{
    "epoch": 1.0,
    "eval_logits/chosen": -3.034407377243042,
    "eval_logits/rejected": -3.069913864135742,
    "eval_logps/chosen": -271.40020751953125,
    "eval_logps/rejected": -175.5244140625,
    "eval_loss": 0.5650191903114319,
    "eval_rewards/accuracies": 0.76953125,
    "eval_rewards/chosen": 0.08157022297382355,
    "eval_rewards/margins": 0.33799096941947937,
    "eval_rewards/rejected": -0.25642073154449463,
    "eval_runtime": 254.1285,
    "eval_samples": 2000,
    "eval_samples_per_second": 7.87,
    "eval_steps_per_second": 0.063,
    "train_loss": 0.5539181610972611,
    "train_runtime": 15602.6148,
    "train_samples": 62064,
    "train_samples_per_second": 3.978,
    "train_steps_per_second": 0.031
}