all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -0.6171875,
    "eval_logits/rejected": -0.671875,
    "eval_logps/chosen": -2320.0,
    "eval_logps/rejected": -2024.0,
    "eval_loss": 0.6403203010559082,
    "eval_rewards/accuracies": 0.5892857313156128,
    "eval_rewards/chosen": 0.609375,
    "eval_rewards/margins": 0.2001953125,
    "eval_rewards/rejected": 0.41015625,
    "eval_runtime": 85.4736,
    "eval_samples": 2000,
    "eval_samples_per_second": 23.399,
    "eval_steps_per_second": 0.491,
    "train_loss": 0.6502101503246783,
    "train_runtime": 8979.6364,
    "train_samples": 61135,
    "train_samples_per_second": 6.808,
    "train_steps_per_second": 0.227
}