Qwen2.5-1.5B-Open-R1-GRPO_Math / train_results.json
Dongwei's picture
Model save
d07cf3c verified
raw
history blame contribute delete
201 Bytes
{
"total_flos": 0.0,
"train_loss": 0.0008235528940029108,
"train_runtime": 9299.7344,
"train_samples": 7500,
"train_samples_per_second": 0.806,
"train_steps_per_second": 0.007
}