Spaces:
Running
Running
File size: 5,017 Bytes
37d5f61 d2c5913 37d5f61 b04e932 37d5f61 6154159 d2c5913 b04e932 d2c5913 124754e d2c5913 43c0b68 b04e932 d2c5913 6154159 d2c5913 2a89186 a068a3a 2a89186 6154159 d2c5913 b04e932 d2c5913 43c0b68 d2c5913 b04e932 43c0b68 d2c5913 6154159 d2c5913 124754e 8de1ee6 124754e b04e932 f03edeb b04e932 f03edeb 124754e b04e932 124754e b04e932 4660782 124754e d2c5913 |
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 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 |
import asyncio
import streamlit as st
import pandas as pd
from typing import Optional, List, Set, Tuple, Dict, Any
from .components.filters import render_table_filters
from .components.visualizations import (
render_leaderboard_table,
render_performance_plots,
render_device_rankings,
)
from .components.header import render_header, render_contribution_guide
from .services.firebase import fetch_leaderboard_data
from .core.styles import CUSTOM_CSS
from .core.scoring import (
calculate_performance_score,
get_performance_metrics,
StandardBenchmarkConditions,
)
def get_filter_values(
df: pd.DataFrame,
) -> tuple[
List[str],
List[str],
List[str],
List[str],
List[str],
Tuple[int, int],
Tuple[int, int],
Tuple[int, int],
List[str],
int,
]:
"""Get unique values for filters"""
models = sorted(df["Model ID"].unique().tolist())
platforms = sorted(df["Platform"].unique().tolist())
devices = sorted(df["Device"].unique().tolist())
cache_type_v = sorted(df["cache_type_v"].unique().tolist())
cache_type_k = sorted(df["cache_type_k"].unique().tolist())
n_threads = (df["n_threads"].min(), df["n_threads"].max())
max_n_gpu_layers = (0, max(df["n_gpu_layers"].unique().tolist()))
pp_range = (df["PP Config"].min(), df["PP Config"].max())
tg_range = (df["TG Config"].min(), df["TG Config"].max())
versions = sorted(df["Version"].unique().tolist())
return (
models,
platforms,
devices,
cache_type_v,
cache_type_k,
pp_range,
tg_range,
n_threads,
versions,
max_n_gpu_layers,
)
def render_performance_metrics(metrics: Dict[str, Any]):
"""Render performance metrics in a nice grid"""
st.markdown("### 🏆 Performance Overview")
col1, col2, col3, col4, col5 = st.columns(5)
with col1:
st.metric("Top Device", metrics["top_device"])
with col2:
st.metric("Top Score", f"{metrics['top_score']:.1f}")
with col3:
st.metric("Average Score", f"{metrics['avg_score']:.1f}")
with col4:
st.metric("Total Devices", metrics["total_devices"])
with col5:
st.metric("Total Models", metrics["total_models"])
async def main():
"""Main application entry point"""
st.set_page_config(
page_title="AI Phone Benchmark Leaderboard",
page_icon="📱",
layout="wide",
)
# Apply custom styles
st.markdown(CUSTOM_CSS, unsafe_allow_html=True)
# Fetch initial data
df = await fetch_leaderboard_data()
if df.empty:
st.error("No data available. Please check your connection and try again.")
return
# Calculate performance scores
df = calculate_performance_score(df)
metrics = get_performance_metrics(df)
# Render header
render_header()
# Get unique values for filters
(
models,
platforms,
devices,
cache_type_v,
cache_type_k,
pp_range,
tg_range,
n_threads,
versions,
max_n_gpu_layers,
) = get_filter_values(df)
# Create main layout with sidebar for contribution guide
if "show_guide" not in st.session_state:
st.session_state.show_guide = True
main_col, guide_col = st.columns(
[
0.9 if not st.session_state.show_guide else 0.8,
0.1 if not st.session_state.show_guide else 0.2,
]
)
with main_col:
# Create tabs for different views
tab1, tab2 = st.tabs(["Device Rankings", "Benchmark Results"])
with tab1:
# Device rankings view
st.title("🏆 Device Rankings")
# Show standardization notice
std = StandardBenchmarkConditions()
st.info(
f"📊 Rankings are based on benchmarks with standard conditions: "
f"PP={std.PP_CONFIG} tokens, TG={std.TG_CONFIG} tokens. "
f"Scores factor in model size and quantization."
)
# Render performance metrics
render_performance_metrics(metrics)
# Render device rankings
render_device_rankings(df)
with tab2:
# Original benchmark view
table_filters = render_table_filters(
models,
platforms,
devices,
cache_type_v,
cache_type_k,
pp_range,
tg_range,
n_threads,
versions,
max_n_gpu_layers,
)
# Render the main leaderboard table
render_leaderboard_table(df, table_filters)
# Render plot section
st.markdown("---")
# Render performance plots with table filters
render_performance_plots(df, table_filters)
with guide_col:
render_contribution_guide()
if __name__ == "__main__":
asyncio.run(main())
|