feat(summary): improve prompt to get more insights #2446
Merged
Conversation
This file contains 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
dosubot
bot
added
the
size:S
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
dosubot
bot
added
the
area: backend
![ellipsis-dev[bot]](https://avatars.githubusercontent.com/in/64358?s=60&v=4){kind=small}
There was a problem hiding this comment.
👍 Looks good to me!
- Reviewed the entire pull request up to adf5541
- Looked at
54lines of code in1files - Took 1 minute and 16 seconds to review
More info
- Skipped
0files when reviewing. - Skipped posting
1additional comments because they didn't meet confidence threshold of50%.
1. backend/modules/assistant/ito/summary.py:154:
- Assessed confidence :
50% - Comment:
The chunk overlap in the text splitter has been changed from 0 to 100. This could potentially lead to redundant processing of text chunks. Please ensure that this change is necessary and justified in terms of performance and accuracy of the text splitting process. - Reasoning:
The PR author has made changes to the map and reduce templates to provide clearer instructions for analyzing each section of the document and generating a consolidated summary. The changes seem to be in line with the PR description and there doesn't seem to be any logical bugs or violations of best practices. However, there is a change in the chunk overlap in the text splitter from 0 to 100. This change might affect the performance of the text splitting process, as overlapping chunks could lead to redundant processing. I need to check if this change is justified and necessary.
Workflow ID: wflow_I9uZxshGA9HDM7aY
Not what you expected? You can customize the content of the reviews using rules. Learn more here.
⌛ You have 18 hours remaining in your free trial. Upgrade at https://app.ellipsis.dev for $20/developer/month. If you have any questions, reach us at help@ellipsis.dev
StanGirard
added a commit
that referenced
this pull request
Apr 19, 2024
🤖 I have created a release *beep* *boop* --- ## 0.0.231 (2024-04-19) ## What's Changed * feat(assistants): Add user usage update and pricing calculation to ITO assistant by @StanGirard in #2433 * feat(assistant): improve prompt summary by @StanGirard in #2435 * feat(assistants): Add PDF generation functionality and nice emails by @StanGirard in #2436 * feat(analytics): rely on sql rather that python loop for brains by @StanGirard in #2437 * fix(assistant): summary now can output 2000 tokens by @StanGirard in #2440 * feat(assistant): check if key of file is same as filename uploaded by @StanGirard in #2439 * feat: Update Docker build commands and dependencies by @StanGirard in #2441 * feat(rag): Refactor DEFAULT_DOCUMENT_PROMPT in quivr_rag.py by @StanGirard in #2442 * Enable Porter Application quivr-back by @porter-deployment-app in #2443 * Enable Porter Application quivr-demo-front by @porter-deployment-app in #2444 * fix(assistants): brain id is null by @StanGirard in #2445 * feat(summary): improve prompt to get more insights by @StanGirard in #2446 * feat(aws): Update CPU and memory configurations for task definitions by @StanGirard in #2447 * feat(frontend): Quivr Assistants by @Zewed in #2448 **Full Changelog**: v0.0.230...v0.0.231 --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please). <!-- ELLIPSIS_HIDDEN --> ---- | 🚀 This description was created by [Ellipsis](https://www.ellipsis.dev) for commit a16fa4d | |--------| ### Summary: This PR releases version 0.0.231, introducing several feature enhancements and bug fixes across the assistant, analytics, Docker, and frontend modules. **Key points**: - Release version 0.0.231 with feature enhancements and bug fixes across multiple modules - User usage update and pricing calculation added to ITO assistant - Improved prompt summary in assistant module - PDF generation functionality and email enhancements added - Analytics optimized by relying on SQL instead of Python loop - Token output limit fixed in assistant summary - Docker build commands and dependencies updated - DEFAULT_DOCUMENT_PROMPT in quivr_rag.py refactored - Porter Applications for quivr-back and quivr-demo-front enabled - Null brain id issue fixed in assistants module - Prompt improved for better insights in summary module - CPU and memory configurations for AWS task definitions updated - Quivr Assistants added in frontend ---- Generated with ❤️ by [ellipsis.dev](https://www.ellipsis.dev) <!-- ELLIPSIS_HIDDEN -->
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 pull request includes a commit that improves the prompt used to get more insights from the document. The commit modifies the map and reduce templates to provide clearer instructions for analyzing each section of the document and generating a consolidated summary. The changes also include adjustments to the text splitter to improve the chunk overlap.
Summary:
This PR enhances the document analysis and summary generation instructions in the
SummaryAssistantclass and improves text splitting by adjusting thechunk_overlapparameter.Key points:
map_templateandreduce_templateinprocess_assistantmethod ofSummaryAssistantclass in/backend/modules/assistant/ito/summary.py.chunk_overlapparameter ofCharacterTextSplitterinstance from 0 to 100.Generated with ❤️ by ellipsis.dev