You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
yea seems like 2,048 tokens from a skim
might be harder to do, but if we can do this, the rest would work for sure.
Also, the cost is close to 0 and high speed hopefully as 0 networking, dedicated local resources, so running unlimited number of smaller queries
It's aspirational.
IDK if something else comes along that is open to running locally
The goal is for running local something so as not to send source code to central companies
Don't care what the final model and such would be as long as it would be able to do the job.
Why
Users don't want to send their code to OPENAI
What
Allow users to connect to different models. Maybe asking users to host the model and provide an API?
For example
Support gpt4all
https://github.com/nomic-ai/gpt4all
https://python.langchain.com/en/latest/modules/models/llms/integrations/gpt4all.html
For use on "secret"/proprietary code with fewer security concerns.
The text was updated successfully, but these errors were encountered: