| { | |
| "epoch": 1.0, | |
| "eval_logits/chosen": 99.07401275634766, | |
| "eval_logits/rejected": 99.07140350341797, | |
| "eval_logps/chosen": -32.35896301269531, | |
| "eval_logps/rejected": -35.895816802978516, | |
| "eval_loss": 2.941925048828125, | |
| "eval_rewards/accuracies": 0.5490033626556396, | |
| "eval_rewards/chosen": 0.03368505463004112, | |
| "eval_rewards/margins": 0.005401855334639549, | |
| "eval_rewards/rejected": 0.028283199295401573, | |
| "eval_runtime": 103.7808, | |
| "eval_samples": 343, | |
| "eval_samples_per_second": 3.305, | |
| "eval_steps_per_second": 0.414, | |
| "train_loss": 2.828607769755574, | |
| "train_runtime": 2555.3286, | |
| "train_samples": 3079, | |
| "train_samples_per_second": 1.205, | |
| "train_steps_per_second": 0.151 | |
| } |