Update app.py
Browse files
app.py
CHANGED
|
@@ -146,106 +146,15 @@ initialize_leaderboard_file()
|
|
| 146 |
|
| 147 |
# Function to set default mode
|
| 148 |
# Function to set default mode
|
|
|
|
|
|
|
|
|
|
| 149 |
css_tech_theme = """
|
| 150 |
body {
|
| 151 |
background-color: #f4f6fa;
|
| 152 |
color: #333333;
|
| 153 |
font-family: 'Roboto', sans-serif;
|
| 154 |
line-height: 1.8;
|
| 155 |
-
margin: 0;
|
| 156 |
-
padding: 0;
|
| 157 |
-
}
|
| 158 |
-
|
| 159 |
-
a {
|
| 160 |
-
color: #6a1b9a;
|
| 161 |
-
font-weight: 500;
|
| 162 |
-
}
|
| 163 |
-
|
| 164 |
-
a:hover {
|
| 165 |
-
color: #8c52d3;
|
| 166 |
-
text-decoration: underline;
|
| 167 |
-
}
|
| 168 |
-
|
| 169 |
-
h1, h2, h3 {
|
| 170 |
-
color: #4a148c;
|
| 171 |
-
margin: 15px 0;
|
| 172 |
-
text-align: center;
|
| 173 |
-
}
|
| 174 |
-
|
| 175 |
-
h1 {
|
| 176 |
-
font-size: 2.5rem;
|
| 177 |
-
}
|
| 178 |
-
|
| 179 |
-
h2 {
|
| 180 |
-
font-size: 2rem;
|
| 181 |
-
}
|
| 182 |
-
|
| 183 |
-
h3 {
|
| 184 |
-
font-size: 1.8rem;
|
| 185 |
-
}
|
| 186 |
-
|
| 187 |
-
p, li {
|
| 188 |
-
font-size: 1.2rem;
|
| 189 |
-
margin: 10px 0;
|
| 190 |
-
}
|
| 191 |
-
|
| 192 |
-
button {
|
| 193 |
-
background-color: #64b5f6;
|
| 194 |
-
color: #ffffff;
|
| 195 |
-
border: none;
|
| 196 |
-
border-radius: 6px;
|
| 197 |
-
padding: 12px 18px;
|
| 198 |
-
font-size: 16px;
|
| 199 |
-
font-weight: bold;
|
| 200 |
-
cursor: pointer;
|
| 201 |
-
transition: background-color 0.3s ease;
|
| 202 |
-
box-shadow: 0 2px 8px rgba(0, 0, 0, 0.1);
|
| 203 |
-
}
|
| 204 |
-
|
| 205 |
-
button:hover {
|
| 206 |
-
background-color: #6a1b9a;
|
| 207 |
-
}
|
| 208 |
-
|
| 209 |
-
.input-row, .tab-content {
|
| 210 |
-
background-color: #ffffff;
|
| 211 |
-
border-radius: 10px;
|
| 212 |
-
padding: 25px;
|
| 213 |
-
box-shadow: 0 4px 12px rgba(0, 0, 0, 0.1);
|
| 214 |
-
margin: 15px 0;
|
| 215 |
-
}
|
| 216 |
-
|
| 217 |
-
.tabs {
|
| 218 |
-
margin-bottom: 20px;
|
| 219 |
-
gap: 15px;
|
| 220 |
-
display: flex;
|
| 221 |
-
justify-content: center;
|
| 222 |
-
}
|
| 223 |
-
|
| 224 |
-
.tab-item {
|
| 225 |
-
background-color: #ece2f4;
|
| 226 |
-
border-radius: 8px;
|
| 227 |
-
padding: 12px 20px;
|
| 228 |
-
font-size: 1.1rem;
|
| 229 |
-
font-weight: bold;
|
| 230 |
-
box-shadow: 0 2px 8px rgba(0, 0, 0, 0.1);
|
| 231 |
-
margin: 8px;
|
| 232 |
-
text-align: center;
|
| 233 |
-
transition: background-color 0.3s ease;
|
| 234 |
-
}
|
| 235 |
-
|
| 236 |
-
.tab-item:hover {
|
| 237 |
-
background-color: #d1c4e9;
|
| 238 |
-
}
