Files
Qwen3-1.7B-Base-SFT-Tulu3-d…/all_results.json

8 lines
217 B
JSON
Raw Permalink Normal View History

{
"total_flos": 2054605533544448.0,
"train_loss": 0.6844859028735806,
"train_runtime": 36489.0201,
"train_samples": 936509,
"train_samples_per_second": 51.331,
"train_steps_per_second": 0.401
}