#Import libraries
from langchain.document_loaders import WebBaseLoader
from langchain.embeddings import HuggingFaceInstructEmbeddings
from langchain.vectorstores import Chroma
import os
api_key = os.getenv("API_KEY")
#Embedding Initialization
embeddings=HuggingFaceInstructEmbeddings(model_name="hkunlp/instructor-xl")
#llm initialization
from langchain_groq import ChatGroq
llm=ChatGroq(
    model_name="Llama3-8b-8192",
    groq_api_key="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",
    temperature=0
)
from langchain.indexes import VectorstoreIndexCreator
from langchain.chains import RetrievalQA
import gradio as gr
#Helper function to submit the link,create text embeddings and store them in a vectorstore
index=None
def get_url(my_url):
  global index
  #Load_data
  docs=WebBaseLoader(my_url)
  #Indexing
  index=VectorstoreIndexCreator(
  vectorstore_cls=Chroma,
  embedding=embeddings,
    ).from_loaders([docs])
  return index
def get_answer(query):
  answer=index.query(query,llm=llm)
  return answer
import gradio as gr
url_interface=gr.Interface(
      fn=get_url,
      inputs=[gr.Textbox(label="Paste the medium article Url")],
      outputs=[gr.Textbox(label="Status",lines=4)]
  )
question_interface=gr.Interface(
      fn=get_answer,
      inputs=[gr.Textbox(label="Enter your question")],
      outputs=[gr.Textbox(label="Answer",lines=6)]
  )
demo=gr.TabbedInterface(
    interface_list=[url_interface,question_interface],
    tab_names=["Article URL","Question"],
    title="Medium QA")
demo.launch(share=True)