| { | |
| "epoch": 0.9991111111111111, | |
| "eval_logits/chosen": 14.29055118560791, | |
| "eval_logits/rejected": 12.499335289001465, | |
| "eval_logps/chosen": -409.5496520996094, | |
| "eval_logps/rejected": -385.20703125, | |
| "eval_loss": 0.6913684010505676, | |
| "eval_rewards/accuracies": 0.591269850730896, | |
| "eval_rewards/chosen": -0.009797219187021255, | |
| "eval_rewards/margins": 0.004927510395646095, | |
| "eval_rewards/rejected": -0.01472472958266735, | |
| "eval_runtime": 89.7641, | |
| "eval_samples": 1000, | |
| "eval_samples_per_second": 11.14, | |
| "eval_steps_per_second": 0.702, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6931305729197438, | |
| "train_runtime": 7518.6256, | |
| "train_samples": 36000, | |
| "train_samples_per_second": 4.788, | |
| "train_steps_per_second": 0.075 | |
| } |