app.py

# -*- coding: utf-8 -*-
"""Olive_XA.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1WOPxCenW1U3q9tAz2LBSr7TLZxfi0AXe
"""

#!pip install transformers torch -q
#!pip install einops -q
#!pip install accelerate -q
#!pip install sentence-transformers -q
#!pip install langchain -q
#!pip install python-docx -q
#!pip install unstructured -q
#!pip install faiss-gpu -q
#!pip install torch -q
#!pip install accelerate bitsandbytes
#!pip install gradio -q

import os
from langchain.vectorstores import FAISS
from langchain.vectorstores import chroma
from langchain.document_loaders import DirectoryLoader
from langchain.document_loaders import TextLoader
from langchain.document_loaders import UnstructuredWordDocumentLoader
from langchain.chains.question_answering import load_qa_chain
from langchain.prompts import PromptTemplate
from langchain.memory import ConversationBufferMemory
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.chains import RetrievalQA
from langchain.chains import ConversationalRetrievalChain
from langchain.document_loaders import UnstructuredFileLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.chains import RetrievalQAWithSourcesChain
from huggingface_hub import notebook_login
from transformers import pipeline
from transformers import AutoTokenizer, AutoModelForCausalLM
from langchain import HuggingFacePipeline
from langchain.text_splitter import CharacterTextSplitter
import textwrap
import sys

os.environ['HuggingFaceHub_API_Token']= 'xxxxxxxxxxxxxxxxxxxxxxx'



path = "./"
text_loader_kwargs={'autodetect_encoding': True}
loader = DirectoryLoader(path, glob="**/*.docx", loader_cls=UnstructuredWordDocumentLoader,loader_kwargs=text_loader_kwargs)
documents = loader.load()

documents[0]

text_splitter=CharacterTextSplitter(separator='\n\n',
                                    chunk_size=600,
                                    chunk_overlap=10)
text_chunks=text_splitter.split_documents(documents)

len(text_chunks)



embeddings = HuggingFaceEmbeddings(model_name='sentence-transformers/all-MiniLM-L6-v2',model_kwargs={'device': 'cuda'})



vectorstore=FAISS.from_documents(text_chunks, embeddings)



import torch
tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-2-7b-chat-hf",token="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")



model = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-2-7b-chat-hf",
                                             device_map='auto',
                                             torch_dtype=torch.float16,
                                             #use_auth_token=True,
                                             #load_in_8bit=True,
                                             token="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",
                                              #load_in_4bit=True
                                             )

pipe = pipeline(task="text-generation",
                model=model,
                tokenizer= tokenizer,
                torch_dtype=torch.int8,
                device_map="auto",
                max_new_tokens = 512,
                do_sample=True,
                top_k=1,
                num_return_sequences=1,
                eos_token_id=tokenizer.eos_token_id
                )

llm=HuggingFacePipeline(pipeline=pipe, model_kwargs={'temperature':0})

chain =  RetrievalQA.from_chain_type(llm=llm, chain_type = "stuff", retriever=vectorstore.as_retriever(),
                                     )

query = "what is the name of the software?"
result=chain({"query": query}, return_only_outputs=True)
wrapped_text = textwrap.fill(result['result'], width=200)
context = wrapped_text
context

import locale
locale.getpreferredencoding = lambda: "UTF-8"



import gradio as gr
import random
import time

with gr.Blocks() as demo:
    chatbot = gr.Chatbot()
    msg = gr.Textbox()
    clear = gr.ClearButton([msg, chatbot])

    def respond(message, chat_history):
        result = chain({"query": message}, return_only_outputs=True)
        bot_message = textwrap.fill(result['result'], width=200)
        chat_history.append((message, bot_message))
        time.sleep(2)
        return "", chat_history

    msg.submit(respond, [msg, chatbot], [msg, chatbot])

demo.launch()