anything-llm/server/utils/vectorDbProviders/qdrant/index.js

398 lines
13 KiB
JavaScript
Raw Normal View History

const { QdrantClient } = require("@qdrant/js-client-rest");
const { RecursiveCharacterTextSplitter } = require("langchain/text_splitter");
const { storeVectorResult, cachedVectorInformation } = require("../../files");
const { v4: uuidv4 } = require("uuid");
const { toChunks, getLLMProvider } = require("../../helpers");
const { chatPrompt } = require("../../chats");
const QDrant = {
name: "QDrant",
connect: async function () {
if (process.env.VECTOR_DB !== "qdrant")
throw new Error("QDrant::Invalid ENV settings");
const client = new QdrantClient({
url: process.env.QDRANT_ENDPOINT,
...(process.env.QDRANT_API_KEY
? { apiKey: process.env.QDRANT_API_KEY }
: {}),
});
const isAlive = (await client.api("cluster")?.clusterStatus())?.ok || false;
if (!isAlive)
throw new Error(
"QDrant::Invalid Heartbeat received - is the instance online?"
);
return { client };
},
heartbeat: async function () {
await this.connect();
return { heartbeat: Number(new Date()) };
},
totalIndicies: async function () {
const { client } = await this.connect();
const { collections } = await client.getCollections();
var totalVectors = 0;
for (const collection of collections) {
if (!collection || !collection.name) continue;
totalVectors +=
(await this.namespace(client, collection.name))?.vectorCount || 0;
}
return totalVectors;
},
namespaceCount: async function (_namespace = null) {
const { client } = await this.connect();
const namespace = await this.namespace(client, _namespace);
return namespace?.vectorCount || 0;
},
similarityResponse: async function (_client, namespace, queryVector) {
const { client } = await this.connect();
const result = {
contextTexts: [],
sourceDocuments: [],
};
const responses = await client.search(namespace, {
vector: queryVector,
limit: 4,
});
responses.forEach((response) => {
result.contextTexts.push(response?.payload?.text || "");
result.sourceDocuments.push({
...(response?.payload || {}),
id: response.id,
});
});
return result;
},
namespace: async function (client, namespace = null) {
if (!namespace) throw new Error("No namespace value provided.");
const collection = await client.getCollection(namespace).catch(() => null);
if (!collection) return null;
return {
name: namespace,
...collection,
vectorCount: collection.vectors_count,
};
},
hasNamespace: async function (namespace = null) {
if (!namespace) return false;
const { client } = await this.connect();
return await this.namespaceExists(client, namespace);
},
namespaceExists: async function (client, namespace = null) {
if (!namespace) throw new Error("No namespace value provided.");
const collection = await client.getCollection(namespace).catch((e) => {
console.error("QDrant::namespaceExists", e.message);
return null;
});
return !!collection;
},
deleteVectorsInNamespace: async function (client, namespace = null) {
await client.deleteCollection(namespace);
return true;
},
getOrCreateCollection: async function (client, namespace) {
if (await this.namespaceExists(client, namespace)) {
return await client.getCollection(namespace);
}
await client.createCollection(namespace, {
vectors: {
size: 1536, //TODO: Fixed to OpenAI models - when other embeddings exist make variable.
distance: "Cosine",
},
});
return await client.getCollection(namespace);
},
addDocumentToNamespace: async function (
namespace,
documentData = {},
fullFilePath = null
) {
const { DocumentVectors } = require("../../../models/vectors");
try {
const { pageContent, docId, ...metadata } = documentData;
if (!pageContent || pageContent.length == 0) return false;
console.log("Adding new vectorized document into namespace", namespace);
const cacheResult = await cachedVectorInformation(fullFilePath);
if (cacheResult.exists) {
const { client } = await this.connect();
const collection = await this.getOrCreateCollection(client, namespace);
if (!collection)
throw new Error("Failed to create new QDrant collection!", {
namespace,
});
const { chunks } = cacheResult;
const documentVectors = [];
for (const chunk of chunks) {
const submission = {
ids: [],
vectors: [],
payloads: [],
};
// Before sending to Qdrant and saving the records to our db
// we need to assign the id of each chunk that is stored in the cached file.
chunk.forEach((chunk) => {
const id = uuidv4();
const { id: _id, ...payload } = chunk.payload;
documentVectors.push({ docId, vectorId: id });
submission.ids.push(id);
submission.vectors.push(chunk.vector);
submission.payloads.push(payload);
});
const additionResult = await client.upsert(namespace, {
wait: true,
batch: { ...submission },
});
if (additionResult?.status !== "completed")
throw new Error("Error embedding into QDrant", additionResult);
}
await DocumentVectors.bulkInsert(documentVectors);
return true;
}
// If we are here then we are going to embed and store a novel document.
// We have to do this manually as opposed to using LangChains `Qdrant.fromDocuments`
// because we then cannot atomically control our namespace to granularly find/remove documents
// from vectordb.
const textSplitter = new RecursiveCharacterTextSplitter({
chunkSize: 1000,
chunkOverlap: 20,
});
const textChunks = await textSplitter.splitText(pageContent);
console.log("Chunks created from document:", textChunks.length);
const LLMConnector = getLLMProvider();
const documentVectors = [];
const vectors = [];
const vectorValues = await LLMConnector.embedChunks(textChunks);
const submission = {
ids: [],
vectors: [],
payloads: [],
};
if (!!vectorValues && vectorValues.length > 0) {
for (const [i, vector] of vectorValues.entries()) {
const vectorRecord = {
id: uuidv4(),
vector: vector,
// [DO NOT REMOVE]
// LangChain will be unable to find your text if you embed manually and dont include the `text` key.
