valdesguefa

i have the same problem

@timothylimyl How would you advise me on which instruction-tuned LLM model to choose

@csunny I've changed the torch version according to the following link: https://pytorch.org/get-started/previous-versions/ it doesn't change anything

@csunny how to load the model in 8 bits like in fastchat ?

 @csunny I get this error when I try to run vicuna-13b (Tribbiani) in 32gb ram, 24gb gpu, RTX 3090

@csunny are there any endpoints that can be accessed via an api?

Webui branch