Qwen2.5-1.5B-Open-R1-GRPO / train_results.json
bruel's picture
Model save
7d163ab verified
{
"total_flos": 0.0,
"train_loss": 379.5353878517854,
"train_runtime": 232403.7931,
"train_samples": 72441,
"train_samples_per_second": 0.312,
"train_steps_per_second": 0.003
}