Qwen2.5-7B-ODA-Math-460k / train_results.json
GX-XinGao's picture
Initial release v1.0
4d227ce verified
raw
history blame contribute delete
210 Bytes
{
"epoch": 3.0,
"total_flos": 1.816985160646656e+16,
"train_loss": 0.20176902750353792,
"train_runtime": 231849.8596,
"train_samples_per_second": 1.195,
"train_steps_per_second": 0.025
}