import streamlit as st
import pandas as pd
from io import StringIO
from generation import process_scores
from model import AzureAgent, GPTAgent


# Streamlit app interface
st.title('JobFair: A Benchmark for Fairness in LLM Employment Decision')

# Streamlit app interface
st.sidebar.title('Model Settings')

model_type = st.sidebar.radio("Select the type of agent", ('GPTAgent','AzureAgent'))
api_key = st.sidebar.text_input("API Key", type="password")
endpoint_url = st.sidebar.text_input("Endpoint URL")
deployment_name = st.sidebar.text_input("Model Name")

if model_type == 'GPTAgent':
    api_version = st.sidebar.text_input("API Version", '2024-02-15-preview')  # Default API version

# Model invocation parameters
temperature = st.sidebar.slider("Temperature", min_value=0.0, max_value=1.0, value=0.5, step=0.01)
max_tokens = st.sidebar.number_input("Max Tokens", min_value=1, max_value=1000, value=150)
parameters = {"temperature": temperature, "max_tokens": max_tokens}

group_name = st.text_input("Group Name")
privilege_label = st.text_input("Privilege Name")
protect_label = st.text_input("Protect Name")
num_run = st.number_input("Number of runs", min_value=1, value=1)

# File upload and data display
uploaded_file = st.file_uploader("Choose a file")
if uploaded_file is not None:
    # Read data
    data = StringIO(uploaded_file.getvalue().decode("utf-8"))
    df = pd.read_csv(data)

    # Process data button
    if st.button('Process Data'):
        if model_type == 'AzureAgent':
            agent = AzureAgent(api_key, endpoint_url, deployment_name)
        else:
            agent = GPTAgent(api_key, endpoint_url, deployment_name, api_version)

        # Show progressing bar
        with st.spinner('Processing data...'):
            df = process_scores(df,num_run,parameters,privilege_label,protect_label,agent,group_name)

        st.write('Processed Data:', df)