Source citation in RAG responses in LangChain - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the LangChain retriever class.

LangChain

from langchain.vectorstores import [1]

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

ATensorFlow

BPandas

CNumPy

DChroma

Attempts:

3 left

💡 Hint

Common Mistakes

Choosing unrelated libraries like Pandas or TensorFlow.

The Chroma class is used as a vector store retriever in LangChain for RAG.

2fill in blank

medium

Complete the code to initialize the retriever with the vector store.

LangChain

retriever = [1](embedding_function=embeddings, persist_directory='db')

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

AChroma

BWeaviate

CFAISS

DPinecone

Attempts:

3 left

💡 Hint

Common Mistakes

Using other vector stores not imported or configured.

Chroma is the vector store class used to create a retriever instance in LangChain.

3fill in blank

hard

Fix the error in the code to enable source citation in the RAG chain.

LangChain

qa_chain = RetrievalQAWithSourcesChain(llm=llm, retriever=[1], return_source_documents=True)

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

Aretriever()

Bretriever

Cretriever.run()

Dretriever.get()

Attempts:

3 left

💡 Hint

Common Mistakes

Calling retriever as a function instead of passing the object.

The retriever should be passed as an object, not called as a function.

4fill in blank

hard

Fill both blanks to extract answer and sources from the RAG output.

LangChain

answer = result[[1]]
sources = result[[2]]

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

A'answer'

B'sources'

C'source_documents'

D'text'

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'sources' key which does not exist in output.

The RAG output dictionary contains the 'answer' key for the answer text and 'source_documents' key for the source documents.

5fill in blank

hard

Fill all three blanks to create a dictionary of source texts from documents.

LangChain

source_texts = [1](doc.page_content for doc in [2])
source_dict = {i: text for i, text in [3](source_texts)}

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

Alist

Bresult['source_documents']

Cenumerate

Ddict

Attempts:

3 left

💡 Hint

Common Mistakes

Using dict() instead of enumerate() in the comprehension.

We first create a list of page contents from source documents, then use enumerate to pair indices with texts for the dictionary comprehension.