Qwen2.5-Math-1.5B-GRPO / all_results.json
YYT-t's picture
Model save
3a60c85 verified
raw
history blame contribute delete
189 Bytes
{
"total_flos": 3.216438415536947e+16,
"train_loss": 0.6463175455729167,
"train_runtime": 630.4846,
"train_samples_per_second": 15.663,
"train_steps_per_second": 0.119
}