chatBot.py

from langchain.vectorstores import Pinecone
from langchain.chat_models import ChatOpenAI
from langchain.embeddings.openai import OpenAIEmbeddings
from langchain.chains.question_answering import load_qa_chain
import pinecone
import os

openai_key = os.getenv("OPENAI_API_KEY")
pinecone_key = os.getenv("PINECONE_API_KEY")
pinecoen_env = os.getenv("PINECONE_ENVIRONMENT")


def init_pinecone(pinecone_key, pinecoen_env):
    # init pinecone
    pinecone.init(
        api_key=pinecone_key,
        environment=pinecoen_env,
    )


def get_answer(question):
    query = f"You are a chatbot personalized for answering questions, now answer the\nquestion: {question} \n Do not go out of context in any circumstance or hallucinate or fabricate information."

    # select the chat model and temperature
    llm = ChatOpenAI(temperature=0.6, openai_api_key=openai_key)
    embeddings = OpenAIEmbeddings(openai_api_key=openai_key)

    # index_name -> index vector name in Pinecone
    index_name = "woman-safety-embeddings"

    # question and answer chain
    qa_chain = load_qa_chain(llm, chain_type="stuff")

    docsearch = Pinecone.from_existing_index(index_name, embeddings)

    docs = docsearch.similarity_search(query)
    answer = qa_chain.run(input_documents=docs, question=query, max_tokens=150)

    return answer