all_results.json

{
    "epoch": 1.0,
    "eval_/kl_divergence": 15.022836685180664,
    "eval_/mean_score": 0.8563764095306396,
    "eval_loss": 0.0,
    "eval_runtime": 12.2347,
    "eval_samples_per_second": 8.173,
    "eval_steps_per_second": 0.327
}