Yoko-7B-Japanese-v0 / README.md
ganchengguang's picture
Update README.md
f93dde7
|
raw
history blame
460 Bytes
metadata
license: mit

This model is traned with guanaco dataset. Improved performance in Chinese and Japanese. Use the QLoRA to fine-tune the vanilla LLaMA2-7B.

Recommend Generation parameters:

  • temperature: 0.5~0.7
  • top p: 0.65~1.0
  • top k: 30~50
  • repeat penalty: 1.03~1.17

Contribute by Yokohama Nationaly University Mori Lab.