Qwen2.5-1.5B-Open-R1-GRPO / eval_results.json
schwamaths's picture
End of training
cb3c595 verified
raw
history blame contribute delete
167 Bytes
{
"eval_loss": 0.012486881576478481,
"eval_runtime": 79.6968,
"eval_samples": 223,
"eval_samples_per_second": 2.798,
"eval_steps_per_second": 0.1
}