Qwen2.5-1.5B-Open-R1-Distill / train_results.json
wooferclaw's picture
Model save
edb440f verified
raw
history blame
214 Bytes
{
"total_flos": 2442892589137920.0,
"train_loss": 0.2958939010054985,
"train_runtime": 111352.539,
"train_samples": 93733,
"train_samples_per_second": 1.54,
"train_steps_per_second": 0.385
}