{ "epoch": 2.999098751126561, "eval_kl/chosen": 0.000334249489242211, "eval_kl/rejected": 0.00022987440752331167, "eval_logits/chosen": -1.4296875, "eval_logits/rejected": -1.3984375, "eval_logps/chosen": -322.0, "eval_logps/rejected": -298.0, "eval_loss": 0.688232421875, "eval_nll_loss": 1.046875, "eval_ref_probs/chosen": 0.0038271306548267603, "eval_ref_probs/rejected": 0.0034569373819977045, "eval_rejected_term/max": 0.00022987440752331167, "eval_rejected_term/min": 0.00022987440752331167, "eval_rewards/accuracies": 0.6079999804496765, "eval_rewards/chosen": 0.01031494140625, "eval_rewards/margins": 0.1435546875, "eval_rewards/rejected": -0.1337890625, "eval_runtime": 25.3139, "eval_samples_per_second": 39.504, "eval_steps_per_second": 9.876 }