| { | |
| "epoch": 0.99, | |
| "eval_logits/chosen": -2.294609308242798, | |
| "eval_logits/rejected": -2.183091878890991, | |
| "eval_logps/chosen": -234.39601135253906, | |
| "eval_logps/rejected": -164.61671447753906, | |
| "eval_loss": 0.6914091110229492, | |
| "eval_rewards/accuracies": 0.6964285969734192, | |
| "eval_rewards/chosen": 0.00801192969083786, | |
| "eval_rewards/margins": 0.003745319088920951, | |
| "eval_rewards/rejected": 0.004266609903424978, | |
| "eval_runtime": 18.9978, | |
| "eval_samples": 204, | |
| "eval_samples_per_second": 10.738, | |
| "eval_steps_per_second": 0.368, | |
| "train_loss": 0.6816894855905087, | |
| "train_runtime": 1070.9433, | |
| "train_samples": 6039, | |
| "train_samples_per_second": 5.639, | |
| "train_steps_per_second": 0.088 | |
| } |