| { | |
| "epoch": 2.999098751126561, | |
| "eval_log_ratio_diff/mean": 0.625, | |
| "eval_logits/chosen": -2.640625, | |
| "eval_logits/rejected": -2.578125, | |
| "eval_logps/chosen": -2.046875, | |
| "eval_logps/rejected": -2.890625, | |
| "eval_loss": 5.372431755065918, | |
| "eval_nll_loss": 2.046875, | |
| "eval_ref_probs/chosen": 0.39426562190055847, | |
| "eval_ref_probs/rejected": 0.3336612582206726, | |
| "eval_rejected_term/max": -0.09389033168554306, | |
| "eval_rejected_term/min": -0.09389033168554306, | |
| "eval_rewards/accuracies": 0.7440000176429749, | |
| "eval_rewards/chosen": -0.099609375, | |
| "eval_rewards/margins": 0.0625, | |
| "eval_rewards/rejected": -0.162109375, | |
| "eval_runtime": 25.3544, | |
| "eval_samples_per_second": 39.441, | |
| "eval_steps_per_second": 9.86 | |
| } |