{ "epoch": 1.0, "eval_logits/chosen": 0.5000983476638794, "eval_logits/rejected": 0.7595670819282532, "eval_logps/chosen": -373.40216064453125, "eval_logps/rejected": -457.4398498535156, "eval_loss": 0.5325239300727844, "eval_rewards/accuracies": 0.765625, "eval_rewards/chosen": -1.2324851751327515, "eval_rewards/margins": 0.8239741921424866, "eval_rewards/rejected": -2.056459426879883, "eval_runtime": 99.4029, "eval_samples": 2000, "eval_samples_per_second": 20.12, "eval_steps_per_second": 0.322 }