| { | |
| "best_metric": 0.5148514851485149, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-75", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 375, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.31773510575294495, | |
| "learning_rate": 0.00019498896290528944, | |
| "loss": 0.0964, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5148514851485149, | |
| "eval_loss": 0.077483631670475, | |
| "eval_runtime": 57.3954, | |
| "eval_samples_per_second": 8.799, | |
| "eval_steps_per_second": 0.07, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.29849082231521606, | |
| "learning_rate": 0.00014624172217896707, | |
| "loss": 0.0774, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5069306930693069, | |
| "eval_loss": 0.07408903539180756, | |
| "eval_runtime": 57.3815, | |
| "eval_samples_per_second": 8.801, | |
| "eval_steps_per_second": 0.07, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.17056742310523987, | |
| "learning_rate": 9.749448145264472e-05, | |
| "loss": 0.0761, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5089108910891089, | |
| "eval_loss": 0.07352373003959656, | |
| "eval_runtime": 56.7092, | |
| "eval_samples_per_second": 8.905, | |
| "eval_steps_per_second": 0.071, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.1569105088710785, | |
| "learning_rate": 4.874724072632236e-05, | |
| "loss": 0.0757, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5089108910891089, | |
| "eval_loss": 0.07332208752632141, | |
| "eval_runtime": 56.6672, | |
| "eval_samples_per_second": 8.912, | |
| "eval_steps_per_second": 0.071, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.2125726044178009, | |
| "learning_rate": 0.0, | |
| "loss": 0.0752, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5089108910891089, | |
| "eval_loss": 0.07323028892278671, | |
| "eval_runtime": 56.4903, | |
| "eval_samples_per_second": 8.94, | |
| "eval_steps_per_second": 0.071, | |
| "step": 375 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 375, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 3889216415700.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.08009223161090917, | |
| "learning_rate": 0.00024373620363161177, | |
| "num_train_epochs": 5, | |
| "temperature": 24 | |
| } | |
| } | |