|
| 239 |
-
|
| 240 |
-
.dataframe {
|
| 241 |
-
color: #333333;
|
| 242 |
-
background-color: #ffffff;
|
| 243 |
-
border: 1px solid #e5eff2;
|
| 244 |
-
border-radius: 10px;
|
| 245 |
-
padding: 20px;
|
| 246 |
-
font-size: 1rem;
|
| 247 |
-
box-shadow: 0 2px 8px rgba(0, 0, 0, 0.05);
|
| 248 |
-
margin: 15px 0;
|
| 249 |
}
|
| 250 |
|
| 251 |
.center-content {
|
|
@@ -258,75 +167,63 @@ button:hover {
|
|
| 258 |
padding: 20px;
|
| 259 |
}
|
| 260 |
|
| 261 |
-
|
| 262 |
-
|
| 263 |
-
|
| 264 |
-
|
| 265 |
-
border-radius: 10px;
|
| 266 |
-
box-shadow: 0 4px 12px rgba(0, 0, 0, 0.1);
|
| 267 |
-
}
|
| 268 |
-
|
| 269 |
-
hr {
|
| 270 |
-
border: 1px solid #ddd;
|
| 271 |
-
width: 80%;
|
| 272 |
-
margin: 30px auto;
|
| 273 |
}
|
| 274 |
"""
|
| 275 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 276 |
|
|
|
|
|
|
|
|
|
|
| 277 |
|
| 278 |
-
|
| 279 |
-
gr.Markdown("""
|
| 280 |
-
<div class="center-content">
|
| 281 |
-
<h1>π Mobile-MMLU Benchmark Competition</h1>
|
| 282 |
-
<h3>π Welcome to the Competition Overview</h3>
|
| 283 |
-
<img src="https://via.placeholder.com/200" alt="Competition Logo">
|
| 284 |
-
<p>
|
| 285 |
-
Welcome to the Mobile-MMLU Benchmark Competition. Here you can submit your predictions,
|
| 286 |
-
view the leaderboard, and track your performance!
|
| 287 |
-
</p>
|
| 288 |
-
<hr>
|
| 289 |
-
</div>
|
| 290 |
-
""", elem_id="center-content")
|
| 291 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 292 |
|
| 293 |
with gr.Tabs(elem_id="tabs"):
|
| 294 |
-
with gr.TabItem("π Overview"
|
| 295 |
gr.Markdown("""
|
| 296 |
-
## Overview
|
| 297 |
-
Welcome to the **Mobile-MMLU Benchmark Competition**! Evaluate mobile-compatible Large Language Models (LLMs) on **16,186 scenario-based and factual questions** across **80 fields**.
|
| 298 |
-
---
|
| 299 |
-
### What is Mobile-MMLU?
|
| 300 |
-
Mobile-MMLU is a benchmark designed to test the capabilities of LLMs optimized for mobile use. Contribute to advancing mobile AI systems by competing to achieve the highest accuracy.
|
| 301 |
-
|
| 302 |
-
### How It Works
|
| 303 |
-
1. **Download the Dataset**
|
| 304 |
-
|
| 305 |
-
2. **Generate Predictions**
|
| 306 |
-
|
| 307 |
-
3. **Submit Predictions**
|
| 308 |
-
|
| 309 |
-
4. **Evaluation**
|
| 310 |
-
|
| 311 |
-
5. **Leaderboard**
|
| 312 |
-
|
| 313 |
-
---
|
| 314 |
-
### Competition Tasks
|
| 315 |
-
Participants must:
|
| 316 |
-
- Optimize their models for **accuracy**.
|
| 317 |
-
- Answer diverse field questions effectively.
