all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.244286060333252,
    "eval_logits/rejected": -2.2393975257873535,
    "eval_logps/chosen": -34.03020095825195,
    "eval_logps/rejected": -37.50043869018555,
    "eval_loss": 0.6983639597892761,
    "eval_rewards/accuracies": 0.4987541437149048,
    "eval_rewards/chosen": 0.0026094578206539154,
    "eval_rewards/margins": -0.007098525296896696,
    "eval_rewards/rejected": 0.00970798172056675,
    "eval_runtime": 145.7128,
    "eval_samples": 343,
    "eval_samples_per_second": 2.354,
    "eval_steps_per_second": 0.295,
    "train_loss": 0.6911570282725544,
    "train_runtime": 3257.344,
    "train_samples": 3079,
    "train_samples_per_second": 0.945,
    "train_steps_per_second": 0.118
}