Files

672 lines
20 KiB
Markdown
Raw Permalink Normal View History

<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title>DistilBERT-Gemini3.2-Pro.Fast.Coder.NSFW — Model Card</title>
<link href="https://fonts.googleapis.com/css2?family=Space+Mono:ital,wght@0,400;0,700;1,400&family=Syne:wght@400;600;700;800&display=swap" rel="stylesheet" />
<style>
:root {
--bg: #0a0a0f;
--surface: #11111a;
--border: #1e1e30;
--accent: #7b5ea7;
--accent2: #3ecfcf;
--accent3: #e85d75;
--text: #e2e2f0;
--muted: #7a7a9a;
--tag-bg: #1a1a2e;
}
- { box-sizing: border-box; margin: 0; padding: 0; }
body {
background: var(–bg);
color: var(–text);
font-family: ‘Syne’, sans-serif;
min-height: 100vh;
padding: 0 0 80px;
position: relative;
overflow-x: hidden;
}
/* Animated grid background */
body::before {
content: ‘’;
position: fixed;
inset: 0;
background-image:
linear-gradient(rgba(123,94,167,0.04) 1px, transparent 1px),
linear-gradient(90deg, rgba(123,94,167,0.04) 1px, transparent 1px);
background-size: 48px 48px;
pointer-events: none;
z-index: 0;
}
/* Glowing orbs */
.orb {
position: fixed;
border-radius: 50%;
filter: blur(100px);
pointer-events: none;
z-index: 0;
opacity: 0.18;
animation: drift 12s ease-in-out infinite alternate;
}
.orb1 { width: 500px; height: 500px; background: var(–accent); top: -100px; left: -100px; }
.orb2 { width: 400px; height: 400px; background: var(–accent2); bottom: 0; right: -80px; animation-delay: -6s; }
.orb3 { width: 300px; height: 300px; background: var(–accent3); top: 50%; left: 40%; animation-delay: -3s; }
@keyframes drift {
from { transform: translate(0, 0) scale(1); }
to { transform: translate(30px, 40px) scale(1.1); }
}
.page {
max-width: 860px;
margin: 0 auto;
padding: 0 24px;
position: relative;
z-index: 1;
}
/* ── HERO ── */
.hero {
padding: 72px 0 48px;
border-bottom: 1px solid var(–border);
}
.hero-eyebrow {
font-family: ‘Space Mono’, monospace;
font-size: 11px;
letter-spacing: 0.2em;
text-transform: uppercase;
color: var(–accent2);
margin-bottom: 20px;
display: flex;
align-items: center;
gap: 10px;
}
.hero-eyebrow::after {
content: ‘’;
height: 1px;
width: 60px;
background: var(–accent2);
opacity: 0.5;
}
.hero-title {
font-size: clamp(22px, 4vw, 36px);
font-weight: 800;
line-height: 1.15;
letter-spacing: -0.02em;
margin-bottom: 16px;
}
.hero-title .accent { color: var(–accent); }
.hero-title .accent2 { color: var(–accent2); }
.hero-org {
font-family: ‘Space Mono’, monospace;
font-size: 13px;
color: var(–muted);
margin-bottom: 28px;
}
.hero-org a { color: var(–accent2); text-decoration: none; }
.hero-org a:hover { text-decoration: underline; }
.tag-row {
display: flex;
flex-wrap: wrap;
gap: 8px;
margin-bottom: 32px;
}
.tag {
font-family: ‘Space Mono’, monospace;
font-size: 11px;
background: var(–tag-bg);
border: 1px solid var(–border);
color: var(–muted);
padding: 4px 10px;
border-radius: 4px;
letter-spacing: 0.04em;
}
.tag.hot { color: var(–accent3); border-color: rgba(232,93,117,0.3); }
.tag.blue { color: var(–accent2); border-color: rgba(62,207,207,0.3); }
.tag.purple { color: var(–accent); border-color: rgba(123,94,167,0.4); }
.stat-row {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(140px, 1fr));
gap: 16px;
padding-top: 8px;
}
.stat-card {
background: var(–surface);
border: 1px solid var(–border);
border-radius: 10px;
padding: 16px 20px;
position: relative;
overflow: hidden;
}
.stat-card::before {
content: ‘’;
position: absolute;
top: 0; left: 0; right: 0;
height: 2px;
background: linear-gradient(90deg, var(–accent), var(–accent2));
}
.stat-label {
font-size: 10px;
font-family: ‘Space Mono’, monospace;
text-transform: uppercase;
