Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[Graph RAG] Init Commit with GraphRag interfaces #3388

Merged
merged 10 commits into from
Sep 4, 2024
Merged

[Graph RAG] Init Commit with GraphRag interfaces #3388

merged 10 commits into from
Sep 4, 2024
+151 −1

Conversation

randombet
Copy link
Collaborator

Why are these changes needed?

Graph RAG is a long-sought feature from the community, we plan to support this and integrate FalkorDB and MS GraphRAG.
Please refer to the design doc for more details.
https://docs.google.com/document/d/15uuNrJrxujX8biPCAZzAtmeMKOFB4SEMkVszcE9nzRo/edit

This is the first PR to set up the graph rag interfaces.

Related issue number

GraphRAG support #3119

Checks

@thinkall thinkall self-assigned this Aug 20, 2024
@thinkall thinkall added the rag retrieve-augmented generative agents label Aug 20, 2024
thinkall
thinkall reviewed Aug 22, 2024
View reviewed changes
Copy link
Collaborator

@thinkall thinkall left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you @randombet for the PR. I've left some minor suggestions.

autogen/agentchat/contrib/graph_rag/graph_rag_agent.py Outdated Show resolved Hide resolved
autogen/agentchat/contrib/graph_rag/graph_rag_agent.py Outdated Show resolved Hide resolved
autogen/agentchat/contrib/graph_rag/graph_rag_agent.py Outdated Show resolved Hide resolved
autogen/agentchat/contrib/graph_rag/graph_rag_agent.py Outdated Show resolved Hide resolved
autogen/agentchat/contrib/graph_rag/graph_store.py Outdated Show resolved Hide resolved
@codecov-commenter
Copy link

codecov-commenter commented Aug 29, 2024 •
edited
Loading

Codecov Report

Attention: Patch coverage is 0% with 33 lines in your changes missing coverage. Please review.

Project coverage is 22.24%. Comparing base (6279247) to head (a9c53aa).
Report is 138 commits behind head on main.

Files with missing lines Patch % Lines
.../agentchat/contrib/graph_rag/graph_query_engine.py 0.00% 14 Missing ⚠️
autogen/agentchat/contrib/graph_rag/document.py 0.00% 12 Missing ⚠️
...gentchat/contrib/graph_rag/graph_rag_capability.py 0.00% 7 Missing ⚠️
Additional details and impacted files 
@@             Coverage Diff             @@
##             main    #3388       +/-   ##
===========================================
- Coverage   32.90%   22.24%   -10.67%     
===========================================
  Files          94      105       +11     
  Lines       10235    11095      +860     
  Branches     2193     2540      +347     
===========================================
- Hits         3368     2468      -900     
- Misses       6580     8383     +1803     
+ Partials      287      244       -43
Flag Coverage Δ
unittest 11.90% <0.00%> (?)
unittests 22.18% <0.00%> (-10.73%) ⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

gagb
gagb reviewed Aug 29, 2024
View reviewed changes
Copy link
Collaborator

@gagb gagb left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is it correct that this PR creates the interfaces for GraphRAG but does not fully implement it? Will example notebooks and use cases be added to this PR?

@thinkall
Copy link
Collaborator

Is it correct that this PR creates the interfaces for GraphRAG but does not fully implement it? Will example notebooks and use cases be added to this PR?

The original plan is to first add interface and then add implementations. In this PR, only interface will be added. WDYT? @gagb

@gagb
Copy link
Collaborator

gagb commented Aug 30, 2024 •
edited
Loading

Thank you @thinkall I trust your expertise on this PR! Is there a reason to merge this before adding an implementation? Also can we remove the changes in the notebook -- they seems to be unrelated :) But the changes in the notebook are also non consequential.

@randombet
Copy link
Collaborator Author

@thinkall I updated the PR, pls review.
I checked FalkorDB, MS GraphRAG, LlamaIndex with NebulaGraph.
The change to use capability is because

  1. Current Graph RAG implementation frameworks integrate retrieve and generate together.
  2. Graph RAG Framework retrieve module assumes they knows about the underling graph hierarchy that we could not separate from Agent side.
  3. This is also easier for us to integrate to agent.

thinkall
thinkall reviewed Sep 4, 2024
View reviewed changes
Copy link
Collaborator

@thinkall thinkall left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you @randombet , would you like to add a test for graph_rag? Thanks.

@thinkall thinkall added this pull request to the merge queue Sep 4, 2024
Merged via the queue into microsoft:main with commit e579a46 Sep 4, 2024
142 of 154 checks passed
@gkorland
Copy link

Thanks! When do you plan to release the next version with this capability?

@thinkall thinkall mentioned this pull request Oct 2, 2024
Open
11 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@gagb gagb gagb left review comments

@2152505 2152505 2152505 approved these changes

@thinkall thinkall thinkall approved these changes

@qingyun-wu qingyun-wu Awaiting requested review from qingyun-wu

@marklysze marklysze Awaiting requested review from marklysze

@ekzhu ekzhu Awaiting requested review from ekzhu

@jackgerrits jackgerrits Awaiting requested review from jackgerrits

Assignees

@thinkall thinkall

Labels
rag retrieve-augmented generative agents
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
6 participants
@randombet @codecov-commenter @thinkall @gagb @gkorland @2152505