{ "epoch": 2.888888888888889, "eval_logits/chosen": -2.388538360595703, "eval_logits/rejected": -2.4200072288513184, "eval_logps/chosen": -271.4289855957031, "eval_logps/rejected": -239.95298767089844, "eval_loss": 0.63017338514328, "eval_rewards/accuracies": 0.6666666865348816, "eval_rewards/chosen": 1.4230774641036987, "eval_rewards/margins": 1.7166643142700195, "eval_rewards/rejected": -0.2935870587825775, "eval_runtime": 24.5831, "eval_samples_per_second": 15.458, "eval_steps_per_second": 0.244 }