Text Generation
Transformers
English
llama
Inference Endpoints

talks to its self after answering the question

#8
by nucleardiffusion - opened

https://huggingface.co./eachadea/ggml-vicuna-13b-1.1

Think it is fixed here, what launch parameters are you using / prompts .txt files

I can confirm that version of vicuna 1.1 no longer continues to talk to its self

Wish vicuna continues self talking.Using it in chat mode annoying but in notebook mode its pretty good like a autogpt. Is there a way to enable it?

Sign up or log in to comment