- Add new repository to GRAD
uv run autods/grad/grad.py add https://github.com/owner/repo_name- Ask repository using GRAD
uv run autods/grad/grad.py ask https://github.com/owner/repo_name "How to use LightAutoML library?"If you encounter an error like:
LanceError(IO): Too many open files (os error 24)
This is a macOS/Linux file descriptor limit issue. LanceDB (the vector database) opens many files for its index, which can exceed the default system limit.
Quick fix - increase the limit for your current terminal session:
ulimit -n 10240Permanent fix - add to your ~/.zshrc or ~/.bashrc:
ulimit -n 10240To check your current limits:
ulimit -n # Current soft limit (often 256 on macOS)
ulimit -Hn # Hard limit (max you can set)If you encounter an error like:
Buffer manager exception: Unable to allocate memory! The buffer pool is full and no memory could be freed!
This is a Kuzu graph database buffer pool limit issue. The buffer pool may fill up during concurrent graph operations.
Option 1: Clear Cognee data (try first):
rm -rf .venv/lib/python3.12/site-packages/cognee/.cognee_system/Option 2: Increase Kuzu buffer pool size (if Option 1 doesn't help):
Edit .venv/lib/python3.12/site-packages/cognee/infrastructure/databases/graph/kuzu/adapter.py and change all occurrences of buffer pool settings:
# Change from:
buffer_pool_size=2048 * 1024 * 1024, # 2GB
max_db_size=4096 * 1024 * 1024, # 4GB
# To:
buffer_pool_size=8 * 1024 * 1024 * 1024, # 8GB
max_db_size=16 * 1024 * 1024 * 1024, # 16GBNote: This change will be lost if you reinstall/update Cognee.