{ "epoch": 2.999098751126561, "eval_kl/chosen": 0.00022940408962313086, "eval_kl/rejected": 0.00017510981706436723, "eval_logits/chosen": -1.4296875, "eval_logits/rejected": -1.3984375, "eval_logps/chosen": -322.0, "eval_logps/rejected": -298.0, "eval_loss": 0.6857578158378601, "eval_nll_loss": 1.046875, "eval_ref_probs/chosen": 0.0038271306548267603, "eval_ref_probs/rejected": 0.0034569373819977045, "eval_rejected_term/max": 0.00017510981706436723, "eval_rejected_term/min": 0.00017510981706436723, "eval_rewards/accuracies": 0.6200000047683716, "eval_rewards/chosen": 0.006134033203125, "eval_rewards/margins": 0.1357421875, "eval_rewards/rejected": -0.1298828125, "eval_runtime": 25.2672, "eval_samples_per_second": 39.577, "eval_steps_per_second": 9.894 }