How to load and save documents using FaissStore from Langchain in the Next.js pages/api ???
Unanswered
Sun bear posted this in #help-forum
Sun bearOP
Trying to run
I might be calling the wrong path from the new TextLoader??
What would be wrong that this code doesn't do anything when I call it?
Complete code here:
node pages/api/indexes.js but it does nothing, it doesn't even give me an error... it is suppose to create 2 documents but I can't see them anywhere...I might be calling the wrong path from the new TextLoader??
const loader = new TextLoader("./restaurant.txt"); <-- the restaurant.txt is under pages/apiWhat would be wrong that this code doesn't do anything when I call it?
Complete code here:
const { OpenAIEmbeddings } = require("langchain/embeddings/openai");
const { FaissStore } = require("langchain/vectorstores/faiss");
const { CharacterTextSplitter } = require("langchain/text_splitter");
const { TextLoader } = require("langchain/document_loaders/fs/text");
const indexes = async (req, res) => {
const loader = new TextLoader("./restaurant.txt"); /// <--- this is the call to the file
console.log(loader);
const docs = await loader.load();
const splitter = new CharacterTextSplitter({
chunkSize: 200,
chunkOverlap: 50,
});
const documents = await splitter.splitDocuments(docs);
console.log(documents);
const embeddings = new OpenAIEmbeddings();
const vectorStore = await FaissStore.fromDocuments(documents, embeddings);
await vectorStore.save("./"); /// <-- here is where it should be saving 2 other files, but I can't find them anywhere.
res.status(200).json({ ok: ok });
};
module.exports = { indexes };