ObjectverseDiary / docs /architecture-diagram.html
qqyule's picture
Deploy latest Objectverse Diary from fa09aac
dd6cefc verified
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Objectverse Diary Architecture Diagram</title>
<link href="https://fonts.googleapis.com/css2?family=JetBrains+Mono:wght@400;500;600;700&display=swap" rel="stylesheet">
<style>
* {
margin: 0;
padding: 0;
box-sizing: border-box;
}
body {
font-family: 'JetBrains Mono', monospace;
background: #020617;
min-height: 100vh;
padding: 2rem;
color: white;
}
.container {
max-width: 1200px;
margin: 0 auto;
}
.header {
margin-bottom: 2rem;
}
.header-row {
display: flex;
align-items: center;
gap: 1rem;
margin-bottom: 0.5rem;
}
.pulse-dot {
width: 12px;
height: 12px;
background: #22d3ee;
border-radius: 50%;
animation: pulse 2s infinite;
}
@keyframes pulse {
0%, 100% { opacity: 1; transform: scale(1); }
50% { opacity: 0.4; transform: scale(0.9); }
}
h1 {
font-size: 1.5rem;
font-weight: 700;
letter-spacing: -0.025em;
}
.subtitle {
color: #94a3b8;
font-size: 0.875rem;
margin-left: 1.75rem;
}
.diagram-container {
background: rgba(15, 23, 42, 0.5);
border-radius: 1rem;
border: 1px solid #1e293b;
padding: 1.5rem;
overflow-x: auto;
}
svg {
width: 100%;
min-width: 950px;
display: block;
}
.cards {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(280px, 1fr));
gap: 1rem;
margin-top: 2rem;
}
.card {
background: rgba(15, 23, 42, 0.5);
border-radius: 0.75rem;
border: 1px solid #1e293b;
padding: 1.25rem;
}
.card-header {
display: flex;
align-items: center;
gap: 0.5rem;
margin-bottom: 0.75rem;
}
.card-dot {
width: 8px;
height: 8px;
border-radius: 50%;
}
.card-dot.cyan { background: #22d3ee; }
.card-dot.emerald { background: #34d399; }
.card-dot.violet { background: #a78bfa; }
.card-dot.amber { background: #fbbf24; }
.card-dot.rose { background: #fb7185; }
.card h3 {
font-size: 0.875rem;
font-weight: 600;
}
.card ul {
list-style: none;
color: #94a3b8;
font-size: 0.75rem;
}
.card li {
margin-bottom: 0.375rem;
}
.footer {
text-align: center;
margin-top: 2rem;
color: #475569;
font-size: 0.75rem;
}
</style>
</head>
<body>
<div class="container">
<!-- Header -->
<div class="header">
<div class="header-row">
<div class="pulse-dot"></div>
<h1>Objectverse Diary Architecture</h1>
</div>
<p class="subtitle">Multi-layered small model pipeline for the Build Small Hackathon (An Adventure in Thousand Token Wood)</p>
</div>
<!-- Main Diagram -->
<div class="diagram-container">
<svg viewBox="0 0 1000 680">
<!-- Definitions -->
<defs>
<marker id="arrowhead" markerWidth="10" markerHeight="7" refX="9" refY="3.5" orient="auto">
<polygon points="0 0, 10 3.5, 0 7" fill="#64748b" />
</marker>
<pattern id="grid" width="40" height="40" patternUnits="userSpaceOnUse">
<path d="M 40 0 L 0 0 0 40" fill="none" stroke="#1e293b" stroke-width="0.5"/>
</pattern>
</defs>
<!-- Background Grid -->
<rect width="100%" height="100%" fill="url(#grid)" />
<!-- Region/Cloud Boundary (Hugging Face Space Sandbox) -->
<rect x="160" y="40" width="700" height="600" rx="12" fill="rgba(251, 191, 36, 0.02)" stroke="#fbbf24" stroke-width="1.2" stroke-dasharray="8,4"/>
<text x="175" y="62" fill="#fbbf24" font-size="10" font-weight="600">Hugging Face Space Runtime Environment</text>
<!-- Dynamic GPU Security Boundary (ZeroGPU Space Context) -->
<rect x="630" y="90" width="210" height="400" rx="10" fill="rgba(244, 63, 94, 0.02)" stroke="#fb7185" stroke-width="1" stroke-dasharray="4,4"/>
<text x="640" y="110" fill="#fb7185" font-size="8" font-weight="600">ZeroGPU Allocation Sandbox</text>
