Qwen2.5-1.5B-Open-R1-GRPO / eval_results.json
hwang595's picture
End of training
ec8a30c verified
{
"eval_loss": 0.9750849008560181,
"eval_runtime": 22.6746,
"eval_samples": 99,
"eval_samples_per_second": 4.366,
"eval_steps_per_second": 0.176
}