.2342, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.3612750885478158, | |
| "grad_norm": 8.495936393737793, | |
| "learning_rate": 1.928826092697742e-07, | |
| "loss": 1.2448, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.558048012593467, | |
| "grad_norm": 8.68142318725586, | |
| "learning_rate": 1.3346097979664598e-07, | |
| "loss": 1.2507, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.7548209366391183, | |
| "grad_norm": 11.802330017089844, | |
| "learning_rate": 7.403935032351777e-08, | |
| "loss": 1.2448, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.9515938606847696, | |
| "grad_norm": 8.9815673828125, | |
| "learning_rate": 1.4617720850389543e-08, | |
| "loss": 1.2538, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6566928697962799, | |
| "eval_loss": 1.3125017881393433, | |
| "eval_model_preparation_time": 0.0054, | |
| "eval_runtime": 1235.4205, | |
| "eval_samples_per_second": 87.73, | |
| "eval_steps_per_second": 2.742, | |
| "step": 7623 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 7623, | |
| "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": 3.781059536914691e+19, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |