Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

✨ Support few-shot classification #247

Open
roma-glushko opened this issue May 6, 2024 · 1 comment
Open

✨ Support few-shot classification #247

roma-glushko opened this issue May 6, 2024 · 1 comment
Labels
area:config model:classifier LLM-based few-shot classiers priority:mid-term The feature is on our radars size:large type:enhancement New feature or request type:epic A high level issue that usually consist of smaller tasks and corresponds to some functionality

Comments

@roma-glushko
Copy link
Member

Just like Glide supports text-to-text, embeddings & speech unified API, let's try to bring a support for zero/few-shot classification.
Cohere supports this case explicitly. Other providers like OpenAI/Anthropic don't.

In case provider doesn't support few-shot classification, we should be able to mimic it with prompt engineering.

Reference:

@roma-glushko roma-glushko added type:enhancement New feature or request area:config size:large type:epic A high level issue that usually consist of smaller tasks and corresponds to some functionality priority:mid-term The feature is on our radars labels May 6, 2024
@roma-glushko
Copy link
Member Author

cc @mkrueger12 just looked at your old idea at different angle. It doesn't feel like a big deal now, so I have filled this epic.

@roma-glushko roma-glushko added the model:classifier LLM-based few-shot classiers label May 13, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
area:config model:classifier LLM-based few-shot classiers priority:mid-term The feature is on our radars size:large type:enhancement New feature or request type:epic A high level issue that usually consist of smaller tasks and corresponds to some functionality
Projects
Status: Backlog
Development

No branches or pull requests

1 participant