Files
DReP-SFT-Qwen3-4B-Thinking-…/all_results.json

8 lines
209 B
JSON
Raw Normal View History

{
"epoch": 4.0,
"total_flos": 2.006101437057283e+19,
"train_loss": 0.5864963867131946,
"train_runtime": 153073.7159,
"train_samples_per_second": 0.687,
"train_steps_per_second": 0.005
}