| { | |
| "best_metric": 1.8800272941589355, | |
| "best_model_checkpoint": "flower_groups_image_detection/checkpoint-7755", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 7755, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 5.792048454284668, | |
| "learning_rate": 9.415963659961064e-07, | |
| "loss": 2.0867, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 3.73576283454895, | |
| "learning_rate": 8.767034393251135e-07, | |
| "loss": 2.0233, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 3.3685128688812256, | |
| "learning_rate": 8.118105126541207e-07, | |
| "loss": 1.9642, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7352745524915336, | |
| "eval_loss": 2.0397422313690186, | |
| "eval_runtime": 331.1652, | |
| "eval_samples_per_second": 99.866, | |
| "eval_steps_per_second": 12.483, | |
| "step": 1551 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 3.8721961975097656, | |
| "learning_rate": 7.469175859831278e-07, | |
| "loss": 1.9523, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 5.311251163482666, | |
| "learning_rate": 6.820246593121349e-07, | |
| "loss": 1.939, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 3.2061917781829834, | |
| "learning_rate": 6.17131732641142e-07, | |
| "loss": 1.8962, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7481555394291244, | |
| "eval_loss": 1.9691122770309448, | |
| "eval_runtime": 321.3062, | |
| "eval_samples_per_second": 102.93, | |
| "eval_steps_per_second": 12.866, | |
| "step": 3102 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 6.1755170822143555, | |
| "learning_rate": 5.522388059701492e-07, | |
| "loss": 1.8795, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 7.479930877685547, | |
| "learning_rate": 4.873458792991564e-07, | |
| "loss": 1.8772, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 3.788714647293091, | |
| "learning_rate": 4.224529526281635e-07, | |
| "loss": 1.8427, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7578313981615868, | |
| "eval_loss": 1.9191893339157104, | |
| "eval_runtime": 323.2055, | |
| "eval_samples_per_second": 102.325, | |
| "eval_steps_per_second": 12.791, | |
| "step": 4653 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 5.707278251647949, | |
| "learning_rate": 3.5756002595717065e-07, | |
| "loss": 1.8256, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 3.444243907928467, | |
| "learning_rate": 2.9266709928617776e-07, | |
| "loss": 1.8318, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 3.4218833446502686, | |
| "learning_rate": 2.2777417261518494e-07, | |
| "loss": 1.8064, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7642416545718432, | |
| "eval_loss": 1.8897231817245483, | |
| "eval_runtime": 325.0959, | |
| "eval_samples_per_second": 101.73, | |
| "eval_steps_per_second": 12.716, | |
| "step": 6204 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "grad_norm": 3.0800297260284424, | |
| "learning_rate": 1.6288124594419208e-07, | |
| "loss": 1.8, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "grad_norm": 5.477210521697998, | |
| "learning_rate": 9.798831927319921e-08, | |
| "loss": 1.8135, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 5.891938209533691, | |
| "learning_rate": 3.309539260220636e-08, | |
| "loss": 1.7991, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7664791969037252, | |
| "eval_loss": 1.8800272941589355, | |
| "eval_runtime": 319.8009, | |
| "eval_samples_per_second": 103.414, | |
| "eval_steps_per_second": 12.927, | |
| "step": 7755 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 7755, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 1.9243077271251886e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |