| { | |
| "epoch": 3.97, | |
| "eval_logits/chosen": 0.8417125940322876, | |
| "eval_logits/rejected": 0.8990198373794556, | |
| "eval_logps/chosen": -256.497314453125, | |
| "eval_logps/rejected": -233.4380340576172, | |
| "eval_loss": 0.0003914303961209953, | |
| "eval_rewards/accuracies": 0.49950000643730164, | |
| "eval_rewards/chosen": 0.001168824383057654, | |
| "eval_rewards/margins": 0.00016647118900436908, | |
| "eval_rewards/rejected": 0.001002353266812861, | |
| "eval_runtime": 561.9533, | |
| "eval_samples": 2000, | |
| "eval_samples_per_second": 3.559, | |
| "eval_steps_per_second": 0.89, | |
| "train_loss": 0.0003399105843263603, | |
| "train_runtime": 4613.339, | |
| "train_samples": 61135, | |
| "train_samples_per_second": 0.867, | |
| "train_steps_per_second": 0.054 | |
| } |