This repository was archived by the owner on Jan 2, 2025. It is now read-only.
Add frequency penalty for proc queries #693
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
calyptobai
approved these changes
Jun 27, 2023
| { | ||
| "name": "proc", | ||
| "description": "Read one or more files and extract the line ranges which are relevant to the search terms.", | ||
| "description": "Read one or more files and extract the line ranges which are relevant to the search terms. Do not proc more than 10 files at a time.", |
Contributor
There was a problem hiding this comment.
Why do we add this in the prompt rather than when consuming the results?
Contributor
Author
There was a problem hiding this comment.
The hypothesis is that it encourages the model to be selective about the files that it processes. Want to avoid situations where the model chooses to process every file in the path history. Have seen an example where it chose to process 54 files.
oppiliappan
approved these changes
Jun 28, 2023
oppiliappan
approved these changes
Jun 28, 2023
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
3 participants
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Outdated suggestions cannot be applied.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.
Occasionally the proc query ignores its instructions and generates hundreds of line ranges. This PR addresses this by setting a low
frequency_penaltyfor proc calls, discouraging the model from generating many ranges. Calling proc with this penalty has been evaluated on a test set which shows that it does not degrade proc performance.Also updates the action selection prompt, instructing the model to proc a maximum of 10 files at once.