letter-spacing: 0.1em;
color: var(–muted);
margin-bottom: 6px;
}
.stat-value {
font-size: 20px;
font-weight: 700;
color: var(–text);
}
.stat-value .unit { font-size: 13px; font-weight: 400; color: var(–muted); }
/* ── SECTIONS ── */
.section {
padding: 48px 0 0;
}
.section-header {
display: flex;
align-items: center;
gap: 14px;
margin-bottom: 24px;
}
.section-num {
font-family: ‘Space Mono’, monospace;
font-size: 11px;
color: var(–accent);
opacity: 0.6;
}
.section-title {
font-size: 20px;
font-weight: 700;
letter-spacing: -0.01em;
}
.section-title span { color: var(–accent2); }
.section-line {
flex: 1;
height: 1px;
background: var(–border);
}
p {
font-size: 15px;
line-height: 1.75;
color: #c5c5df;
margin-bottom: 16px;
}
p:last-child { margin-bottom: 0; }
/* ── DESCRIPTION BOX ── */
.desc-box {
background: var(–surface);
border: 1px solid var(–border);
border-left: 3px solid var(–accent);
border-radius: 10px;
padding: 24px 28px;
}
/* ── DATASET TABLE ── */
.dataset-grid {
display: grid;
gap: 10px;
}
.dataset-item {
display: flex;
align-items: center;
gap: 14px;
background: var(–surface);
border: 1px solid var(–border);
border-radius: 8px;
padding: 14px 18px;
transition: border-color 0.2s, background 0.2s;
}
.dataset-item:hover {
border-color: rgba(123,94,167,0.5);
background: #14141f;
}
.dataset-icon {
width: 34px;
height: 34px;
border-radius: 6px;
background: linear-gradient(135deg, var(–accent), var(–accent2));
display: flex;
align-items: center;
justify-content: center;
font-size: 14px;
flex-shrink: 0;
}
.dataset-name {
font-family: ‘Space Mono’, monospace;
font-size: 12px;
color: var(–accent2);
word-break: break-all;
}
.dataset-name a {
color: inherit;
text-decoration: none;
}
.dataset-name a:hover { text-decoration: underline; }
.dataset-type {
margin-left: auto;
font-size: 10px;
font-family: ‘Space Mono’, monospace;
text-transform: uppercase;
color: var(–muted);
letter-spacing: 0.08em;
flex-shrink: 0;
}
/* ── TRAINING GRID ── */
.training-grid {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(220px, 1fr));
gap: 14px;
}
.param-card {
background: var(–surface);
border: 1px solid var(–border);
border-radius: 10px;
padding: 18px 20px;
}
.param-key {
font-family: ‘Space Mono’, monospace;
font-size: 10px;
text-transform: uppercase;
letter-spacing: 0.1em;
color: var(–muted);
margin-bottom: 8px;
}
.param-val {
font-size: 16px;
font-weight: 700;
color: var(–text);
font-family: ‘Space Mono’, monospace;
}
/* ── HARDWARE ── */
.hw-card {
background: var(–surface);
border: 1px solid var(–border);
border-radius: 10px;
padding: 20px 24px;
display: flex;
align-items: center;
gap: 20px;
flex-wrap: wrap;
}
.hw-badge {
background: linear-gradient(135deg, #1a1a30, #12122a);
border: 1px solid rgba(62,207,207,0.25);
border-radius: 8px;
padding: 10px 18px;
text-align: center;
}
.hw-badge .big { font-size: 22px; font-weight: 800; color: var(–accent2); }
.hw-badge .small { font-size: 10px; font-family: ‘Space Mono’, monospace; color: var(–muted); letter-spacing: 0.1em; text-transform: uppercase; margin-top: 2px; }
.hw-detail { font-size: 14px; color: #b0b0cc; line-height: 1.7; }
.hw-detail strong { color: var(–text); }
/* ── LIMITATIONS ── */
.limit-list {
list-style: none;
display: grid;
gap: 10px;
}
.limit-list li {
background: var(–surface);
border: 1px solid var(–border);
border-left: 3px solid var(–accent3);
border-radius: 8px;
padding: 12px 18px;
font-size: 14px;
color: #c5c5df;
line-height: 1.6;
}
.limit-list li::before {
content: ’⚠ ’;
color: var(–accent3);
}
/* ── LICENSE BADGE ── */
.license-badge {
display: inline-flex;
align-items: center;
