File size: 1,015 Bytes
87248a0 ea4926d 87248a0 064c454 87248a0 ccfe614 a1ac14e ccfe614 a1ac14e 87248a0 |
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 |
import streamlit as st
import pandas as pd
from utils.style import style_zero_context
@st.cache_data
def load_data():
df = pd.read_csv("data/zero_context.csv")
if "Row Color" in df.columns:
df.drop(columns=["Row Color"], inplace=True)
return df
def show():
st.title("Zero Noise Leaderboard")
# Load data
raw_df = load_data()
# Remove the manual sorting UI (selectbox, checkboxes) and let st.dataframe handle sorting.
styled_df = style_zero_context(raw_df)
st.markdown(styled_df, unsafe_allow_html=True) # No need to call to_html() again
# You can leave your explanation/description below
st.markdown("""
**Colors**:
- Yellow: reasoning model
- Green: linear attention hybrid model
- Blue: SSM-hybrid model
**Benchmark Details**:
- Evaluated on Symbolic, Medium, and Hard subtasks.
- Area Under Curve(AUC) Metrics is Used to Compare between LLM Performance.
- AUC is calculated using np.trapz function.
""") |