Implementation:Intel Ipex llm LangChain RAG Chain

Knowledge Sources	IPEX-LLM LangChain Documentation
Domains	NLP, RAG
Last Updated	2026-02-09 00:00 GMT

Overview

LangChain LCEL chain composing retriever, prompt template, LLM, and output parser for RAG on Intel XPU.

Description

This is a Wrapper Doc for LangChain's LCEL chain composition used in the IPEX-LLM RAG pipeline. The chain uses RunnablePassthrough to pass the question through, pipes retrieved documents through a format function, combines with a prompt template from LangChain Hub ("rlm/rag-prompt"), passes through the IpexLLM, and parses with StrOutputParser.

External Reference

• LangChain LCEL Documentation

Usage

Use after creating the retriever and LLM to assemble the complete RAG inference chain.

Code Reference

Source Location

• Repository: IPEX-LLM
• File: python/llm/example/GPU/LangChain/rag.py
• Lines: 77-88

Signature

from langchain import hub
from langchain_core.runnables import RunnablePassthrough
from langchain_core.output_parsers import StrOutputParser

def format_docs(docs):
    return "\n\n".join(doc.page_content for doc in docs)

prompt = hub.pull("rlm/rag-prompt")

rag_chain = (
    {"context": retriever | format_docs, "question": RunnablePassthrough()}
    | prompt
    | llm
    | StrOutputParser()
)

result = rag_chain.invoke(query: str) -> str

Import

from langchain import hub
from langchain_core.runnables import RunnablePassthrough
from langchain_core.output_parsers import StrOutputParser

I/O Contract

Inputs

Name	Type	Required	Description
retriever	VectorStoreRetriever	Yes	Retriever from Chroma vector store
llm	IpexLLM	Yes	IPEX-LLM LangChain LLM instance
query	str	Yes	User question to answer

Outputs

Name	Type	Description
result	str	Generated answer grounded in retrieved context

Usage Examples

from langchain import hub
from langchain_core.runnables import RunnablePassthrough
from langchain_core.output_parsers import StrOutputParser

def format_docs(docs):
    return "\n\n".join(doc.page_content for doc in docs)

# Pull RAG prompt template from LangChain Hub
prompt = hub.pull("rlm/rag-prompt")

# Assemble LCEL RAG chain
rag_chain = (
    {"context": retriever | format_docs, "question": RunnablePassthrough()}
    | prompt
    | llm
    | StrOutputParser()
)

# Invoke the chain
answer = rag_chain.invoke("What is IPEX-LLM?")
print(answer)

Related Pages

Implements Principle

• Principle:Intel_Ipex_llm_RAG_Chain_Assembly

Requires Environment

• Environment:Intel_Ipex_llm_RAG_LangChain_Environment