{ "epoch": 2.999098751126561, "eval_kl/chosen": 0.000522215326782316, "eval_kl/rejected": 0.0004575403581839055, "eval_logits/chosen": -1.4140625, "eval_logits/rejected": -1.375, "eval_logps/chosen": -322.0, "eval_logps/rejected": -298.0, "eval_loss": 0.711800754070282, "eval_nll_loss": 1.0546875, "eval_ref_probs/chosen": 0.0038271306548267603, "eval_ref_probs/rejected": 0.0034569373819977045, "eval_rejected_term/max": 0.0004575403581839055, "eval_rejected_term/min": 0.0004575403581839055, "eval_rewards/accuracies": 0.5799999833106995, "eval_rewards/chosen": -0.051513671875, "eval_rewards/margins": 0.11962890625, "eval_rewards/rejected": -0.1708984375, "eval_runtime": 25.5394, "eval_samples_per_second": 39.155, "eval_steps_per_second": 9.789 }