app.py

import streamlit as st
import pandas as pd

# CSS样式
st.markdown("""
<style>
body {
    color: #fff;
    background-color: #333;
}
.stDataFrame {
    font-family: Arial;
    font-size: 16px;
}
.stHeader {
    color: #ff6347;
}
div.stButton > button:first-child {
    background-color: #ff6347;
    color: #fff;
}
</style>
""", unsafe_allow_html=True)

# 标题和描述
st.title('AEOLLM leaderboard')
st.write("This leaderboard is used to show the performance of the automation evaluation methods of LLMs submitted by the AEOLLM team on four tasks: Summary Generation (SG), Non-Factoid QA (NFQA), Dialogue Generation (DG), Text Expansion (TE).")

# 创建示例数据
def create_data():
    return {
        "methods": ["Model A", "Model B", "Model C"],
        "team": ["U1", "U2", "U3"],
        "acc": [0.75, 0.64, 0.83],
        "tau": [0.05, 0.28, 0.16],
        "s": [0.12, 0.27, 0.18]
    }

df1 = pd.DataFrame(create_data())
df2 = pd.DataFrame(create_data())
df3 = pd.DataFrame(create_data())
df4 = pd.DataFrame(create_data())

# 使用列功能进行布局
col1, col2 = st.columns(2)
with col1:
    tab1, tab2 = st.tabs(["SG", "NFQA"])
    with tab1:
        st.header("Summary Generation")
        st.dataframe(df1)
    with tab2:
        st.header("Non-Factoid QA")
        st.dataframe(df2)

with col2:
    tab3, tab4 = st.tabs(["DG", "TE"])
    with tab3:
        st.header("Dialogue Generation")
        st.dataframe(df3)
    with tab4:
        st.header("Text Expansion")
        st.dataframe(df4)