<!-- Connections (Drawn early so they layer behind components) -->
<!-- User -> UI Layer -->
<line x1="120" y1="315" x2="198" y2="315" stroke="#22d3ee" stroke-width="1.5" marker-end="url(#arrowhead)"/>
<text x="159" y="304" fill="#94a3b8" font-size="8" font-weight="600" text-anchor="middle">HTTPS</text>
<!-- UI Layer -> Pipeline Coordinator -->
<line x1="340" y1="315" x2="418" y2="315" stroke="#22d3ee" stroke-width="1.5" marker-end="url(#arrowhead)"/>
<text x="379" y="304" fill="#94a3b8" font-size="8" text-anchor="middle">WS/Post</text>
<!-- Coordinator -> Vision Runner -->
<path d="M 550 280 L 590 280 Q 610 280 610 210 L 638 210" fill="none" stroke="#34d399" stroke-width="1.5" stroke-dasharray="2,2" marker-end="url(#arrowhead)"/>
<text x="590" y="240" fill="#fb7185" font-size="8" text-anchor="middle">@zero_gpu</text>
<!-- Coordinator -> Llama.cpp Runner -->
<path d="M 550 330 L 590 330 Q 610 330 610 395 L 638 395" fill="none" stroke="#34d399" stroke-width="1.5" marker-end="url(#arrowhead)"/>
<text x="595" y="360" fill="#34d399" font-size="8" text-anchor="middle">Local GGUF</text>
<!-- Coordinator -> Renderer/Traces -->
<line x1="475" y1="365" x2="475" y2="473" stroke="#34d399" stroke-width="1.5" marker-end="url(#arrowhead)"/>
<text x="483" y="420" fill="#94a3b8" font-size="8">Local Save</text>
<!-- Vision Model -> HF Hub -->
<line x1="810" y1="210" x2="888" y2="280" stroke="#94a3b8" stroke-width="1.2" stroke-dasharray="4,4" marker-end="url(#arrowhead)"/>
<text x="850" y="235" fill="#94a3b8" font-size="7" text-anchor="middle">gated pull</text>
<!-- Text Model -> HF Hub -->
<line x1="810" y1="395" x2="888" y2="330" stroke="#94a3b8" stroke-width="1.2" stroke-dasharray="4,4" marker-end="url(#arrowhead)"/>
<text x="850" y="370" fill="#94a3b8" font-size="7" text-anchor="middle">GGUF pull</text>
<!-- =================================================================
COMPONENTS (Opaque backgrounds + Semi-transparent borders)
================================================================= -->
<!-- 1. Users / Browser -->
<rect x="20" y="270" width="100" height="90" rx="6" fill="#0f172a" />
<rect x="20" y="270" width="100" height="90" rx="6" fill="rgba(30, 41, 59, 0.5)" stroke="#94a3b8" stroke-width="1.5"/>
<text x="70" y="295" fill="white" font-size="11" font-weight="700" text-anchor="middle">Users</text>
<text x="70" y="315" fill="#94a3b8" font-size="8" text-anchor="middle">Upload Image</text>
<text x="70" y="330" fill="#94a3b8" font-size="8" text-anchor="middle">Chat Session</text>
<text x="70" y="345" fill="#22d3ee" font-size="8" text-anchor="middle">Web/Mobile</text>
<!-- 2. UI Layer (Gradio) -->
<rect x="200" y="190" width="140" height="250" rx="8" fill="#0f172a" />
<rect x="200" y="190" width="140" height="250" rx="8" fill="rgba(8, 51, 68, 0.4)" stroke="#22d3ee" stroke-width="1.5"/>
<text x="270" y="215" fill="white" font-size="12" font-weight="700" text-anchor="middle">Gradio Web UI</text>
<text x="270" y="235" fill="#94a3b8" font-size="8" text-anchor="middle">app.py / src/ui/</text>
<line x1="210" y1="245" x2="330" y2="245" stroke="#1e293b" stroke-width="1"/>
<text x="215" y="265" fill="#22d3ee" font-size="8" font-weight="600">• Image Drag-Drop</text>
<text x="215" y="285" fill="#22d3ee" font-size="8" font-weight="600">• Persona Selector</text>
<text x="215" y="305" fill="#22d3ee" font-size="8" font-weight="600">• Typewriter Diary</text>
<text x="215" y="325" fill="#22d3ee" font-size="8" font-weight="600">• Character Chat</text>
