{
    "eval_loss": 1.41169273853302,
    "eval_runtime": 58.9222,
    "eval_samples": 462,
    "eval_samples_per_second": 7.841,
    "perplexity": 4.102894657126741
}