{
    "epoch": 0.99,
    "eval_logits/chosen": -3.1902854442596436,
    "eval_logits/rejected": -3.0295298099517822,
    "eval_logps/chosen": -123.95388793945312,
    "eval_logps/rejected": -352.96429443359375,
    "eval_loss": 0.5983969569206238,
    "eval_rewards/accuracies": 0.84375,
    "eval_rewards/chosen": 0.025536777451634407,
    "eval_rewards/margins": 0.23942241072654724,
    "eval_rewards/rejected": -0.2138856202363968,
    "eval_runtime": 22.1643,
    "eval_samples": 241,
    "eval_samples_per_second": 10.873,
    "eval_steps_per_second": 0.361,
    "train_loss": 0.6001577688300092,
    "train_runtime": 543.7858,
    "train_samples": 2973,
    "train_samples_per_second": 5.467,
    "train_steps_per_second": 0.085
}