view --main rag-system-builder-skill-keshirovaniya-embeddingov.md
rag-system-builder: Скилл кэширования эмбеддингов
readonly
--- lines
---
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
```python
# Load all embeddings into memory at startup
self.embedding_cache = self._load_all_embeddings()
```
## 2. Use FAISS for Large Collections
```python
import faiss
# Build FAISS index for fast similarity search
index = faiss.IndexFlatIP(dimension) # Inner product for cosine sim
index.add(embeddings)
```
## 3. Batch Queries
```python
# Process multiple questions efficiently
questions = ["Q1", "Q2", "Q3"]
query_embeddings = model.embed_batch(questions)
```
Инициализация мануала...
package.json
$ install --global
skills.sh
npx skills add https://github.com/vamseeachanta/workspace-hub/tree/main/.claude/skills/_archive/data/documents/rag-system-builder/1-cache-embeddings
$ download --local
man
[HINT] Скачивает всю директорию скилла с GitHub: SKILL.md и все связанные файлы