Vicuna-based AutoGPT
#3
by
Refugium
- opened
I feel that one of the largest barriers to widespread usage of AutoGPT is the cost of the OpenAI API. I feel that Vicuna, being locally run, would be a good alternative to said API to create a fully local AutoGPT. Could anyone suggest whether this is possible, and if so what changes would need to be made to Vicuna/AutoGPT source code?
Can we use, localy, with Vicuna 4bit GPTQ or Koala ?
thanks
This comment has been hidden