| { | |
| "epoch": 1.0, | |
| "eval_logits/chosen": 13.824411392211914, | |
| "eval_logits/rejected": 13.813151359558105, | |
| "eval_logps/chosen": -13722.0166015625, | |
| "eval_logps/rejected": -11596.5400390625, | |
| "eval_loss": 0.011624496430158615, | |
| "eval_pred_label": 14285.83984375, | |
| "eval_rewards/accuracies": 0.4740000069141388, | |
| "eval_rewards/chosen": -1343.776123046875, | |
| "eval_rewards/margins": -210.05210876464844, | |
| "eval_rewards/rejected": -1133.72412109375, | |
| "eval_runtime": 450.0493, | |
| "eval_samples": 2000, | |
| "eval_samples_per_second": 4.444, | |
| "eval_steps_per_second": 0.278, | |
| "eval_use_label": 1746.1600341796875, | |
| "train_loss": 0.08065580570807007, | |
| "train_runtime": 25025.0638, | |
| "train_samples": 61135, | |
| "train_samples_per_second": 2.443, | |
| "train_steps_per_second": 0.038 | |
| } |