model/SL/eval_results.json

{
    "epoch": 1.0,
    "eval_loss": 0.4317435324192047,
    "eval_runtime": 9.6749,
    "eval_samples_per_second": 515.972,
    "eval_steps_per_second": 2.067
}