Qwen3-8B-math220k-run7 / train_results.json
HectorHe's picture
Model save
b3b6b30 verified
raw
history blame contribute delete
218 Bytes
{
"total_flos": 4.249743249117741e+18,
"train_loss": 0.3027180843966751,
"train_runtime": 80986.1439,
"train_samples": 93733,
"train_samples_per_second": 3.472,
"train_steps_per_second": 0.109
}