MetaGPT
MetaGPT copied to clipboard
Integrated Milvus with MetaGPT
Features
- add another vector database milvus into MetaGPT
Feature Docs
Influence
Result
Other
:warning: Please install the to ensure uploads and comments are reliably processed by Codecov.
Codecov Report
Attention: Patch coverage is 35.95506% with 57 lines in your changes missing coverage. Please review.
Project coverage is 60.04%. Comparing base (
ab846f6) to head (4d92fdc). Report is 6 commits behind head on main.
:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.
Additional details and impacted files
@@ Coverage Diff @@
## main #1457 +/- ##
==========================================
- Coverage 62.59% 60.04% -2.55%
==========================================
Files 287 289 +2
Lines 17589 17677 +88
==========================================
- Hits 11009 10615 -394
- Misses 6580 7062 +482
| Flag | Coverage Δ | |
|---|---|---|
60.04% <35.95%> (?) |
Flags with carried forward coverage won't be shown. Click here to find out more.
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Can anyone help to review this PR? We are appreciate this MetaGPT project and hope Milvus can be integrated into it.
We found that if pymilvus dependency is added, it will conflict with some existing dependencies, so in the code, pymilvus dependency is commented out and lazy imported.
@better629
Hey, bro. I have meet the similar problem when I install it, so I have make some changes on dependency, in case that less version specified in .whl, my repo is MetaGPT-simple and you can install it with python -m pip install metagpt-simple, the pypi official link is metagpt-simple. Of course, I only do some simple test on it, if some error occur, please let me know.
Thanks for your suggestion. I saw that you lifted the version restrictions of all dependencies. But I think it may introduce new problems, that is, some of the latest versions of some dependencies may not be backward compatible, resulting in them not supporting the code of this project. I think the maintainer of this project needs to ensure the operability of the newer dependencies, and use some more powerful dependency management tools to manage dependencies, such as poetry. Maybe this is a better solution.
I performed the following operations:
- pip install llama-index-vector-stores-milvus==0.1.23
- Modify
milvus_connectionto my own milvus - pytest tests/metagpt/document_store/test_milvus_store.py
Finally, the Collection was successfully created.
LGTM