| { | |
| "best_metric": 0.0647471621632576, | |
| "best_model_checkpoint": "asl_alphabet_image_detection/checkpoint-6525", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 6525, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 9.305019305019305e-06, | |
| "loss": 2.5567, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 8.532818532818533e-06, | |
| "loss": 1.116, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 7.760617760617761e-06, | |
| "loss": 0.5592, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 6.988416988416989e-06, | |
| "loss": 0.3304, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9991954022988506, | |
| "eval_loss": 0.2387184053659439, | |
| "eval_runtime": 184.9784, | |
| "eval_samples_per_second": 94.065, | |
| "eval_steps_per_second": 11.758, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 6.2162162162162164e-06, | |
| "loss": 0.2285, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 5.444015444015444e-06, | |
| "loss": 0.1713, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 4.671814671814672e-06, | |
| "loss": 0.1359, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "learning_rate": 3.8996138996139e-06, | |
| "loss": 0.1108, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9997126436781609, | |
| "eval_loss": 0.09148428589105606, | |
| "eval_runtime": 182.2169, | |
| "eval_samples_per_second": 95.491, | |
| "eval_steps_per_second": 11.936, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 3.1274131274131274e-06, | |
| "loss": 0.0934, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 2.3552123552123553e-06, | |
| "loss": 0.0813, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 1.583011583011583e-06, | |
| "loss": 0.0733, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "learning_rate": 8.108108108108109e-07, | |
| "loss": 0.0679, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "learning_rate": 3.861003861003861e-08, | |
| "loss": 0.0647, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9997701149425288, | |
| "eval_loss": 0.0647471621632576, | |
| "eval_runtime": 184.0798, | |
| "eval_samples_per_second": 94.524, | |
| "eval_steps_per_second": 11.816, | |
| "step": 6525 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6525, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 1.6184242948762829e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |