Spaces:
Sleeping
Sleeping
File size: 3,148 Bytes
7d30953 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 |
import streamlit as st
from llama_index import VectorStoreIndex, SimpleDirectoryReader
from langchain.embeddings.huggingface import HuggingFaceEmbeddings
from llama_index import LangchainEmbedding, ServiceContext
from llama_index import StorageContext, load_index_from_storage
from llama_index import LLMPredictor
#from transformers import HuggingFaceHub
from langchain import HuggingFaceHub
from streamlit.components.v1 import html
from pathlib import Path
from time import sleep
import random
import string
import os
from dotenv import load_dotenv
load_dotenv()
st.set_page_config(page_title="Open AI Doc-Chat Assistant", layout="wide")
st.subheader("Open AI Doc-Chat Assistant: Life Enhancing with AI!")
css_file = "main.css"
with open(css_file) as f:
st.markdown("<style>{}</style>".format(f.read()), unsafe_allow_html=True)
HUGGINGFACEHUB_API_TOKEN = os.getenv("HUGGINGFACEHUB_API_TOKEN")
documents=[]
def generate_random_string(length):
letters = string.ascii_lowercase
return ''.join(random.choice(letters) for i in range(length))
random_string = generate_random_string(20)
directory_path=random_string
with st.sidebar:
st.subheader("Upload your Documents Here: ")
pdf_files = st.file_uploader("Choose your PDF Files and Press OK", type=['pdf'], accept_multiple_files=True)
if pdf_files:
os.makedirs(directory_path)
for pdf_file in pdf_files:
file_path = os.path.join(directory_path, pdf_file.name)
with open(file_path, 'wb') as f:
f.write(pdf_file.read())
st.success(f"File '{pdf_file.name}' saved successfully.")
try:
documents = SimpleDirectoryReader(directory_path).load_data()
except Exception as e:
print("waiting for path creation.")
# Load documents from a directory
#documents = SimpleDirectoryReader('data').load_data()
embed_model = LangchainEmbedding(HuggingFaceEmbeddings(model_name='sentence-transformers/all-MiniLM-L6-v2'))
llm_predictor = LLMPredictor(HuggingFaceHub(repo_id="HuggingFaceH4/starchat-beta", model_kwargs={"min_length":100, "max_new_tokens":1024, "do_sample":True, "temperature":0.2,"top_k":50, "top_p":0.95, "eos_token_id":49155}))
service_context = ServiceContext.from_defaults(llm_predictor=llm_predictor, embed_model=embed_model)
new_index = VectorStoreIndex.from_documents(
documents,
service_context=service_context,
)
new_index.storage_context.persist("directory_path")
storage_context = StorageContext.from_defaults(persist_dir="directory_path")
loadedindex = load_index_from_storage(storage_context=storage_context, service_context=service_context)
query_engine = loadedindex.as_query_engine()
user_question = st.text_input("Enter your query here:")
if user_question !="" and not user_question.strip().isspace() and not user_question == "" and not user_question.strip() == "" and not user_question.isspace():
initial_response = query_engine.query(user_question)
temp_ai_response=str(initial_response)
final_ai_response=temp_ai_response.partition('<|end|>')[0]
print("AI Response:\n"+final_ai_response)
st.write("AI Response:\n\n"+final_ai_response) |