I have created a chatbot using langchain and custom prompt template but when I send a message “Who are you” the bot responded “I am an AI assistant programmed to provide information and answer questions. How can I assist you today?” and here is my code
let retriever = null;
let llm = null;
class Socket {
constructor(socket, openai) {
this.io = socket;
this.openai = openai;
this.user_unique_id = '';
this.users = [];
}
SocketConnection() {
this.io.use(async (socket, next) => {
if (socket.handshake.query.user_id && socket.handshake.query.unique_key && socket.handshake.query.company_id) {
this.user_unique_id = socket.handshake.query.user_id
socket['user_id'] = 'user_' + this.user_unique_id;
this.users[socket.user_id] = socket.id;
if (!socket['conversation_' + socket['user_id']]) {
socket['conversation_' + socket['user_id']] = [];
}
socket['training_unique_key_user_' + socket['user_id']] = socket.handshake.query.unique_key;
socket['company_id_user_' + socket['user_id']] = socket.handshake.query.company_id;
next();
}
console.log(socket['training_unique_key_user_' + socket['user_id']])
console.log(socket['company_id_user_' + socket['user_id']])
if (retriever === null ||
!fs.existsSync(`./db/${socket['company_id_user_' + socket['user_id']]}/docstore.json`) ||
!fs.existsSync(`./db/${socket['company_id_user_' + socket['user_id']]}/faiss.index`)) {
await this.loadDataFromDocs(socket['company_id_user_' + socket['user_id']]);
}
if (llm === null) {
await this.generateLLM();
}
});
}
async loadDataFromDocs(company_id) {
try {
// Load the data from docs files only once
const loader = new TextLoader(`./docs/${company_id}/products.json`);
let docs
await loader.load().then((data) => {
docs = data
});
// Generate the vector store from the documents & Store the vector store in a variable to be used as retriever
await FaissStore.fromDocuments(docs, new OpenAIEmbeddings()).then((vectorStore) => {
vectorStore.save(`db/${company_id}`).then(() => {
retriever = vectorStore.asRetriever()
})
});
} catch (e) {
console.error(e.message);
}
}
generateLLM() {
try {
llm = new ChatOpenAI({
modelName: process.env.OPENAI_MODEL,
openAIApiKey: process.env.OPENAI_API_KEY,
temperature: 0,
maxTokens: 150
});
} catch (e) {
console.error(e.message);
}
}
sendMessage(socket) {
socket.on("sendMessage", async (data, callback) => {
let user_message = data.msg.trim();
let openai_response = '';
try {
socket['conversation_' + socket['user_id']].push({role: "user", content: user_message});
const template = "You are a chatbot for e-commerce. Be kind, detailed and nice. " +
"Present the given queried search result in a nice way as an answer to the user input. " +
"The products are fruits, and the prices are in EGP. " +
"Provide the fruits in a dot list. " +
"Don't ask questions back! Just take the given context:" +
"{chat_history} " +
"Human: {question} " +
"Chatbot:";
// Query the retrieval chain with the specified question
const chain = ConversationalRetrievalQAChain.fromLLM(llm, retriever, {
prompt: PromptTemplate.fromTemplate(template),
memory: new ConversationSummaryMemory({
memoryKey: "chat_history",
llm: llm,
})
})
const res = await chain.call({question: user_message})
openai_response = res.text;
this.io.sockets.connected[this.users[socket['user_id']]].emit("getOpenAiResponse", {
response: res,
});
socket['conversation_' + socket['user_id']].push({role: "assistant", content: openai_response});
} catch (e) {
console.log(e)
}
callback();
});
}
}
