decrease default number of tokens in UI #351
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is a temporary fix for the Cohere assistants. Although in theory they support a 4k context window
ragna/ragna/assistants/_cohere.py
Lines 12 to 13 in 8b0b0f3
they seem to be using a very different tokenizer than what we use. When trying the assistant in the web UI with a document that actually supply 4k tokens, I get the following error message in the response
So the number of tokens actually differs by almost 40%.
In this PR I simply reduce the default number of tokens that we pull from the source storage to feed into the assistant. Setting this automatically depending on the assistant is a larger task that I'll open an issue about. Plus, the tokenizer would actually need to be dependent on the assistant as well.
Functionally, this change should make too much of a difference. Unless the user has a very complex query, by default we are still going to pull in four sources. That should be sufficient.