Explore advanced indexing techniques for similarity search in Faiss through this informative 26-minute video. Dive into composite indexes, including IVFADC, Multi-D-ADC, and IVF-HNSW, to optimize vector search performance. Learn how to balance recall, latency, and memory usage by combining different indexing methods and vector transformations. Understand the implementation of these techniques in Faiss and gain insights into building robust, high-performance vector similarity search applications. Discover when and how to apply various indexes and vector processing steps to create an ideal composite index for your specific needs.
Overview
Syllabus
Intro
IVFADC
IVFADC in Faiss
Multi-D-ADC
Multi-D-ADC in Faiss
IVF-HNSW
IVF-HNSW in Faiss
Outro
Taught by
James Briggs
Related Courses
-
Composite Indexes and the Faiss Index Factory
-
HNSW for Vector Search Explained and Implemented with Faiss - Python
-
Retrieval Optimization: Tokenization to Vector Quantization
-
Choosing Indexes for Similarity Search - Faiss in Python
-
Finding Relevant Items Using Approximate Nearest Neighbor Search Algorithms
-
Product Quantization for Vector Similarity Search - Python