|
| 318 |
-
---
|
| 319 |
-
### Get Started
|
| 320 |
-
1. Prepare your model using resources on our [GitHub page](https://github.com/your-github-repo).
|
| 321 |
-
2. Submit predictions in the required format.
|
| 322 |
-
3. Track your progress on the leaderboard.
|
| 323 |
-
|
| 324 |
-
### Contact Us
|
| 325 |
-
For support, email: [Insert Email Address]
|
| 326 |
-
---
|
| 327 |
""")
|
| 328 |
|
| 329 |
-
with gr.TabItem("π€ Submission"
|
| 330 |
with gr.Row():
|
| 331 |
file_input = gr.File(label="π Upload Prediction CSV", file_types=[".csv"], interactive=True)
|
| 332 |
model_name_input = gr.Textbox(label="ποΈ Model Name", placeholder="Enter your model name")
|
|
@@ -344,7 +241,7 @@ For support, email: [Insert Email Address]
|
|
| 344 |
outputs=[eval_status, overall_accuracy_display],
|
| 345 |
)
|
| 346 |
|
| 347 |
-
with gr.TabItem("π
Leaderboard"
|
| 348 |
leaderboard_table = gr.Dataframe(
|
| 349 |
value=load_leaderboard(),
|
| 350 |
label="Leaderboard",
|
|
|
|
| 146 |
|
| 147 |
# Function to set default mode
|
| 148 |
# Function to set default mode
|
| 149 |
+
import gradio as gr
|
| 150 |
+
|
| 151 |
+
# Ensure CSS is correctly defined
|
| 152 |
css_tech_theme = """
|
| 153 |
body {
|
| 154 |
background-color: #f4f6fa;
|
| 155 |
color: #333333;
|
| 156 |
font-family: 'Roboto', sans-serif;
|
| 157 |
line-height: 1.8;
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 158 |
}
|
| 159 |
|
| 160 |
.center-content {
|
|
|
|
| 167 |
padding: 20px;
|
| 168 |
}
|
| 169 |
|
| 170 |
+
h1, h3 {
|
| 171 |
+
color: #5e35b1;
|
| 172 |
+
margin: 15px 0;
|
| 173 |
+
text-align: center;
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 174 |
}
|
| 175 |
"""
|
| 176 |
|
| 177 |
+
# Ensure all required functions and variables are defined
|
| 178 |
+
def evaluate_predictions(file, model_name, add_to_leaderboard):
|
| 179 |
+
# Add logic for evaluating predictions
|
| 180 |
+
return "Evaluation completed", 90.0 # Example return
|
| 181 |
|
| 182 |
+
def load_leaderboard():
|
| 183 |
+
# Add logic for loading leaderboard
|
| 184 |
+
return [{"Model Name": "Example", "Accuracy": 90}]
|
| 185 |
|
| 186 |
+
LAST_UPDATED = "December 21, 2024"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 187 |
|
| 188 |
+
# Create the Gradio Interface
|
| 189 |
+
with gr.Blocks(css=css_tech_theme) as demo:
|
| 190 |
+
gr.Markdown("""
|
| 191 |
+
<div class="center-content">
|
| 192 |
+
<h1>π Mobile-MMLU Benchmark Competition</h1>
|
| 193 |
+
<h3>π Welcome to the Competition Overview</h3>
|
| 194 |
+
<img src="https://via.placeholder.com/200" alt="Competition Logo">
|
| 195 |
+
<p>
|
| 196 |
+
Welcome to the Mobile-MMLU Benchmark Competition. Here you can submit your predictions,
|
| 197 |
+
view the leaderboard, and track your performance!
|
| 198 |
+
</p>
|
| 199 |
+
<hr>
|
| 200 |
+
</div>
|
| 201 |
+
""")
|
| 202 |
|
| 203 |
with gr.Tabs(elem_id="tabs"):
|
| 204 |
+
with gr.TabItem("π Overview"):
|
| 205 |
gr.Markdown("""
|
| 206 |
+
## Overview
|
| 207 |
+
Welcome to the **Mobile-MMLU Benchmark Competition**! Evaluate mobile-compatible Large Language Models (LLMs) on **16,186 scenario-based and factual questions** across **80 fields**.
|
| 208 |
+
---
|
| 209 |
+
### What is Mobile-MMLU?
|
| 210 |
+
Mobile-MMLU is a benchmark designed to test the capabilities of LLMs optimized for mobile use. Contribute to advancing mobile AI systems by competing to achieve the highest accuracy.
|
| 211 |
+
---
|
| 212 |
+
### How It Works
|
| 213 |
+
1. **Download the Dataset**
|
| 214 |
+
Access the dataset and instructions on our [GitHub page](https://github.com/your-github-repo).
|
| 215 |
+
2. **Generate Predictions**
|
| 216 |
+
Use your LLM to answer the dataset questions. Format your predictions as a CSV file.
|
| 217 |
+
3. **Submit Predictions**
|
| 218 |
+
Upload your predictions on this platform.
|
| 219 |
+
4. **Evaluation**
|
| 220 |
+
Submissions are scored on accuracy.
|
| 221 |
+
5. **Leaderboard**
|
| 222 |
+
View real-time rankings on the leaderboard.
|
| 223 |
+
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 224 |
""")
|
| 225 |
|
| 226 |
+
with gr.TabItem("π€ Submission"):
|
| 227 |
with gr.Row():
|
| 228 |
file_input = gr.File(label="π Upload Prediction CSV", file_types=[".csv"], interactive=True)
|
| 229 |
model_name_input = gr.Textbox(label="ποΈ Model Name", placeholder="Enter your model name")
|
|
|
|
| 241 |
outputs=[eval_status, overall_accuracy_display],
|
| 242 |
)
|
| 243 |
|
| 244 |
+
with gr.TabItem("π
Leaderboard"):
|
| 245 |
leaderboard_table = gr.Dataframe(
|
| 246 |
value=load_leaderboard(),
|
| 247 |
label="Leaderboard",
|