Why the RAG chain connects retrieval to generation in LangChain - Test Your Understanding

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a RAG chain that connects retrieval to generation.

LangChain

from langchain.chains import RetrievalQA
from langchain.llms import OpenAI

retriever = get_retriever()
llm = OpenAI()
rag_chain = RetrievalQA(llm=llm, retriever=[1])

Drag options to blanks, or click blank then click option'

Aretriever

Bllm

Cquery

Ddocument

Attempts:

3 left

💡 Hint

Common Mistakes

Passing the language model instead of the retriever to the chain.

Using an undefined variable instead of the retriever.

The RAG chain connects the retriever to the language model by passing the retriever object to the RetrievalQA chain.

2fill in blank

medium

Complete the code to retrieve documents before generating an answer.

LangChain

query = "What is RAG?"
docs = [1].get_relevant_documents(query)
answer = llm.generate(docs)

Drag options to blanks, or click blank then click option'

Aretriever

Bllm

Cquery

Drag_chain

Attempts:

3 left

💡 Hint

Common Mistakes

Calling get_relevant_documents on the language model.

Using the query variable instead of the retriever.

The retriever fetches relevant documents for the query before the language model generates an answer.

3fill in blank

hard

Fix the error in connecting retrieval to generation in the RAG chain.

LangChain

from langchain.chains import RetrievalQA
from langchain.llms import OpenAI

retriever = get_retriever()
llm = OpenAI()
rag_chain = RetrievalQA(llm=[1], retriever=llm)
result = rag_chain.run("Explain RAG")

Drag options to blanks, or click blank then click option'

Aresult

BOpenAI

Cquery

Dretriever

Attempts:

3 left

💡 Hint

Common Mistakes

Swapping llm and retriever arguments.

Passing the wrong variable to retriever.

The retriever must be passed as the retriever argument, not the language model.

4fill in blank

hard

Fill both blanks to create a dictionary comprehension that maps documents to their content length.

LangChain

doc_lengths = { [1]: len([2]) for doc in docs}

Drag options to blanks, or click blank then click option'

Adoc.content

Bdoc

Ddoc.text

Attempts:

3 left

💡 Hint

Common Mistakes

Using doc.text instead of doc.content.

Swapping key and value in the comprehension.

The dictionary comprehension uses doc as key and length of doc.content as value.

5fill in blank

hard

Fill all three blanks to filter documents with content longer than 100 characters and create a summary dictionary.

LangChain

summary = {doc.id: [1] for doc in docs if len([2]) [3] 100}

Drag options to blanks, or click blank then click option'

Adoc.content[:50]

Bdoc.content

C>

D<

Attempts:

3 left

💡 Hint

Common Mistakes

Using '<' instead of '>' in the condition.

Using doc.id as value instead of key.

Not slicing the content for summary.

We take the first 50 characters of doc.content for docs where content length is greater than 100.