LangChain Example

ingest.mjs

import { ClickHouseStore } from "@langchain/community/vectorstores/clickhouse";
import { createRetrievalChain } from "langchain/chains/retrieval";
import { OpenAI, OpenAIEmbeddings } from "@langchain/openai";
import { ChatPromptTemplate } from "@langchain/core/prompts";
import { createStuffDocumentsChain } from "langchain/chains/combine_documents";
import { RecursiveCharacterTextSplitter } from "langchain/text_splitter";

import { Document } from '@langchain/core/documents'

const openAIApiKey = "sk-xxx"
const vectorStore = await ClickHouseStore.fromExistingIndex(
  new OpenAIEmbeddings({ openAIApiKey }),
  {
    host: "localhost",
    port: 8123,
    protocol: "http://",
    indexType: "hypothesis"
  }
);

import { CheerioWebBaseLoader } from "langchain/document_loaders/web/cheerio";

const loader = new CheerioWebBaseLoader(
  "https://docs.smith.langchain.com/user_guide"
);

const docs = await loader.load();


const splitter = new RecursiveCharacterTextSplitter();

const splitDocs = await splitter.splitDocuments(
  docs  
);

vectorStore.addDocuments(
  splitDocs
)

script.mjs

import { ClickHouseStore } from "@langchain/community/vectorstores/clickhouse";
import { createRetrievalChain } from "langchain/chains/retrieval";
import { OpenAI, OpenAIEmbeddings } from "@langchain/openai";
import { ChatPromptTemplate } from "@langchain/core/prompts";
import { createStuffDocumentsChain } from "langchain/chains/combine_documents";

const openAIApiKey = "sk-xxx"
const vectorStore = await ClickHouseStore.fromExistingIndex(
  new OpenAIEmbeddings({ openAIApiKey }),
  {
    host: "localhost",
    port: 8123,
    protocol: "http://",
    indexType: "hypothesis"
  }
);

const model = new OpenAI({ temperature: 0, openAIApiKey: openAIApiKey, modelName: "gpt-3.5-turbo-0125" });
const retriever = vectorStore.asRetriever()

const prompt =
  ChatPromptTemplate.fromTemplate(`Answer the following question based only on the provided context:

<context>
{context}
</context>

Question: {input}`);

const documentChain = await createStuffDocumentsChain({
  llm: model,
  prompt,
});


const retrievalChain = await createRetrievalChain({
  retriever: vectorStore.asRetriever(),
  combineDocsChain: documentChain
});

const response = await retrievalChain.invoke({ input: "what is LangSmith?" });

console.log(response)