Xinyong Guo

How to load ChatGLM in 4-bit?

5

### Describe the bug I want to load this model in 8G VRAM. But it seems that they have different quantize. Shall I copy Tokenizer class ChatGLMTokenizer from "models\THUDM_chatglm3-6btokenization_chatglm.py" into...

This is a great plugin. However, obviously, when I open a new tab, what I need is to create a new tab, not to initiate a search. Hopefully, the corresponding...