Qwen-2.5-7B_Math / train_results.json
Dongwei's picture
Model save
5b18b72 verified
raw
history blame
198 Bytes
{
"total_flos": 0.0,
"train_loss": 0.2598499658503639,
"train_runtime": 9503.7836,
"train_samples": 7500,
"train_samples_per_second": 0.789,
"train_steps_per_second": 0.007
}