{
    "epoch": 3.0,
    "eval_dpo_losses": 0.6658358573913574,
    "eval_logits/chosen": -2.7120673656463623,
    "eval_logits/rejected": -2.6741116046905518,
    "eval_logps/chosen": -278.5768127441406,
    "eval_logps/rejected": -258.9605712890625,
    "eval_loss": 0.9543376564979553,
    "eval_positive_losses": 2.5736184120178223,
    "eval_rewards/accuracies": 0.6299999952316284,
    "eval_rewards/chosen": 0.06016625836491585,
    "eval_rewards/margins": 0.06398254632949829,
    "eval_rewards/margins_max": 0.34728285670280457,
    "eval_rewards/margins_min": -0.18240024149417877,
    "eval_rewards/margins_std": 0.1765824854373932,
    "eval_rewards/rejected": -0.003816291457042098,
    "eval_runtime": 427.7131,
    "eval_samples": 2000,
    "eval_samples_per_second": 4.676,
    "eval_steps_per_second": 0.292,
    "train_loss": 0.6146632369135467,
    "train_runtime": 13546.83,
    "train_samples": 5678,
    "train_samples_per_second": 1.257,
    "train_steps_per_second": 0.079
}