microsoft
[Bug]: KeyError: '\n "title"' during community report extraction
Do you need to file an issue?
- [x] I have searched the existing issues and this bug is not already filed.
- [x] My model is hosted on OpenAI or Azure. If not, please look at the "model providers" issue and don't file a new one here.
- [x] I believe this is a legitimate bug, not just a question. If this is a question, please use the Discussions area.
Describe the bug
At the graphrag.index.operations.summarize_communities.community_reports_extractor step, I sometimes get an error:
Traceback (most recent call last): File "/home/admin/code/graphrag-build-update/.venv/lib/python3.12/site-packages/graphrag/index/operations/summarize_communities/community_reports_extractor.py", line 76, in __call__ prompt = self._extraction_prompt.format(**{ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ KeyError: '\n "title"'
What happens during auto-prompt tuning is that the tuned 'community_report_text.txt' contains both double {{}} and single {} for formatting placeholders sometimes, probably due to the LLM making a mistake during tuning. In those cases, the above-mentioned step goes wrong and the whole process stops. We now fix this by manually making all {{}} single (so {}), but I can imagine that it's preferable to fix this in the actual graphrag package.
Also see this issue which was marked stale, which seems to be the same issue: https://github.com/microsoft/graphrag/issues/1912.
Steps to reproduce
No response
Expected Behavior
No response
GraphRAG Config Used
# Paste your config here
Logs and screenshots
No response
Additional Information
- GraphRAG Version:
- Operating System:
- Python Version:
- Related Issues:
