Files
OpenRS-GRPO-S-2/train_results.json

8 lines
200 B
JSON
Raw Permalink Normal View History

{
"total_flos": 0.0,
"train_loss": -1.2733016510141806,
"train_runtime": 18317.2078,
"train_samples": 7000,
"train_samples_per_second": 0.459,
"train_steps_per_second": 0.019
}