| { | |
| "epoch": 2.999098751126561, | |
| "eval_kl/chosen": 0.0006644949316978455, | |
| "eval_kl/rejected": -0.00017435444169677794, | |
| "eval_logits/chosen": -1.1953125, | |
| "eval_logits/rejected": -1.1640625, | |
| "eval_logps/chosen": -328.0, | |
| "eval_logps/rejected": -300.0, | |
| "eval_loss": 1.0569921731948853, | |
| "eval_nll_loss": 1.0703125, | |
| "eval_ref_probs/chosen": 0.0038271306548267603, | |
| "eval_ref_probs/rejected": 0.0034569373819977045, | |
| "eval_rejected_term/max": -0.00017435444169677794, | |
| "eval_rejected_term/min": -0.00017435444169677794, | |
| "eval_rewards/accuracies": 0.37599998712539673, | |
| "eval_rewards/chosen": -0.5859375, | |
| "eval_rewards/margins": -0.2275390625, | |
| "eval_rewards/rejected": -0.357421875, | |
| "eval_runtime": 25.5229, | |
| "eval_samples_per_second": 39.181, | |
| "eval_steps_per_second": 9.795 | |
| } |