| { | |
| "epoch": 0.9981298423724285, | |
| "eval_logits/chosen": -5.339814186096191, | |
| "eval_logits/rejected": -5.241891384124756, | |
| "eval_logps/chosen": -1.3389328718185425, | |
| "eval_logps/rejected": -1.3919367790222168, | |
| "eval_loss": 1.659356713294983, | |
| "eval_rewards/accuracies": 0.5282257795333862, | |
| "eval_rewards/chosen": -3.347332239151001, | |
| "eval_rewards/margins": 0.13250981271266937, | |
| "eval_rewards/rejected": -3.479841947555542, | |
| "eval_runtime": 29.2769, | |
| "eval_samples": 1961, | |
| "eval_samples_per_second": 66.981, | |
| "eval_steps_per_second": 2.118, | |
| "total_flos": 0.0, | |
| "train_loss": 1.7099600828689476, | |
| "train_runtime": 3104.7014, | |
| "train_samples": 59876, | |
| "train_samples_per_second": 19.286, | |
| "train_steps_per_second": 0.15 | |
| } |