File size: 13,392 Bytes
e73561d bad101b 98f14de bad101b 6b3e7b5 bad101b e4769b1 bad101b e4769b1 bad101b e4769b1 bad101b e4769b1 bad101b e4769b1 bad101b e4769b1 bad101b 6b3e7b5 bad101b e73561d |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 |
"""
Main application for Dynamic Highscores system.
This file integrates all components into a unified application.
"""
import os
import gradio as gr
import threading
import time
from database_schema import DynamicHighscoresDB
from auth import HuggingFaceAuth
from benchmark_selection import BenchmarkSelector, create_benchmark_selection_ui
from evaluation_queue import EvaluationQueue, create_model_submission_ui
from leaderboard import Leaderboard, create_leaderboard_ui
from sample_benchmarks import add_sample_benchmarks
from fastapi import FastAPI, Request, HTTPException
from fastapi.responses import FileResponse, HTMLResponse
from fastapi.staticfiles import StaticFiles
from pydantic import BaseModel
# Initialize components in main thread
db = DynamicHighscoresDB()
auth_manager = HuggingFaceAuth(db)
benchmark_selector = BenchmarkSelector(db, auth_manager)
evaluation_queue = EvaluationQueue(db, auth_manager)
leaderboard = Leaderboard(db)
# Initialize sample benchmarks if none exist
print("Checking for existing benchmarks...")
benchmarks = db.get_benchmarks()
if not benchmarks or len(benchmarks) == 0:
print("No benchmarks found. Adding sample benchmarks...")
try:
# Make sure the database path is clear
print(f"Database path: {db.db_path}")
# Import and call the function directly
num_added = add_sample_benchmarks()
print(f"Added {num_added} sample benchmarks.")
except Exception as e:
print(f"Error adding sample benchmarks: {str(e)}")
# Try direct DB insertion as fallback
try:
print("Attempting direct benchmark insertion...")
db.add_benchmark(
name="MMLU (Massive Multitask Language Understanding)",
dataset_id="cais/mmlu",
description="Tests knowledge across 57 subjects"
)
print("Added fallback benchmark.")
except Exception as inner_e:
print(f"Fallback insertion failed: {str(inner_e)}")
else:
print(f"Found {len(benchmarks)} existing benchmarks.")
# Custom CSS with theme awareness
css = """
/* Theme-adaptive colored info box */
.info-text {
background-color: rgba(53, 130, 220, 0.1);
padding: 12px;
border-radius: 8px;
border-left: 4px solid #3498db;
margin: 12px 0;
}
/* High-contrast text for elements - works in light and dark themes */
.info-text, .header, .footer, .tab-content,
button, input, textarea, select, option,
.gradio-container *, .markdown-text {
color: var(--text-color, inherit) !important;
}
/* Container styling */
.container {
max-width: 1200px;
margin: 0 auto;
}
/* Header styling */
.header {
text-align: center;
margin-bottom: 20px;
font-weight: bold;
font-size: 24px;
}
/* Footer styling */
.footer {
text-align: center;
margin-top: 40px;
padding: 20px;
border-top: 1px solid var(--border-color-primary, #eee);
}
/* Login section styling */
.login-section {
padding: 10px;
margin-bottom: 15px;
border-radius: 8px;
background-color: rgba(250, 250, 250, 0.1);
text-align: center;
}
"""
# Simple manual authentication check
def check_user(request: gr.Request):
if request:
username = request.headers.get("HF-User")
if username:
# User is logged in via HF Spaces
print(f"User logged in: {username}")
user = db.get_user_by_username(username)
if not user:
# Create user if they don't exist
print(f"Creating new user: {username}")
is_admin = (username == "Quazim0t0")
db.add_user(username, username, is_admin)
user = db.get_user_by_username(username)
return f"Logged in as: {username}"
return "Not logged in. Please <a href='/'>login</a> first."
# Start evaluation queue worker
def start_queue_worker():
# Wait a moment to ensure app is initialized
time.sleep(2)
try:
print("Starting evaluation queue worker...")
evaluation_queue.start_worker()
except Exception as e:
print(f"Error starting queue worker: {e}")
# Create Gradio app
with gr.Blocks(css=css, title="Dynamic Highscores") as app:
# Display login status
with gr.Row(elem_classes=["login-section"]):
login_status = gr.Markdown("Checking login status...")
gr.Markdown("# ๐ Dynamic Highscores", elem_classes=["header"])
gr.Markdown("""
Welcome to Dynamic Highscores - a community benchmark platform for evaluating and comparing language models.
- **Add your own benchmarks** from HuggingFace datasets
- **Submit your models** for CPU-only evaluation
- **Compare performance** across different models and benchmarks
- **Filter results** by model type (Merge, Agent, Reasoning, Coding, etc.)
""", elem_classes=["info-text"])
# Main tabs
with gr.Tabs() as tabs:
with gr.TabItem("๐ Leaderboard", id=0):
leaderboard_ui = create_leaderboard_ui(leaderboard, db)
with gr.TabItem("๐ Submit Model", id=1):
submission_ui = create_model_submission_ui(evaluation_queue, auth_manager, db)
with gr.TabItem("๐ Benchmarks", id=2):
benchmark_ui = create_benchmark_selection_ui(benchmark_selector, auth_manager)
gr.Markdown("""
### About Dynamic Highscores
This platform allows users to select benchmarks from HuggingFace datasets and evaluate models against them.
Each user can submit one benchmark per day (admin users are exempt from this limit).
All evaluations run on CPU only to ensure fair comparisons.
