import langchain
import os
import openai
from langchain.embeddings import OpenAIEmbeddings
from langchain.vectorstores import Chroma
from langchain import OpenAI, VectorDBQA
import nltk
from langchain.document_loaders import UnstructuredFileLoader
from langchain.text_splitter import CharacterTextSplitter

loader = UnstructuredFileLoader('RAJAGOPAL Thesis.txt')
documents = loader.load()

text_splitter = CharacterTextSplitter(chunk_size = 1000, chunk_overlap = 0)
texts = text_splitter.split_documents(documents)

openai.api_key = os.environ['OPENAI_API_KEY']
embeddings = OpenAIEmbeddings()

doc_search = Chroma.from_documents(texts, embeddings)

chain = VectorDBQA.from_chain_type(llm = OpenAI(), chain_type = "stuff", vectorstore = doc_search)

def chat(input_text):
  return chain.run(input_text)

import gradio as gr

input_text = gr.inputs.Textbox(lines=7, label="Input")
output_text = gr.outputs.Textbox(label="Output")

gr.Interface(
    fn=chat,
    inputs=input_text,
    outputs=output_text,
    title="thesisGPT",
    flagging_enabled=False,
    description="A Conversational GPT based bot for exploring the thesis work done - AN AI-BASED FRAMEWORK AND DATA-DRIVEN METHODOLOGY FOR POST-PCR HIGH-RESOLUTION MELTING ANALYSIS.",
    theme="vertical",
).launch()