{
    "epoch": 0.9994242947610823,
    "eval_logits/chosen": -1.1188652515411377,
    "eval_logits/rejected": -0.6186092495918274,
    "eval_logps/chosen": -639.7113037109375,
    "eval_logps/rejected": -952.0885009765625,
    "eval_loss": 0.3328971564769745,
    "eval_rewards/accuracies": 0.828125,
    "eval_rewards/chosen": -3.571798324584961,
    "eval_rewards/margins": 3.05015230178833,
    "eval_rewards/rejected": -6.621951103210449,
    "eval_runtime": 330.4232,
    "eval_samples": 7126,
    "eval_samples_per_second": 21.566,
    "eval_steps_per_second": 0.339,
    "total_flos": 0.0,
    "train_loss": 0.17276417467451316,
    "train_runtime": 14185.6914,
    "train_samples": 111134,
    "train_samples_per_second": 7.834,
    "train_steps_per_second": 0.031
}