Qwen2.5-1.5B-Open-R1-GRPO / train_results.json
andrewsiah's picture
Model save
a6fbf6b verified
raw
history blame
200 Bytes
{
"total_flos": 0.0,
"train_loss": 0.11728219354807815,
"train_runtime": 26244.8941,
"train_samples": 72441,
"train_samples_per_second": 2.76,
"train_steps_per_second": 0.011
}