// https://github.com/hwchase17/langchainjs/blob/2def486af734c0ca87285a48f1a04c057ab74bdf/langchain/src/vectorstores/pinecone.ts#L64
payload: { ...metadata, text: textChunks[i] },
};
submission.ids.push(vectorRecord.id);
submission.vectors.push(vectorRecord.vector);
submission.payloads.push(vectorRecord.payload);
vectors.push(vectorRecord);
documentVectors.push({ docId, vectorId: vectorRecord.id });
}
} else {
console.error(
"Could not use OpenAI to embed document chunks! This document will not be recorded."
);
}
const { client } = await this.connect();
const collection = await this.getOrCreateCollection(client, namespace);
if (!collection)
throw new Error("Failed to create new QDrant collection!", {
namespace,
});
if (vectors.length > 0) {
const chunks = [];
console.log("Inserting vectorized chunks into QDrant collection.");
for (const chunk of toChunks(vectors, 500)) chunks.push(chunk);
const additionResult = await client.upsert(namespace, {
wait: true,
batch: {
ids: submission.ids,
vectors: submission.vectors,
payloads: submission.payloads,
},
});
if (additionResult?.status !== "completed")
throw new Error("Error embedding into QDrant", additionResult);
await storeVectorResult(chunks, fullFilePath);
}
await DocumentVectors.bulkInsert(documentVectors);
return true;
} catch (e) {
console.error("addDocumentToNamespace", e.message);
return false;
}
},
deleteDocumentFromNamespace: async function (namespace, docId) {
const { DocumentVectors } = require("../../../models/vectors");
const { client } = await this.connect();
if (!(await this.namespaceExists(client, namespace))) return;
const knownDocuments = await DocumentVectors.where(`docId = '${docId}'`);
if (knownDocuments.length === 0) return;
const vectorIds = knownDocuments.map((doc) => doc.vectorId);
await client.delete(namespace, {
wait: true,
points: vectorIds,
});
const indexes = knownDocuments.map((doc) => doc.id);
await DocumentVectors.deleteIds(indexes);
return true;
},
query: async function (reqBody = {}) {
const { namespace = null, input, workspace = {} } = reqBody;
if (!namespace || !input) throw new Error("Invalid request body");
const { client } = await this.connect();
if (!(await this.namespaceExists(client, namespace))) {
return {
response: null,
sources: [],
message: "Invalid query - no documents found for workspace!",
};
}
const LLMConnector = getLLMProvider();
const queryVector = await LLMConnector.embedTextInput(input);
const { contextTexts, sourceDocuments } = await this.similarityResponse(
client,
namespace,
queryVector
);
const prompt = {
role: "system",
content: `${chatPrompt(workspace)}
Context:
${contextTexts
.map((text, i) => {
return `[CONTEXT ${i}]:\n${text}\n[END CONTEXT ${i}]\n\n`;
})
.join("")}`,
};
const memory = [prompt, { role: "user", content: input }];
const responseText = await LLMConnector.getChatCompletion(memory, {
temperature: workspace?.openAiTemp ?? 0.7,
});
return {
response: responseText,
sources: this.curateSources(sourceDocuments),
message: false,
};
},
// This implementation of chat uses the chat history and modifies the system prompt at execution
// this is improved over the regular langchain implementation so that chats do not directly modify embeddings
// because then multi-user support will have all conversations mutating the base vector collection to which then
// the only solution is replicating entire vector databases per user - which will very quickly consume space on VectorDbs
chat: async function (reqBody = {}) {
const {
namespace = null,
input,
workspace = {},
chatHistory = [],
} = reqBody;
if (!namespace || !input) throw new Error("Invalid request body");
const { client } = await this.connect();
if (!(await this.namespaceExists(client, namespace))) {
return {
response: null,
sources: [],
message: "Invalid query - no documents found for workspace!",
};
}
const LLMConnector = getLLMProvider();
const queryVector = await LLMConnector.embedTextInput(input);
const { contextTexts, sourceDocuments } = await this.similarityResponse(
client,
namespace,
queryVector
);
const prompt = {
role: "system",
content: `${chatPrompt(workspace)}
Context:
${contextTexts
.map((text, i) => {
return `[CONTEXT ${i}]:\n${text}\n[END CONTEXT ${i}]\n\n`;
})
.join("")}`,
};
const memory = [prompt, ...chatHistory, { role: "user", content: input }];
const responseText = await LLMConnector.getChatCompletion(memory, {
temperature: workspace?.openAiTemp ?? 0.7,
});
return {
response: responseText,
sources: this.curateSources(sourceDocuments),
message: false,
};
},
"namespace-stats": async function (reqBody = {}) {
const { namespace = null } = reqBody;
if (!namespace) throw new Error("namespace required");
const { client } = await this.connect();
if (!(await this.namespaceExists(client, namespace)))
throw new Error("Namespace by that name does not exist.");
const stats = await this.namespace(client, namespace);
return stats
? stats
: { message: "No stats were able to be fetched from DB for namespace" };
},
"delete-namespace": async function (reqBody = {}) {
const { namespace = null } = reqBody;
const { client } = await this.connect();
if (!(await this.namespaceExists(client, namespace)))
throw new Error("Namespace by that name does not exist.");
const details = await this.namespace(client, namespace);
await this.deleteVectorsInNamespace(client, namespace);
return {
message: `Namespace ${namespace} was deleted along with ${details?.vectorCount} vectors.`,
};
},
reset: async function () {
const { client } = await this.connect();
const response = await client.getCollections();
for (const collection of response.collections) {
await client.deleteCollection(collection.name);
}
return { reset: true };
},
curateSources: function (sources = []) {
const documents = [];
for (const source of sources) {
if (Object.keys(source).length > 0) {
documents.push({
...source,
});
}
}
return documents;
},
};
module.exports.QDrant = QDrant;