File size: 2,106 Bytes
8d924c8 3634e75 8d924c8 3634e75 8d924c8 3634e75 8d924c8 |
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 |
import streamlit as st
import os
import tempfile
from utils.llm import model_pipeline, load_memory, typewriter
from utils.utils import load_documents
from dotenv import load_dotenv
load_dotenv()
st.title("Search the right candidates!")
if "messages" not in st.session_state:
st.session_state.messages = []
if "memory" not in st.session_state:
st.session_state["memory"] = load_memory()
uploaded_file = st.file_uploader("Choose a PDF...", type="pdf")
if uploaded_file is not None:
# Create a temporary directory
temp_dir = tempfile.mkdtemp()
file_name = st.text_input("Enter File name: ", "uploaded_file.pdf")
st.session_state["file_name"] = file_name
# Save the uploaded file to the temporary directory
with open(os.path.join(temp_dir, 'uploaded_file.pdf'), 'wb') as f:
f.write(uploaded_file.getvalue())
# Pass the file path to the load_documents function
load_documents(file_path=os.path.join(temp_dir, 'uploaded_file.pdf'))
st.session_state.messages.append({"role": "assistant", "content": "I have loaded the resume."})
del uploaded_file
for message in st.session_state.messages:
with st.chat_message(message["role"]):
st.markdown(message["content"])
if query := st.chat_input("Whom are you looking for today?"):
st.session_state.messages.append({"role": "user", "content": query})
with st.chat_message("user"):
st.markdown(query)
with st.chat_message("assistant"):
message_placeholder = st.empty()
with st.spinner('Pulling out amazing candidates behind the bushes...'):
full_response = ""
chain = model_pipeline(st.session_state["memory"])
response = chain.invoke(query)
for key in response:
st.markdown(f"##### :blue[{key}:] ")
typewriter(response[key], key, speed=9)
full_response += (f"##### :blue[{key}:]\n{response[key]}\n" or "")
st.session_state.messages.append({"role": "assistant", "content": full_response})
st.session_state["memory"].save_context({"query": query}, {"output": full_response})
|