rag-system-builder-1-cache-embeddings

name	rag-system-builder-1-cache-embeddings
description	Sub-skill of rag-system-builder: 1. Cache Embeddings (+2).
version	1.2.0
category	data
type	reference
scripts_exempt	true

1. Cache Embeddings (+2)

1. Cache Embeddings

# Load all embeddings into memory at startup
self.embedding_cache = self._load_all_embeddings()

2. Use FAISS for Large Collections

import faiss

# Build FAISS index for fast similarity search
index = faiss.IndexFlatIP(dimension)  # Inner product for cosine sim
index.add(embeddings)

3. Batch Queries

# Process multiple questions efficiently
questions = ["Q1", "Q2", "Q3"]
query_embeddings = model.embed_batch(questions)