⚡ Bolt: Optimize vector DB search with vectorized operations - #6
⚡ Bolt: Optimize vector DB search with vectorized operations#6google-labs-jules[bot] wants to merge 1 commit into
Conversation
💡 What: - Replaced iterative cosine similarity calculation in `HNSWIndex.search` with vectorized NumPy operations. - Implemented `_vector_matrix` caching to avoid rebuilding the matrix on every search. - Used `np.argpartition` for efficient top-k selection. 🎯 Why: - The previous implementation iterated over the dictionary of vectors in pure Python, which is O(N) and slow for large datasets. - Profiling showed that search time for 10k vectors was ~86ms per query. 📊 Impact: - Reduces search time for 10k vectors from ~86ms to ~5ms (17x speedup). - Increases query throughput from ~11 QPS to ~193 QPS. 🔬 Measurement: - Verified using a benchmark script `benchmark_vector_db_v3.py` (deleted before commit). - Verified correctness by running existing integration tests.
|
👋 Jules, reporting for duty! I'm here to lend a hand with this pull request. When you start a review, I'll add a 👀 emoji to each comment to let you know I've read it. I'll focus on feedback directed at me and will do my best to stay out of conversations between you and other bots or reviewers to keep the noise down. I'll push a commit with your requested changes shortly after. Please note there might be a delay between these steps, but rest assured I'm on the job! For more direct control, you can switch me to Reactive Mode. When this mode is on, I will only act on comments where you specifically mention me with New to Jules? Learn more at jules.google/docs. For security, I will only act on instructions from the user who triggered this task. |
Optimized
memory/lightweight_vector_db.pyby vectorizing the search operation using NumPy.Introduced matrix caching to minimize overhead during search.
Verified significant performance improvement via benchmarking.
PR created automatically by Jules for task 9544349186448933250 started by @Rohith-Shimori