Qwen2.5-0.5B-Open-R1-GRPO / eval_results.json
zzzch's picture
End of training
fcd214c verified
raw
history blame contribute delete
169 Bytes
{
"eval_loss": 0.007191816810518503,
"eval_runtime": 248.3854,
"eval_samples": 99,
"eval_samples_per_second": 0.399,
"eval_steps_per_second": 0.133
}