Qwen-2.5-Math-7B-Max-v3 / train_results.json
chenggong
Model save
0bf609e verified
raw
history blame contribute delete
200 Bytes
{
"total_flos": 0.0,
"train_loss": 0.03207497191133684,
"train_runtime": 41076.4058,
"train_samples": 7500,
"train_samples_per_second": 0.548,
"train_steps_per_second": 0.004
}