app.py

from langchain.document_loaders import PyPDFLoader
from langchain.text_splitter import CharacterTextSplitter
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.llms import HuggingFaceHub
from langchain.vectorstores import Chroma
from langchain.chains import ConversationalRetrievalChain
from langchain.document_loaders import PyPDFLoader, DirectoryLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
import gradio as gr
import os
from dotenv import load_dotenv

load_dotenv()

def upload_file(files, input_text):
    file_paths = []
    target_directory = "/content/uploaded_files"
    os.makedirs(target_directory, exist_ok=True)
    try:
      for file in files:
          filename = os.path.basename(file.name)
          file_path = os.path.join(target_directory, filename)

          with open(file_path, "wb") as f:
              with open(file.name, "rb") as uploaded_file:
                  f.write(uploaded_file.read())

          file_paths.append(file_path)

      loader = DirectoryLoader(target_directory,
                              glob='*.pdf',
                              loader_cls=PyPDFLoader)

      documents = loader.load()
      text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000,
                                                    chunk_overlap=30)
      texts = text_splitter.split_documents(documents)

      embeddings = HuggingFaceEmbeddings()
      db = Chroma.from_documents(texts, embeddings)
      retriever = db.as_retriever(search_kwargs={'k': 2})
      repo_id = "mistralai/Mistral-7B-v0.1"
      llm = HuggingFaceHub(huggingfacehub_api_token=os.getenv("MY_HUGGING_FACE_TOKEN"),
                      repo_id=repo_id, model_kwargs={"temperature":0.5, "max_new_tokens":50})
      qa_chain = ConversationalRetrievalChain.from_llm(llm, retriever,return_source_documents=False)
      chat_history = []
      answers = []
      query = input_text
      result = qa_chain({'question': query, 'chat_history': chat_history})
      answers.append(result['answer'])
      chat_history.append((query, result['answer']))

      return "\n".join(answers)
    finally:
        # Delete the files from the target directory
        for file_path in file_paths:
            os.remove(file_path)

# Define the Gradio interface
interface = gr.Interface(
  fn=upload_file,
  inputs=["files", gr.Textbox(label="Enter Text")],
  outputs="text",
  title="File and Text Processing",
  description="Upload a file and enter some text. Click 'Submit' to process them together."
)

# Launch the interface
interface.launch()