Files
OpenRS-GRPO/train_phase1_results.json

8 lines
200 B
JSON
Raw Permalink Normal View History

{
"total_flos": 0.0,
"train_loss": 0.04891544524503115,
"train_runtime": 53118.1614,
"train_samples": 7000,
"train_samples_per_second": 0.339,
"train_steps_per_second": 0.005
}