skills.rest
zerwiz
zerwizCOMMUNITY
Explore all skills by this author →
GitHub

faiss

Optimizes FAISS index selection and GPU-accelerated similarity search for dense embeddings.

1|Updated Apr 10, 2026
Search Tags:#embeddings#vector search#similarity search#faiss#gpu acceleration#index tuning#nearest neighbor
One-click install
npx skills add https://github.com/zerwiz/wayofpi --skill faiss-zerwiz
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: faiss
Source: https://github.com/zerwiz/wayofpi/tree/main/.hermes/hermes-agent/optional-skills/mlops/faiss
Command: npx skills add https://github.com/zerwiz/wayofpi --skill faiss-zerwiz

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps you perform fast similarity search and clustering on dense vector embeddings when exact lookup, large-scale retrieval, or low-latency nearest-neighbor search becomes difficult to manage manually.

Core Features & Use Cases

  • Index Selection: Choose between Flat, IVF, HNSW, PQ, and combined index strategies based on dataset size, memory limits, and accuracy goals.
  • High-Scale Retrieval: Build and query vector indexes for semantic search, recommendation systems, retrieval-augmented generation, and embedding-driven matching.
  • Performance Tuning: Adjust training, probing, normalization, and GPU usage to balance speed, recall, and memory efficiency.
  • Use Case: If you need to search millions of document embeddings for the most relevant passages, this Skill guides you to the right FAISS index and retrieval workflow.

Quick Start

Use the faiss skill to select the best index for your vector dataset and run a similarity search workflow for your embeddings.

Frequently Asked Questions about faiss

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I choose the right vector index for similarity search?▼

Vector index selection depends on dataset size, memory limits, and accuracy goals. Choose between Flat, IVF, HNSW, PQ, or combined index strategies to balance search speed, recall, and memory efficiency for dense embeddings.

Can I use FAISS for similarity search with LangChain and LlamaIndex?▼

Yes, FAISS supports direct integration with LangChain and LlamaIndex. You can build and query vector indexes for semantic search and retrieval-augmented generation workflows within these frameworks.

How do I run nearest neighbor search on billions of vectors?▼

To run nearest neighbor search on billions of vectors, use GPU-accelerated retrieval and approximate indexes like IVF or PQ. Adjust probing and training parameters to optimize latency and recall at massive scale.

What is the difference between exact and approximate nearest neighbor search?▼

Exact search uses Flat indexes to guarantee perfect recall by scanning all vectors, while approximate search uses IVF, HNSW, or PQ indexes to trade minimal accuracy for significantly faster retrieval speeds and lower memory usage.

How do I tune FAISS performance for dense embeddings?▼

Tune FAISS performance by adjusting index training, probing parameters, and cosine normalization. Enable GPU acceleration to balance search speed, recall, and memory efficiency for your dense vector embeddings.

Related Skills

Explore More Software Engineering Skills→
haJ1thaJ1t

embedding-manager

Generate and store dense and sparse embeddings for similarity search.

Community
Advanced
twjarviszyra-webtwjarviszyra-web
3

faiss

Build exact and approximate FAISS similarity search indexes for dense vectors.

Community
Intermediate
ZardLi1115ZardLi1115
4

chroma

Stores embeddings and metadata for filtered, similarity-based retrieval.

Community
Basic
zechenzhangAGIzechenzhangAGI
11.5k

pinecone

Manage a serverless vector database with hybrid search and metadata filtering.

Community
Intermediate