| { | |
| "best_metric": 0.8969631236442517, | |
| "best_model_checkpoint": "swin-tiny-patch4-window7-224-MM_Classification/checkpoint-146", | |
| "epoch": 9.846153846153847, | |
| "eval_steps": 500, | |
| "global_step": 160, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 2.110320568084717, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.9712, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.9846153846153847, | |
| "eval_accuracy": 0.7472885032537961, | |
| "eval_loss": 0.5833942890167236, | |
| "eval_runtime": 119.6965, | |
| "eval_samples_per_second": 7.703, | |
| "eval_steps_per_second": 0.067, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 2.4040279388427734, | |
| "learning_rate": 4.8611111111111115e-05, | |
| "loss": 0.6442, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 3.5479025840759277, | |
| "learning_rate": 4.5138888888888894e-05, | |
| "loss": 0.4751, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.9692307692307693, | |
| "eval_accuracy": 0.8644251626898047, | |
| "eval_loss": 0.34649166464805603, | |
| "eval_runtime": 122.7615, | |
| "eval_samples_per_second": 7.51, | |
| "eval_steps_per_second": 0.065, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 2.6174745559692383, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.3677, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.953846153846154, | |
| "eval_accuracy": 0.8785249457700651, | |
| "eval_loss": 0.31836390495300293, | |
| "eval_runtime": 122.4616, | |
| "eval_samples_per_second": 7.529, | |
| "eval_steps_per_second": 0.065, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 3.0105929374694824, | |
| "learning_rate": 3.8194444444444444e-05, | |
| "loss": 0.3472, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 2.121572256088257, | |
| "learning_rate": 3.472222222222222e-05, | |
| "loss": 0.3052, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8828633405639913, | |
| "eval_loss": 0.30051109194755554, | |
| "eval_runtime": 122.6551, | |
| "eval_samples_per_second": 7.517, | |
| "eval_steps_per_second": 0.065, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 2.34260892868042, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.3105, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 1.8295146226882935, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.2935, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 4.984615384615385, | |
| "eval_accuracy": 0.8796095444685467, | |
| "eval_loss": 0.3044392466545105, | |
| "eval_runtime": 121.9892, | |
| "eval_samples_per_second": 7.558, | |
| "eval_steps_per_second": 0.066, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 5.538461538461538, | |
| "grad_norm": 2.232234239578247, | |
| "learning_rate": 2.4305555555555558e-05, | |
| "loss": 0.2811, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 5.969230769230769, | |
| "eval_accuracy": 0.89587852494577, | |
| "eval_loss": 0.28295043110847473, | |
| "eval_runtime": 121.4895, | |
| "eval_samples_per_second": 7.589, | |
| "eval_steps_per_second": 0.066, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 2.462690830230713, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 0.2815, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.769230769230769, | |
| "grad_norm": 3.978851556777954, | |
| "learning_rate": 1.736111111111111e-05, | |
| "loss": 0.2659, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 6.953846153846154, | |
| "eval_accuracy": 0.89587852494577, | |
| "eval_loss": 0.2783919870853424, | |
| "eval_runtime": 120.5663, | |
| "eval_samples_per_second": 7.647, | |
| "eval_steps_per_second": 0.066, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 7.384615384615385, | |
| "grad_norm": 4.286837577819824, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.2649, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 4.650554656982422, | |
| "learning_rate": 1.0416666666666668e-05, | |
| "loss": 0.2584, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8937093275488069, | |
| "eval_loss": 0.282360702753067, | |
| "eval_runtime": 119.2752, | |
| "eval_samples_per_second": 7.73, | |
| "eval_steps_per_second": 0.067, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 8.615384615384615, | |
| "grad_norm": 4.211384296417236, | |
| "learning_rate": 6.944444444444445e-06, | |
| "loss": 0.2623, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 8.984615384615385, | |
| "eval_accuracy": 0.8969631236442517, | |
| "eval_loss": 0.27903664112091064, | |
| "eval_runtime": 125.6315, | |
| "eval_samples_per_second": 7.339, | |
| "eval_steps_per_second": 0.064, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "grad_norm": 2.3360867500305176, | |
| "learning_rate": 3.4722222222222224e-06, | |
| "loss": 0.2617, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 9.846153846153847, | |
| "grad_norm": 2.65038800239563, | |
| "learning_rate": 0.0, | |
| "loss": 0.2487, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 9.846153846153847, | |
| "eval_accuracy": 0.8937093275488069, | |
| "eval_loss": 0.276502400636673, | |
| "eval_runtime": 129.1467, | |
| "eval_samples_per_second": 7.139, | |
| "eval_steps_per_second": 0.062, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 9.846153846153847, | |
| "step": 160, | |
| "total_flos": 2.0293244994235208e+18, | |
| "train_loss": 0.36495583951473237, | |
| "train_runtime": 24166.1649, | |
| "train_samples_per_second": 3.43, | |
| "train_steps_per_second": 0.007 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 160, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "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": 2.0293244994235208e+18, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |