LangChain Exercises — Practice Coding

1

Complete the LCEL chain construction using the pipe operator

const chain = prompt model outputParser;

2

Create a ChatPromptTemplate from a template string

const prompt = ChatPromptTemplate.("Tell me about {topic}");

3

Name the output parser that extracts plain text from model output

import { } from "@langchain/core/output_parsers";

const parser = new StringOutputParser();

4

Invoke a chain with variables

const result = await chain.({

topic: "recursion",

language: "Python"

});

5

Name the text splitter that uses multiple separators recursively

const splitter = new CharacterTextSplitter({

chunkSize: 1000,

chunkOverlap: 200

});

6

Get a retriever from a vector store

const retriever = vectorStore.({

k: 4,

searchType: "similarity"

});

7

Name the memory class that stores all conversation messages

import { } from "langchain/memory";

const memory = new BufferMemory();

8

Create a LangChain RAG chain using the retrieval chain factory

const ragChain = await ({

retriever: vectorStore.asRetriever(),

combineDocsChain: documentChain

});