| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.0, |
| "eval_steps": 1999, |
| "global_step": 5997, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.25, |
| "learning_rate": 4.583124895781224e-05, |
| "loss": 0.4645, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 4.1662497915624484e-05, |
| "loss": 0.2725, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 3.749374687343672e-05, |
| "loss": 0.2168, |
| "step": 1500 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.9100734976221357, |
| "eval_loss": 0.35582369565963745, |
| "eval_runtime": 64.7426, |
| "eval_samples_per_second": 214.356, |
| "eval_steps_per_second": 13.407, |
| "step": 1999 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 3.332499583124896e-05, |
| "loss": 0.1988, |
| "step": 2000 |
| }, |
| { |
| "epoch": 1.25, |
| "learning_rate": 2.91562447890612e-05, |
| "loss": 0.1172, |
| "step": 2500 |
| }, |
| { |
| "epoch": 1.5, |
| "learning_rate": 2.4987493746873438e-05, |
| "loss": 0.1262, |
| "step": 3000 |
| }, |
| { |
| "epoch": 1.75, |
| "learning_rate": 2.0818742704685675e-05, |
| "loss": 0.123, |
| "step": 3500 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.9247009655569967, |
| "eval_loss": 0.3541441559791565, |
| "eval_runtime": 65.104, |
| "eval_samples_per_second": 213.167, |
| "eval_steps_per_second": 13.333, |
| "step": 3998 |
| }, |
| { |
| "epoch": 2.0, |
| "learning_rate": 1.6649991662497916e-05, |
| "loss": 0.0974, |
| "step": 4000 |
| }, |
| { |
| "epoch": 2.25, |
| "learning_rate": 1.2481240620310155e-05, |
| "loss": 0.0531, |
| "step": 4500 |
| }, |
| { |
| "epoch": 2.5, |
| "learning_rate": 8.312489578122396e-06, |
| "loss": 0.0514, |
| "step": 5000 |
| }, |
| { |
| "epoch": 2.75, |
| "learning_rate": 4.143738535934634e-06, |
| "loss": 0.0526, |
| "step": 5500 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.9388240380458279, |
| "eval_loss": 0.33043617010116577, |
| "eval_runtime": 65.1232, |
| "eval_samples_per_second": 213.104, |
| "eval_steps_per_second": 13.329, |
| "step": 5997 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 5997, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 500, |
| "total_flos": 2.5078963831327536e+16, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
