qwen2-0.5b-sft / train_results.json
yinmingzhang's picture
Model save
664c53d verified
raw
history blame
249 Bytes
{
"epoch": 0.9993049349617714,
"total_flos": 106161864966144.0,
"train_loss": 1.5477431688475496,
"train_runtime": 10888.6684,
"train_samples": 207864,
"train_samples_per_second": 22.196,
"train_steps_per_second": 0.116
}