| { |
| "best_metric": 0.8571428571428571, |
| "best_model_checkpoint": "roberta-based/checkpoint-925", |
| "epoch": 5.0, |
| "eval_steps": 500, |
| "global_step": 925, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 1.0, |
| "grad_norm": 5.441559314727783, |
| "learning_rate": 1.6000000000000003e-05, |
| "loss": 1.7332, |
| "step": 185 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.7997557997557998, |
| "eval_loss": 1.0176318883895874, |
| "eval_runtime": 5.1197, |
| "eval_samples_per_second": 159.97, |
| "eval_steps_per_second": 5.078, |
| "step": 185 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 11.958271980285645, |
| "learning_rate": 1.2e-05, |
| "loss": 0.6767, |
| "step": 370 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.8290598290598291, |
| "eval_loss": 0.7341179847717285, |
| "eval_runtime": 5.1381, |
| "eval_samples_per_second": 159.396, |
| "eval_steps_per_second": 5.06, |
| "step": 370 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 8.717308044433594, |
| "learning_rate": 8.000000000000001e-06, |
| "loss": 0.4469, |
| "step": 555 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.8473748473748474, |
| "eval_loss": 0.6683215498924255, |
| "eval_runtime": 5.2018, |
| "eval_samples_per_second": 157.446, |
| "eval_steps_per_second": 4.998, |
| "step": 555 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 14.097502708435059, |
| "learning_rate": 4.000000000000001e-06, |
| "loss": 0.3435, |
| "step": 740 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.8522588522588522, |
| "eval_loss": 0.6408819556236267, |
| "eval_runtime": 5.2227, |
| "eval_samples_per_second": 156.816, |
| "eval_steps_per_second": 4.978, |
| "step": 740 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 7.630906105041504, |
| "learning_rate": 0.0, |
| "loss": 0.2829, |
| "step": 925 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_accuracy": 0.8571428571428571, |
| "eval_loss": 0.6410605311393738, |
| "eval_runtime": 5.1314, |
| "eval_samples_per_second": 159.606, |
| "eval_steps_per_second": 5.067, |
| "step": 925 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 925, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 500, |
| "total_flos": 2736984690806400.0, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
