Files
Qwen3-4B_Paper_Impact_paten…/train_results.json

8 lines
209 B
JSON
Raw Normal View History

{
"epoch": 1.0,
"total_flos": 3.259472961077248e+17,
"train_loss": 0.08579332347738618,
"train_runtime": 1625.0751,
"train_samples_per_second": 11.273,
"train_steps_per_second": 0.177
}