{ "epoch": 2.999098751126561, "eval_kl/chosen": 6.359943654388189e-05, "eval_kl/rejected": -0.00048386561684310436, "eval_logits/chosen": -1.1015625, "eval_logits/rejected": -1.0625, "eval_logps/chosen": -328.0, "eval_logps/rejected": -298.0, "eval_loss": 3.514777421951294, "eval_nll_loss": 1.0703125, "eval_ref_probs/chosen": 0.0038271306548267603, "eval_ref_probs/rejected": 0.0034569373819977045, "eval_rejected_term/max": -0.00048386561684310436, "eval_rejected_term/min": -0.00048386561684310436, "eval_rewards/accuracies": 0.2639999985694885, "eval_rewards/chosen": -0.56640625, "eval_rewards/margins": -0.375, "eval_rewards/rejected": -0.1923828125, "eval_runtime": 25.5462, "eval_samples_per_second": 39.145, "eval_steps_per_second": 9.786 }