phowell
/
pocketflow
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

update rag doc

This commit is contained in:
zachary62 2025-03-08 01:26:31 -05:00
parent 8c88175dc5
commit 79f3d4c867
1 changed files with 4 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
docs/design_pattern/rag.md
View File

@ -7,23 +7,15 @@ nav_order: 4
# RAG (Retrieval Augmented Generation) # RAG (Retrieval Augmented Generation)
For certain LLM tasks like answering questions, providing context is essential. For certain LLM tasks like answering questions, providing relevant context is essential. One common architecture is a **two-stage** RAG pipeline:
Most common way to retrive text-based context is through embedding:
1. Given texts, you first [chunk](../utility_function/chunking.md) them. 1. **Offline stage**: Preprocess and index documents ("building the index").
2. Next, you [embed](../utility_function/embedding.md) each chunk. 2. **Online stage**: Given a question, generate answers by retrieving the most relevant context from the index.
3. Then you store the chunks in [vector databases](../utility_function/vector.md).
4. Finally, given a query, you embed the query and find the closest chunk in the vector databases.
<div align="center"> <div align="center">
<img src="https://github.com/the-pocket/PocketFlow/raw/main/assets/rag.png?raw=true" width="250"/> <img src="https://github.com/the-pocket/PocketFlow/raw/main/assets/rag.png?raw=true" width="250"/>
</div> </div>
# RAG (Retrieval Augmented Generation)
For certain LLM tasks like answering questions, providing relevant context is essential. One common architecture is a **two-stage** RAG pipeline:
1. **Offline stage**: Preprocess and index documents ("building the index").
2. **Online stage**: Given a question, generate answers by retrieving the most relevant context from the index.
--- ---
## Stage 1: Offline Indexing ## Stage 1: Offline Indexing