| { | |
| "best_metric": 0.4597996771335602, | |
| "best_model_checkpoint": "./vit-base-brain-mri\\checkpoint-1440", | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 1440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6097560975609756, | |
| "eval_loss": 0.998555600643158, | |
| "eval_runtime": 4.892, | |
| "eval_samples_per_second": 117.334, | |
| "eval_steps_per_second": 14.718, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 2.622373342514038, | |
| "learning_rate": 0.00027916666666666666, | |
| "loss": 1.098, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7003484320557491, | |
| "eval_loss": 0.8445045948028564, | |
| "eval_runtime": 4.8929, | |
| "eval_samples_per_second": 117.312, | |
| "eval_steps_per_second": 14.715, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 4.336460590362549, | |
| "learning_rate": 0.00025833333333333334, | |
| "loss": 0.7895, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7526132404181185, | |
| "eval_loss": 0.7317853569984436, | |
| "eval_runtime": 4.7479, | |
| "eval_samples_per_second": 120.896, | |
| "eval_steps_per_second": 15.165, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7473867595818815, | |
| "eval_loss": 0.6842443943023682, | |
| "eval_runtime": 4.5716, | |
| "eval_samples_per_second": 125.558, | |
| "eval_steps_per_second": 15.749, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 4.166666666666667, | |
| "grad_norm": 3.367997169494629, | |
| "learning_rate": 0.00023749999999999997, | |
| "loss": 0.6629, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7857142857142857, | |
| "eval_loss": 0.6328176856040955, | |
| "eval_runtime": 4.5619, | |
| "eval_samples_per_second": 125.825, | |
| "eval_steps_per_second": 15.783, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 5.555555555555555, | |
| "grad_norm": 2.0712759494781494, | |
| "learning_rate": 0.00021666666666666666, | |
| "loss": 0.5966, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8101045296167247, | |
| "eval_loss": 0.5956693291664124, | |
| "eval_runtime": 4.6073, | |
| "eval_samples_per_second": 124.585, | |
| "eval_steps_per_second": 15.627, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 6.944444444444445, | |
| "grad_norm": 1.491408348083496, | |
| "learning_rate": 0.00019583333333333331, | |
| "loss": 0.5546, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8118466898954704, | |
| "eval_loss": 0.5646191835403442, | |
| "eval_runtime": 4.5982, | |
| "eval_samples_per_second": 124.83, | |
| "eval_steps_per_second": 15.658, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8048780487804879, | |
| "eval_loss": 0.5646994709968567, | |
| "eval_runtime": 4.6362, | |
| "eval_samples_per_second": 123.809, | |
| "eval_steps_per_second": 15.53, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 1.5481159687042236, | |
| "learning_rate": 0.000175, | |
| "loss": 0.5113, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8275261324041812, | |
| "eval_loss": 0.5340307354927063, | |
| "eval_runtime": 4.6122, | |
| "eval_samples_per_second": 124.453, | |
| "eval_steps_per_second": 15.611, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 9.722222222222221, | |
| "grad_norm": 2.323460817337036, | |
| "learning_rate": 0.00015416666666666663, | |
| "loss": 0.4882, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8327526132404182, | |
| "eval_loss": 0.5189912915229797, | |
| "eval_runtime": 4.6128, | |
| "eval_samples_per_second": 124.435, | |
| "eval_steps_per_second": 15.609, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8327526132404182, | |
| "eval_loss": 0.5197045803070068, | |
| "eval_runtime": 4.6509, | |
| "eval_samples_per_second": 123.417, | |
| "eval_steps_per_second": 15.481, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 11.11111111111111, | |
| "grad_norm": 2.1213157176971436, | |
| "learning_rate": 0.0001333333333333333, | |
| "loss": 0.4789, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8257839721254355, | |
| "eval_loss": 0.5001842975616455, | |
| "eval_runtime": 4.5564, | |
| "eval_samples_per_second": 125.977, | |
| "eval_steps_per_second": 15.802, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 1.9881811141967773, | |
| "learning_rate": 0.0001125, | |
| "loss": 0.4582, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8310104529616724, | |
| "eval_loss": 0.4956616163253784, | |
| "eval_runtime": 4.5548, | |
| "eval_samples_per_second": 126.02, | |
| "eval_steps_per_second": 15.807, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 13.88888888888889, | |
| "grad_norm": 2.0128438472747803, | |
| "learning_rate": 9.166666666666667e-05, | |
| "loss": 0.4426, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.8310104529616724, | |
| "eval_loss": 0.4820682108402252, | |
| "eval_runtime": 4.6027, | |
| "eval_samples_per_second": 124.708, | |
| "eval_steps_per_second": 15.643, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.8466898954703833, | |
| "eval_loss": 0.4706496000289917, | |
| "eval_runtime": 4.5952, | |
| "eval_samples_per_second": 124.913, | |
| "eval_steps_per_second": 15.669, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 15.277777777777779, | |
| "grad_norm": 1.9610830545425415, | |
| "learning_rate": 7.083333333333332e-05, | |
| "loss": 0.4328, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8153310104529616, | |
| "eval_loss": 0.4820646047592163, | |
| "eval_runtime": 4.6291, | |
| "eval_samples_per_second": 123.999, | |
| "eval_steps_per_second": 15.554, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 16.666666666666668, | |
| "grad_norm": 1.850261926651001, | |
| "learning_rate": 4.9999999999999996e-05, | |
| "loss": 0.432, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.8275261324041812, | |
| "eval_loss": 0.4991794228553772, | |
| "eval_runtime": 4.5987, | |
| "eval_samples_per_second": 124.818, | |
| "eval_steps_per_second": 15.657, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.8344947735191638, | |
| "eval_loss": 0.4799434542655945, | |
| "eval_runtime": 4.621, | |
| "eval_samples_per_second": 124.216, | |
| "eval_steps_per_second": 15.581, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 18.055555555555557, | |
| "grad_norm": 1.6390336751937866, | |
| "learning_rate": 2.9166666666666663e-05, | |
| "loss": 0.4196, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.8310104529616724, | |
| "eval_loss": 0.4837837815284729, | |
| "eval_runtime": 4.5576, | |
| "eval_samples_per_second": 125.943, | |
| "eval_steps_per_second": 15.798, | |
| "step": 1368 | |
| }, | |
| { | |
| "epoch": 19.444444444444443, | |
| "grad_norm": 3.292024612426758, | |
| "learning_rate": 8.333333333333332e-06, | |
| "loss": 0.4287, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.8658536585365854, | |
| "eval_loss": 0.4597996771335602, | |
| "eval_runtime": 4.6098, | |
| "eval_samples_per_second": 124.517, | |
| "eval_steps_per_second": 15.619, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 1440, | |
| "total_flos": 3.558495949305938e+18, | |
| "train_loss": 0.5529726452297634, | |
| "train_runtime": 477.7876, | |
| "train_samples_per_second": 96.11, | |
| "train_steps_per_second": 3.014 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 1440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "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": 3.558495949305938e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |