{
    "epoch": 0.9991111111111111,
    "eval_logits/chosen": 13.841368675231934,
    "eval_logits/rejected": 12.528715133666992,
    "eval_logps/chosen": -448.14361572265625,
    "eval_logps/rejected": -459.6432189941406,
    "eval_loss": 0.6129743456840515,
    "eval_rewards/accuracies": 0.7539682388305664,
    "eval_rewards/chosen": -0.41938650608062744,
    "eval_rewards/margins": 0.1739136427640915,
    "eval_rewards/rejected": -0.5933001041412354,
    "eval_runtime": 89.7858,
    "eval_samples": 1000,
    "eval_samples_per_second": 11.138,
    "eval_steps_per_second": 0.702,
    "total_flos": 0.0,
    "train_loss": 0.6475430355801701,
    "train_runtime": 7558.4306,
    "train_samples": 36000,
    "train_samples_per_second": 4.763,
    "train_steps_per_second": 0.074
}