Files
Qwen3-8B-finetuned/train_results.json

8 lines
206 B
JSON
Raw Normal View History

{
"epoch": 1.0,
"total_flos": 5228992760119296.0,
"train_loss": 0.07239722475082117,
"train_runtime": 152754.4145,
"train_samples_per_second": 1.13,
"train_steps_per_second": 0.071
}