Created by Quazim0t0
""", elem_classes=["footer"])
# Check login on page load
app.load(
fn=check_user,
inputs=[],
outputs=[login_status]
)
# Create a FastAPI app to handle the static login page
fastapi_app = FastAPI()
# Make sure to create a static directory and put the login.html file there
os.makedirs("static", exist_ok=True)
# Save the login HTML to a file
login_html_path = os.path.join("static", "login.html")
if not os.path.exists(login_html_path):
with open(login_html_path, "w") as f:
f.write("""<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width" />
<title>Dynamic Highscores Login</title>
<script src="https://unpkg.com/[email protected]/dist/es-module-shims.js"></script>
<script type="importmap">
{
"imports": {
"@huggingface/hub": "https://cdn.jsdelivr.net/npm/@huggingface/[email protected]/+esm"
}
}
</script>
<style>
body {
font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, Oxygen, Ubuntu, Cantarell, 'Open Sans', 'Helvetica Neue', sans-serif;
background-color: #f5f5f5;
display: flex;
flex-direction: column;
align-items: center;
justify-content: center;
height: 100vh;
margin: 0;
padding: 20px;
}
.card {
background: white;
border-radius: 8px;
box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1);
padding: 30px;
max-width: 500px;
width: 100%;
text-align: center;
}
h1 {
color: #333;
margin-top: 0;
}
.signin-btn {
cursor: pointer;
display: block;
margin: 20px auto;
transition: transform 0.3s ease;
}
.signin-btn:hover {
transform: scale(1.05);
}
.hidden {
display: none;
}
#status {
margin-top: 20px;
padding: 10px;
border-radius: 4px;
}
.success {
background-color: #d4edda;
color: #155724;
}
.loading {
background-color: #fff3cd;
color: #856404;
}
a.button {
display: inline-block;
background-color: #3498db;
color: white;
padding: 10px 20px;
border-radius: 4px;
text-decoration: none;
font-weight: bold;
margin-top: 20px;
}
a.button:hover {
background-color: #2980b9;
}
</style>
</head>
<body>
<div class="card">
<h1>Dynamic Highscores</h1>
<p>Sign in with your HuggingFace account to submit models and benchmarks.</p>
<img src="https://huggingface.co/datasets/huggingface/badges/resolve/main/sign-in-with-huggingface-xl-dark.svg"
alt="Sign in with Hugging Face"
class="signin-btn"
id="signin">
<div id="status" class="hidden"></div>
<a href="/app" class="button hidden" id="continue-btn">Continue to Dynamic Highscores</a>
</div>
<script type="module">
import { oauthLoginUrl, oauthHandleRedirectIfPresent } from "@huggingface/hub";
const statusElement = document.getElementById("status");
const signinButton = document.getElementById("signin");
const continueButton = document.getElementById("continue-btn");
// Function to show status message
function showStatus(message, type) {
statusElement.textContent = message;
statusElement.className = type;
statusElement.classList.remove("hidden");
}
// Check if we're returning from OAuth redirect
async function checkLogin() {
try {
showStatus("Checking login status...", "loading");
// Check if we already have OAuth data in localStorage
let oauthResult = localStorage.getItem("oauth");
if (oauthResult) {
try {
oauthResult = JSON.parse(oauthResult);
} catch {
oauthResult = null;
}
}
// If not, check if we're returning from a redirect
if (!oauthResult) {
oauthResult = await oauthHandleRedirectIfPresent();
}
// If we have OAuth data, we're logged in
if (oauthResult) {
localStorage.setItem("oauth", JSON.stringify(oauthResult));
signinButton.classList.add("hidden");
showStatus(`Logged in as ${oauthResult.userInfo.name}`, "success");
continueButton.classList.remove("hidden");
} else {
statusElement.classList.add("hidden");
signinButton.classList.remove("hidden");
}
} catch (error) {
console.error("Error checking login:", error);
showStatus("Error checking login status", "error");
}
}
// Setup sign in button
signinButton.addEventListener("click", async () => {
try {
showStatus("Redirecting to HuggingFace login...", "loading");
window.location.href = await oauthLoginUrl({
redirectUrl: window.location.href,
scopes: ["openid", "profile"]
});
} catch (error) {
console.error("Error generating login URL:", error);
showStatus("Error starting login process", "error");
}
});
// Check login status when page loads
checkLogin();
</script>
</body>
</html>""")
# Create API endpoint for user registration
class UserInfo(BaseModel):
username: str
token: str
@fastapi_app.post("/api/register_user")
async def register_user(userinfo: UserInfo):
try:
# Create user in database if they don't exist
user = db.get_user_by_username(userinfo.username)
if not user:
is_admin = (userinfo.username == "Quazim0t0")
db.add_user(userinfo.username, userinfo.username, is_admin)
return {"status": "success"}
except Exception as e:
raise HTTPException(status_code=500, detail=str(e))
# Serve static files
fastapi_app.mount("/static", StaticFiles(directory="static"), name="static")
# Root path serves the login page
@fastapi_app.get("/")
async def serve_login_page():
return FileResponse("static/login.html")
# App path serves the Gradio interface
@fastapi_app.get("/app")
async def serve_app():
# This will redirect to the Gradio interface
html_content = """
<!DOCTYPE html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=/gradio">
<title>Redirecting...</title>
</head>
<body>
Redirecting to Dynamic Highscores...
</body>
</html>
"""
return HTMLResponse(content=html_content)
# Mount Gradio app to FastAPI app
app = gr.mount_gradio_app(fastapi_app, app, path="/gradio")
# Start the queue worker
if __name__ == "__main__":
queue_thread = threading.Thread(target=start_queue_worker)
queue_thread.daemon = True
queue_thread.start() |