openthoughts3_100k_qwen25_1b_bsz256_lr8e5_epochs5/train_results.json

{
    "epoch": 5.0,
    "total_flos": 7122204608430080.0,
    "train_loss": 1.0287959775656386,
    "train_runtime": 35452.5316,
    "train_samples_per_second": 14.103,
    "train_steps_per_second": 0.055
}