[Feature] Add a new model

[cacaview]

I found an AI model on GitHub:

https://github.com/baichuan-inc/Baichuan2

I using llama.cpp convert the model to gguf, but the results aren't ideal and feedback is all wrong.
Used the following parameters: --ctx 4096 --outtype q8_0.

I think this might be my problem, and I want to make it usable.
I want it to be available on gpt4all and add it to the download list.

[cebtenzzre]

This model should already work, including GPU support. Try downloading a Q4_0 GGUF, e.g. the one from here (this is a finetune but it's the same model architecture). Make sure to set the prompt template (including the blank line after):

|Human|: %1
|Bot|: %2