Files
ronpo-qwen3-8b-fair-inpo-av…/train_results.json

9 lines
232 B
JSON
Raw Permalink Normal View History

{
"epoch": 0.8599068434252956,
"total_flos": 0.0,
"train_loss": 24.184869588216145,
"train_runtime": 7109.0777,
"train_samples": 16746,
"train_samples_per_second": 2.026,
"train_steps_per_second": 0.127
}