ganchengguang/Yoko-7B-Japanese-v1

This model is traned with guanaco dataset. And this model used whole guanaco dataset by 49000 chat samples and 280000 non chat samples.
Improved performance in Chinese and Japanese.
Use the QLoRA to fine-tune the vanilla LLaMA2-7B.
And you can use test.py to test the model.

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.