{
    "epoch": 0.999244142101285,
    "eval_epsilon_dpo/beta": 0.003164831316098571,
    "eval_epsilon_dpo/beta_margin_grad_mean": -0.4500769376754761,
    "eval_epsilon_dpo/beta_margin_grad_std": 0.08875461667776108,
    "eval_epsilon_dpo/beta_margin_mean": 0.20660468935966492,
    "eval_epsilon_dpo/beta_margin_std": 0.3723542392253876,
    "eval_epsilon_dpo/loss_margin_mean": 66.05204772949219,
    "eval_kl/n_epsilon_steps": 0.28917253017425537,
    "eval_kl/p_epsilon_steps": 0.7099471688270569,
    "eval_logits/chosen": -3.157954216003418,
    "eval_logits/rejected": -2.9584426879882812,
    "eval_logps/chosen": -173.7250213623047,
    "eval_logps/ref_chosen": -77.40868377685547,
    "eval_logps/ref_rejected": -73.52816772460938,
    "eval_logps/rejected": -235.89654541015625,
    "eval_loss": 0.6127588152885437,
    "eval_rewards/accuracies": 0.71875,
    "eval_rewards/chosen": -0.30623942613601685,
    "eval_rewards/margins": 0.20660467445850372,
    "eval_rewards/rejected": -0.5128440856933594,
    "eval_runtime": 36.9095,
    "eval_samples": 2303,
    "eval_samples_per_second": 62.396,
    "eval_steps_per_second": 1.951,
    "total_flos": 0.0,
    "train_loss": 1.0336073330418967,
    "train_runtime": 2985.5995,
    "train_samples": 42336,
    "train_samples_per_second": 14.18,
    "train_steps_per_second": 0.221
}