all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -0.21875108778476715,
    "eval_logits/rejected": -0.22542043030261993,
    "eval_logps/chosen": -2148.42724609375,
    "eval_logps/rejected": -1824.6278076171875,
    "eval_loss": 0.04916780814528465,
    "eval_rewards/accuracies": 0.5699999928474426,
    "eval_rewards/chosen": 0.06797851622104645,
    "eval_rewards/margins": 0.016193805262446404,
    "eval_rewards/rejected": 0.05178471654653549,
    "eval_runtime": 511.5062,
    "eval_samples": 2000,
    "eval_samples_per_second": 3.91,
    "eval_steps_per_second": 0.978,
    "train_loss": 0.05238237046290443,
    "train_runtime": 26355.2814,
    "train_samples": 30567,
    "train_samples_per_second": 1.16,
    "train_steps_per_second": 0.072
}