| { | |
| "best_metric": 0.4716978371143341, | |
| "best_model_checkpoint": "./vit-lr-0.0001/checkpoint-642", | |
| "epoch": 12.0, | |
| "eval_steps": 500, | |
| "global_step": 3852, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.450945854187012, | |
| "learning_rate": 9.549150281252611e-06, | |
| "loss": 0.682, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8099861303744799, | |
| "eval_f1": 0.798120450266025, | |
| "eval_loss": 0.5192127823829651, | |
| "eval_precision": 0.8034262669589041, | |
| "eval_recall": 0.8099861303744799, | |
| "eval_runtime": 37.3933, | |
| "eval_samples_per_second": 77.126, | |
| "eval_steps_per_second": 9.654, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.513205051422119, | |
| "learning_rate": 6.394955530196143e-05, | |
| "loss": 0.4386, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8283633841886269, | |
| "eval_f1": 0.8213379813460607, | |
| "eval_loss": 0.4716978371143341, | |
| "eval_precision": 0.8230999577196753, | |
| "eval_recall": 0.8283633841886269, | |
| "eval_runtime": 37.886, | |
| "eval_samples_per_second": 76.123, | |
| "eval_steps_per_second": 9.529, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.9458593726158142, | |
| "learning_rate": 6.840622763423387e-05, | |
| "loss": 0.2621, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8349514563106796, | |
| "eval_f1": 0.8201100816778963, | |
| "eval_loss": 0.5676096081733704, | |
| "eval_precision": 0.8262616496556073, | |
| "eval_recall": 0.8349514563106796, | |
| "eval_runtime": 37.2214, | |
| "eval_samples_per_second": 77.482, | |
| "eval_steps_per_second": 9.699, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 7.709182262420654, | |
| "learning_rate": 6.9628986498044885e-06, | |
| "loss": 0.1891, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8495145631067961, | |
| "eval_f1": 0.8436304900485414, | |
| "eval_loss": 0.5022268891334534, | |
| "eval_precision": 0.8555501567834226, | |
| "eval_recall": 0.8495145631067961, | |
| "eval_runtime": 37.4592, | |
| "eval_samples_per_second": 76.99, | |
| "eval_steps_per_second": 9.637, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.030552733689546585, | |
| "learning_rate": 9.960573506570156e-05, | |
| "loss": 0.1052, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8398058252427184, | |
| "eval_f1": 0.8357173097291434, | |
| "eval_loss": 0.6660671830177307, | |
| "eval_precision": 0.8482768361910064, | |
| "eval_recall": 0.8398058252427184, | |
| "eval_runtime": 39.2791, | |
| "eval_samples_per_second": 73.423, | |
| "eval_steps_per_second": 9.191, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.1932728290557861, | |
| "learning_rate": 1.4644660940675861e-05, | |
| "loss": 0.0785, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8592233009708737, | |
| "eval_f1": 0.8574256259898136, | |
| "eval_loss": 0.5570405125617981, | |
| "eval_precision": 0.8573736727029027, | |
| "eval_recall": 0.8592233009708737, | |
| "eval_runtime": 37.1007, | |
| "eval_samples_per_second": 77.734, | |
| "eval_steps_per_second": 9.73, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 17.325477600097656, | |
| "learning_rate": 5.626666167820289e-05, | |
| "loss": 0.0481, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8547156726768377, | |
| "eval_f1": 0.8523115263952025, | |
| "eval_loss": 0.6496189832687378, | |
| "eval_precision": 0.8547876753873573, | |
| "eval_recall": 0.8547156726768377, | |
| "eval_runtime": 36.7758, | |
| "eval_samples_per_second": 78.421, | |
| "eval_steps_per_second": 9.816, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6180940270423889, | |
| "learning_rate": 7.545207078756922e-05, | |
| "loss": 0.0281, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8571428571428571, | |
| "eval_f1": 0.8544924977438809, | |
| "eval_loss": 0.7423234581947327, | |
| "eval_precision": 0.8569592454687236, | |
| "eval_recall": 0.8571428571428571, | |
| "eval_runtime": 37.383, | |
| "eval_samples_per_second": 77.147, | |
| "eval_steps_per_second": 9.657, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.014829314313828945, | |
| "learning_rate": 3.5111757055883184e-06, | |
| "loss": 0.0439, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8540221914008321, | |
| "eval_f1": 0.8518021160859013, | |
| "eval_loss": 0.7677786946296692, | |
| "eval_precision": 0.851139101545856, | |
| "eval_recall": 0.8540221914008321, | |
| "eval_runtime": 37.2162, | |
| "eval_samples_per_second": 77.493, | |
| "eval_steps_per_second": 9.7, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.003476586891338229, | |
| "learning_rate": 9.801468428386933e-05, | |
| "loss": 0.0297, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8467406380027739, | |
| "eval_f1": 0.8464070946689463, | |
| "eval_loss": 0.8647756576538086, | |
| "eval_precision": 0.8478472152581428, | |
| "eval_recall": 0.8467406380027739, | |
| "eval_runtime": 37.1575, | |
| "eval_samples_per_second": 77.616, | |
| "eval_steps_per_second": 9.715, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.09397422522306442, | |
| "learning_rate": 2.0610737385380886e-05, | |
| "loss": 0.037, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8561026352288488, | |
| "eval_f1": 0.8548878398804619, | |
| "eval_loss": 0.7682856321334839, | |
| "eval_precision": 0.8552181984518148, | |
| "eval_recall": 0.8561026352288488, | |
| "eval_runtime": 37.5045, | |
| "eval_samples_per_second": 76.898, | |
| "eval_steps_per_second": 9.626, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.002134101465344429, | |
| "learning_rate": 4.999999999998897e-05, | |
| "loss": 0.0322, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8512482662968099, | |
| "eval_f1": 0.8396847788052201, | |
| "eval_loss": 1.0125652551651, | |
| "eval_precision": 0.8561110421693547, | |
| "eval_recall": 0.8512482662968099, | |
| "eval_runtime": 37.3877, | |
| "eval_samples_per_second": 77.138, | |
| "eval_steps_per_second": 9.656, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "step": 3852, | |
| "total_flos": 4.768760767819088e+18, | |
| "train_loss": 0.16453982563157443, | |
| "train_runtime": 1738.7619, | |
| "train_samples_per_second": 294.922, | |
| "train_steps_per_second": 18.461 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 32100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "total_flos": 4.768760767819088e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |