支持batch形式的对话

#60
by curlyfu - opened

— 添加chat_batch方法,可以同时进行多次多伦对话

1683539812923.png

3090,fp16,并行跑100条只要3.2s,当然文本比较短,不过也比循环一百次快多了

Ready to merge
This branch is ready to get merged automatically.

Sign up or log in to comment