{ "epoch": 0.9994756161510225, "eval_log_odds_chosen": 0.4455896019935608, "eval_log_odds_ratio": -0.6093403697013855, "eval_logits/chosen": -2.943448543548584, "eval_logits/rejected": -2.9263010025024414, "eval_logps/chosen": -0.7282419204711914, "eval_logps/rejected": -0.9978163242340088, "eval_loss": 0.4700748324394226, "eval_nll_loss": 0.4356931149959564, "eval_rewards/accuracies": 0.658730149269104, "eval_rewards/chosen": -0.03641209378838539, "eval_rewards/margins": 0.013478721491992474, "eval_rewards/rejected": -0.04989081248641014, "eval_runtime": 137.5022, "eval_samples": 1994, "eval_samples_per_second": 14.502, "eval_steps_per_second": 0.458 }