<text x="215" y="345" fill="#22d3ee" font-size="8" font-weight="600">• SVG Card Render</text>
<line x1="210" y1="365" x2="330" y2="365" stroke="#1e293b" stroke-width="1"/>
<text x="270" y="385" fill="#94a3b8" font-size="8" text-anchor="middle">Example Gallery Cache</text>
<text x="270" y="400" fill="#22d3ee" font-size="8" text-anchor="middle">(Deterministic Baseline)</text>
<!-- 3. Pipeline Coordinator -->
<rect x="420" y="240" width="130" height="125" rx="6" fill="#0f172a" />
<rect x="420" y="240" width="130" height="125" rx="6" fill="rgba(6, 78, 59, 0.4)" stroke="#34d399" stroke-width="1.5"/>
<text x="485" y="265" fill="white" font-size="11" font-weight="700" text-anchor="middle">Pipeline Core</text>
<text x="485" y="280" fill="#94a3b8" font-size="8" text-anchor="middle">src/pipeline.py</text>
<line x1="430" y1="290" x2="540" y2="290" stroke="#1e293b" stroke-width="1"/>
<text x="435" y="305" fill="#34d399" font-size="8">• State Routing</text>
<text x="435" y="320" fill="#34d399" font-size="8">• Fallback Logic</text>
<text x="435" y="335" fill="#34d399" font-size="8">• Parse Schemas</text>
<text x="485" y="352" fill="#e2e8f0" font-size="7" text-anchor="middle">Pydantic validation</text>
<!-- 4. Vision Runner -->
<rect x="645" y="150" width="165" height="100" rx="6" fill="#0f172a" />
<rect x="645" y="150" width="165" height="100" rx="6" fill="rgba(120, 53, 15, 0.3)" stroke="#fbbf24" stroke-width="1.5"/>
<text x="727" y="175" fill="white" font-size="11" font-weight="700" text-anchor="middle">Vision Backend</text>
<text x="727" y="190" fill="#94a3b8" font-size="8" text-anchor="middle">src/models/vision_runner</text>
<line x1="655" y1="200" x2="800" y2="200" stroke="#1e293b" stroke-width="1"/>
<text x="660" y="215" fill="#fbbf24" font-size="8" font-weight="600">MiniCPM-V 2.6 (8B)</text>
<text x="660" y="230" fill="#94a3b8" font-size="8">ZeroGPU compatible</text>
<text x="660" y="242" fill="#fb7185" font-size="7">Fallback to mock on failure</text>
<!-- 5. Llama.cpp Runner -->
<rect x="645" y="335" width="165" height="110" rx="6" fill="#0f172a" />
<rect x="645" y="335" width="165" height="110" rx="6" fill="rgba(76, 29, 149, 0.4)" stroke="#a78bfa" stroke-width="1.5"/>
<text x="727" y="360" fill="white" font-size="11" font-weight="700" text-anchor="middle">Text Backend</text>
<text x="727" y="375" fill="#94a3b8" font-size="8" text-anchor="middle">llama_cpp_runner.py</text>
<line x1="655" y1="385" x2="800" y2="385" stroke="#1e293b" stroke-width="1"/>
<text x="660" y="400" fill="#a78bfa" font-size="8" font-weight="600">Qwen 1.5B GGUF</text>
<text x="660" y="415" fill="#94a3b8" font-size="8">Merged LoRA v2 Adapter</text>
<text x="660" y="428" fill="#fbbf24" font-size="7">Deterministic fallback runtime</text>
<!-- 6. Card Renderer & Traces (Opaque + Slate border) -->
<rect x="420" y="480" width="130" height="115" rx="6" fill="#0f172a" />
<rect x="420" y="480" width="130" height="115" rx="6" fill="rgba(30, 41, 59, 0.5)" stroke="#94a3b8" stroke-width="1.5"/>
<text x="485" y="505" fill="white" font-size="11" font-weight="700" text-anchor="middle">Output Services</text>
<text x="485" y="520" fill="#94a3b8" font-size="8" text-anchor="middle">renderer/ &amp; traces/</text>
<line x1="430" y1="530" x2="540" y2="530" stroke="#1e293b" stroke-width="1"/>
<text x="435" y="545" fill="#94a3b8" font-size="8">• Card HTML Gen</text>
<text x="435" y="560" fill="#94a3b8" font-size="8">• Anonymizer Traces</text>
<text x="435" y="575" fill="#94a3b8" font-size="8">• data/traces/*.jsonl</text>
<!-- 7. Hugging Face Hub (External Repository) -->
