Files
full_sft_qwen2-5_7b_openr1_…/train_results.json

8 lines
202 B
JSON
Raw Normal View History

{
"epoch": 3.0,
"total_flos": 74683443707904.0,
"train_loss": 0.3264822006225586,
"train_runtime": 5153.2999,
"train_samples_per_second": 1.746,
"train_steps_per_second": 0.146
}