| { | |
| "best_metric": 0.8337155963302753, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-24/checkpoint-527", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 2635, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 17.628143310546875, | |
| "learning_rate": 0.00018567149458255904, | |
| "loss": 1.3968, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8337155963302753, | |
| "eval_loss": 1.3282397985458374, | |
| "eval_runtime": 2.8214, | |
| "eval_samples_per_second": 309.061, | |
| "eval_steps_per_second": 2.481, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 10.100229263305664, | |
| "learning_rate": 0.00013925362093691925, | |
| "loss": 0.6522, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8291284403669725, | |
| "eval_loss": 1.401127576828003, | |
| "eval_runtime": 2.8077, | |
| "eval_samples_per_second": 310.572, | |
| "eval_steps_per_second": 2.493, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": NaN, | |
| "learning_rate": 9.292382674791261e-05, | |
| "loss": 0.4794, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8325688073394495, | |
| "eval_loss": 1.4821361303329468, | |
| "eval_runtime": 2.8127, | |
| "eval_samples_per_second": 310.028, | |
| "eval_steps_per_second": 2.489, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 7.141517162322998, | |
| "learning_rate": 4.650595310227285e-05, | |
| "loss": 0.3848, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8291284403669725, | |
| "eval_loss": 1.4644421339035034, | |
| "eval_runtime": 2.8155, | |
| "eval_samples_per_second": 309.71, | |
| "eval_steps_per_second": 2.486, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 11.939757347106934, | |
| "learning_rate": 8.80794566330925e-08, | |
| "loss": 0.3303, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.823394495412844, | |
| "eval_loss": 1.5062029361724854, | |
| "eval_runtime": 2.8218, | |
| "eval_samples_per_second": 309.024, | |
| "eval_steps_per_second": 2.481, | |
| "step": 2635 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2635, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 40447180739880.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.4205009910693592, | |
| "learning_rate": 0.00023208936822819877, | |
| "num_train_epochs": 5, | |
| "temperature": 21 | |
| } | |
| } | |