| { | |
| "epoch": 2.999098751126561, | |
| "eval_log_ratio_diff/mean": 0.5859375, | |
| "eval_logits/chosen": -2.671875, | |
| "eval_logits/rejected": -2.625, | |
| "eval_logps/chosen": -1.9765625, | |
| "eval_logps/rejected": -2.78125, | |
| "eval_loss": 4.597711086273193, | |
| "eval_nll_loss": 1.9765625, | |
| "eval_ref_probs/chosen": 0.39426562190055847, | |
| "eval_ref_probs/rejected": 0.3336612582206726, | |
| "eval_rejected_term/max": -0.09478497505187988, | |
| "eval_rejected_term/min": -0.09478497505187988, | |
| "eval_rewards/accuracies": 0.7440000176429749, | |
| "eval_rewards/chosen": -0.0927734375, | |
| "eval_rewards/margins": 0.058837890625, | |
| "eval_rewards/rejected": -0.1513671875, | |
| "eval_runtime": 25.3391, | |
| "eval_samples_per_second": 39.465, | |
| "eval_steps_per_second": 9.866 | |
| } |