qwen3-8b-base-sft-hh-helpful-4xh200-batch-64-20260417-214452/train_results.json

{
    "epoch": 1.0,
    "total_flos": 1.000686074462208e+17,
    "train_loss": 1.5741833559199336,
    "train_runtime": 402.4792,
    "train_samples": 43598,
    "train_samples_per_second": 42.723,
    "train_steps_per_second": 0.668
}