all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": 105.66804504394531,
    "eval_logits/rejected": 90.64244842529297,
    "eval_logps/chosen": -1006.0906982421875,
    "eval_logps/rejected": -1118.860107421875,
    "eval_loss": 0.4697866141796112,
    "eval_rewards/accuracies": 0.7698412537574768,
    "eval_rewards/chosen": -1.0027239322662354,
    "eval_rewards/margins": 1.331162929534912,
    "eval_rewards/rejected": -2.3338871002197266,
    "eval_runtime": 99.7367,
    "eval_samples": 2000,
    "eval_samples_per_second": 20.053,
    "eval_steps_per_second": 0.632,
    "train_loss": 0.5152078939433867,
    "train_runtime": 5321.4428,
    "train_samples": 61135,
    "train_samples_per_second": 11.488,
    "train_steps_per_second": 0.09
}