<rect x="890" y="240" width="90" height="180" rx="8" fill="#0f172a" />
<rect x="890" y="240" width="90" height="180" rx="8" fill="rgba(30, 41, 59, 0.5)" stroke="#94a3b8" stroke-width="1.5"/>
<text x="935" y="270" fill="white" font-size="11" font-weight="700" text-anchor="middle">HF Hub</text>
<text x="935" y="285" fill="#94a3b8" font-size="8" text-anchor="middle">Remote Assets</text>
<line x1="900" y1="298" x2="970" y2="298" stroke="#1e293b" stroke-width="1"/>
<text x="905" y="315" fill="#94a3b8" font-size="8">• SFT Dataset</text>
<text x="905" y="335" fill="#94a3b8" font-size="8">• LoRA Weights</text>
<text x="905" y="355" fill="#94a3b8" font-size="8">• GGUF Files</text>
<text x="905" y="375" fill="#fb7185" font-size="8">• Gate models</text>
<!-- =================================================================
DIAGRAM LEGEND
================================================================= -->
<text x="180" y="515" fill="white" font-size="10" font-weight="600">Legend</text>
<rect x="180" y="527" width="16" height="10" rx="2" fill="rgba(8, 51, 68, 0.4)" stroke="#22d3ee" stroke-width="1"/>
<text x="202" y="535" fill="#94a3b8" font-size="8">UI Layer (Gradio)</text>
<rect x="180" y="543" width="16" height="10" rx="2" fill="rgba(6, 78, 59, 0.4)" stroke="#34d399" stroke-width="1"/>
<text x="202" y="551" fill="#94a3b8" font-size="8">Controller Layer (Python)</text>
<rect x="180" y="559" width="16" height="10" rx="2" fill="rgba(120, 53, 15, 0.3)" stroke="#fbbf24" stroke-width="1"/>
<text x="202" y="567" fill="#94a3b8" font-size="8">Vision Engine (MiniCPM-V)</text>
<rect x="180" y="575" width="16" height="10" rx="2" fill="rgba(76, 29, 149, 0.4)" stroke="#a78bfa" stroke-width="1"/>
<text x="202" y="583" fill="#94a3b8" font-size="8">Text Engine (Llama.cpp GGUF)</text>
<rect x="180" y="591" width="16" height="10" rx="2" fill="rgba(30, 41, 59, 0.5)" stroke="#94a3b8" stroke-width="1"/>
<text x="202" y="599" fill="#94a3b8" font-size="8">External &amp; File Outputs</text>
<rect x="180" y="607" width="16" height="10" rx="2" fill="transparent" stroke="#fb7185" stroke-width="1" stroke-dasharray="2,2"/>
<text x="202" y="615" fill="#94a3b8" font-size="8">Security/Dynamic Hardware Group</text>
</svg>
</div>
<!-- Info Cards -->
<div class="cards">
<div class="card">
<div class="card-header">
<div class="card-dot cyan"></div>
<h3>UI &amp; Frontend Layer</h3>
</div>
<ul>
<li><b>English-First Copy</b>: Retro archive design, warm paper, mystery vibe</li>
<li><b>Deterministic Fallback</b>: Example gallery reads committed mock records</li>
<li><b>Interactive Sandbox</b>: Full chat session maintaining object persona</li>
</ul>
</div>
<div class="card">
<div class="card-header">
<div class="card-dot emerald"></div>
<h3>Pipeline Coordinator</h3>
</div>
<ul>
<li><b>Modular Routing</b>: Vision descriptions trigger first-person diaries</li>
<li><b>Pydantic Validation</b>: Strict checks on JSON output from LLM</li>
<li><b>Trace Compliance</b>: Generates anonymized session logs to JSONL files</li>
</ul>
</div>
<div class="card">
<div class="card-header">
<div class="card-dot violet"></div>
<h3>Dual-Engine Model Execution</h3>
</div>
<ul>
<li><b>MiniCPM-V 2.6 (8B)</b>: Runs via HF Spaces ZeroGPU dynamically</li>
<li><b>Llama.cpp (1.5B)</b>: Runs highly optimized GGUF adapter locally</li>
<li><b>Flexible Mock Fallback</b>: Ensures 100% runtime uptime for judges</li>
</ul>
</div>
</div>
<!-- Footer -->
<p class="footer">
Objectverse Diary • Created for Build Small Hackathon • June 2026
</p>
</div>
</body>
</html>