Files
Qwen3-8B-ODA-Math-460k/all_results.json

8 lines
211 B
JSON
Raw Normal View History

{
"epoch": 3.0,
"total_flos": 1.6970304064115966e+19,
"train_loss": 0.20868314632318158,
"train_runtime": 603947.8558,
"train_samples_per_second": 0.459,
"train_steps_per_second": 0.014
}