Could you add a setting to the document settings to set the amount of text snippets that get used in the context?
Is it only one right now?
And it would be great to be able to see the full RAG text for troubleshooting. Answering RAG questions is really slow for me right now, would be great if could see what was inserted as context in the conversation
Originally created by @Venkman42 on GitHub (Feb 26, 2024).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/923
Originally assigned to: @tjbck on GitHub.
Could you add a setting to the document settings to set the amount of text snippets that get used in the context?
Is it only one right now?
And it would be great to be able to see the full RAG text for troubleshooting. Answering RAG questions is really slow for me right now, would be great if could see what was inserted as context in the conversation
Seems like at the moment it is hardcoded to 4: k: Optional[int] = 4 in RAG main.py
For now you can see the snippets sent to the LLM for RAG in the browser console after sending your request.
<!-- gh-comment-id:1966456831 -->
@jannikstdl commented on GitHub (Feb 27, 2024):
Seems like at the moment it is hardcoded to 4:
`k: Optional[int] = 4` in RAG main.py
For now you can see the snippets sent to the LLM for RAG in the browser console after sending your request.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @Venkman42 on GitHub (Feb 26, 2024).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/923
Originally assigned to: @tjbck on GitHub.
Could you add a setting to the document settings to set the amount of text snippets that get used in the context?
Is it only one right now?
And it would be great to be able to see the full RAG text for troubleshooting. Answering RAG questions is really slow for me right now, would be great if could see what was inserted as context in the conversation
@jannikstdl commented on GitHub (Feb 27, 2024):
Seems like at the moment it is hardcoded to 4:
k: Optional[int] = 4in RAG main.pyFor now you can see the snippets sent to the LLM for RAG in the browser console after sending your request.