{ "epoch": 0.9984301412872841, "eval_logits/chosen": -0.5977884531021118, "eval_logits/rejected": -0.5622259378433228, "eval_logps/chosen": -401.6270751953125, "eval_logps/rejected": -481.71246337890625, "eval_loss": 0.5371974110603333, "eval_rewards/accuracies": 0.765625, "eval_rewards/chosen": -1.1911048889160156, "eval_rewards/margins": 0.8110275268554688, "eval_rewards/rejected": -2.0021324157714844, "eval_runtime": 186.7025, "eval_samples": 2000, "eval_samples_per_second": 10.712, "eval_steps_per_second": 0.171 }