YuanGongND / ltu

Code, Dataset, and Pretrained Models for Audio and Speech Large Language Model "Listen, Think, and Understand".
389 stars 36 forks source link

Question about vicuna version #28

Closed CleyLyChen closed 7 months ago

CleyLyChen commented 7 months ago

Hi, great work! I have a small question that what version of vicuna you used in ltu and ltu-as? Thanks.

YuanGongND commented 7 months ago

hi there,

It should be v1.1 https://huggingface.co/lmsys/vicuna-7b-v1.1/tree/main.

This work is mostly finished in April 2023, when 1.1 is the latest version. You can check the hash to confirm this.

-Yuan

CleyLyChen commented 7 months ago

OK, thank you for your reply! : )