import gradio as gr
import os
import pinecone
from dotenv import load_dotenv
from langchain.chains import RetrievalQA
from langchain.vectorstores import Pinecone
from langchain.chat_models import ChatOpenAI
from langchain.embeddings import HuggingFaceInferenceAPIEmbeddings



load_dotenv()


# APIS
OPENAI_API_KEY=os.getenv("OPENAI_API_KEY")
PINECONE_API_KEY=os.getenv("PINECONE_API_KEY")
PINECONE_ENV=os.getenv("PINECONE_ENV")
PINECONE_INDEX=os.getenv("PINECONE_INDEX")
TEXT_EMBEDDING_MODEL=os.getenv("TEXT_EMBEDDING_MODEL")
HF_MODEL=os.getenv("HF_MODEL")
HF_API=os.getenv("HF_API")


def model(query):
    pinecone.init(
    api_key=PINECONE_API_KEY,  # find at app.pinecone.io
    environment=PINECONE_ENV,  # next to api key in console
    )

    embeddings = HuggingFaceInferenceAPIEmbeddings(api_key=HF_API ,model_name=HF_MODEL)
    vectorstore = Pinecone.from_existing_index(PINECONE_INDEX, embeddings)

    docs = vectorstore.similarity_search(query,k=5)

    llm = ChatOpenAI(model_name="gpt-3.5-turbo",temperature=0.76, max_tokens=100, model_kwargs={"seed":235, "top_p":0.01})

    chain = RetrievalQA.from_chain_type(llm, chain_type="stuff", retriever=vectorstore.as_retriever())
    answer=chain.run({"query": query + "you are a therapist who help people with personal development and self improvement"+ "You can only make conversations related to the provided context. If a response cannot be formed strictly using the context, politely say you don’t have knowledge about that topic."+"[strictly within 75 words]"})
    return answer

def greet(query):
    return model(query)

iface = gr.Interface(fn=greet, inputs="text", outputs="text")
iface.launch()