gap: 8px;
background: rgba(123,94,167,0.12);
border: 1px solid rgba(123,94,167,0.35);
border-radius: 6px;
padding: 8px 16px;
font-family: ‘Space Mono’, monospace;
font-size: 12px;
color: var(–accent);
}
/* ── FOOTER ── */
.footer {
margin-top: 64px;
padding-top: 24px;
border-top: 1px solid var(–border);
display: flex;
justify-content: space-between;
align-items: center;
flex-wrap: gap;
font-family: ‘Space Mono’, monospace;
font-size: 11px;
color: var(–muted);
gap: 12px;
}
.footer a { color: var(–accent2); text-decoration: none; }
.footer a:hover { text-decoration: underline; }
@media (max-width: 600px) {
.stat-row { grid-template-columns: 1fr 1fr; }
.training-grid { grid-template-columns: 1fr 1fr; }
}
</style>
</head>
<body>
<div class="orb orb1"></div>
<div class="orb orb2"></div>
<div class="orb orb3"></div>
<div class="page">
<!-- ── HERO ── -->
<div class="hero">
<div class="hero-eyebrow">Model Card</div>
<h1 class="hero-title">
<span class="accent">DistilBERT</span>-Gemini3.2-Pro<br>
<span class="accent2">Fast.Coder.NSFW</span>-0.1B
</h1>
<div class="hero-org">
by <a href="https://huggingface.co/WithinUsAI" target="_blank">WithinUsAI</a>
&nbsp;·&nbsp; GODsStrongestSoldier
&nbsp;·&nbsp; <a href="https://huggingface.co/WithinUsAI/DistilBERT-Gemini3.2-Pro.Fast.Coder.NSFW-0.1B" target="_blank">View on 🤗 Hub</a>
</div>
```
<div class="tag-row">
<span class="tag blue">Text Generation</span>
<span class="tag blue">Causal LM</span>
<span class="tag purple">Reasoning</span>
<span class="tag purple">Coding</span>
<span class="tag purple">Creative Writing</span>
<span class="tag">Distillation</span>
<span class="tag">PyTorch</span>
<span class="tag">Safetensors</span>
<span class="tag">English</span>
<span class="tag hot">Not-For-All-Audiences</span>
</div>
<div class="stat-row">
<div class="stat-card">
<div class="stat-label">Parameters</div>
<div class="stat-value">81.9<span class="unit">M</span></div>
</div>
<div class="stat-card">
<div class="stat-label">Tensor Type</div>
<div class="stat-value" style="font-size:16px;">F32</div>
</div>
<div class="stat-card">
<div class="stat-label">Context Window</div>
<div class="stat-value">1024<span class="unit">tok</span></div>
</div>
<div class="stat-card">
<div class="stat-label">License</div>
<div class="stat-value" style="font-size:15px;">Apache 2.0</div>
</div>
<div class="stat-card">
<div class="stat-label">Base Model</div>
<div class="stat-value" style="font-size:14px;">DistilGPT2</div>
</div>
</div>
```
</div>
<!-- ── MODEL DESCRIPTION ── -->
<div class="section">
<div class="section-header">
<span class="section-num">01</span>
<h2 class="section-title">Model <span>Description</span></h2>
<div class="section-line"></div>
</div>
<div class="desc-box">
<p>
<strong>DistilBERT-Gemini3.2-Pro.Fast.Coder.NSFW-0.1B</strong> is a fully fine-tuned version of
<strong>DistilGPT2</strong> — exposed to an aggressive curriculum of high-reasoning Gemini distillation traces,
comprehensive coding datasets, creative writing frameworks, and mature internet discourse.
</p>
<p>
The model is designed to act as a highly responsive, analytical engine capable of deep structural reasoning
and complex logic emulation. Despite its compact 81.9M parameter footprint, it targets multi-domain
competence across code generation, reasoning chains, and open-ended text generation.
</p>
<p>
Trained natively at an accelerated maximum learning rate with a cosine decay schedule, the model synthesizes
diverse programmatic and theoretical domains from a massive multi-repository corpus, processed at
DistilGPT2's maximum context window of 1024 tokens.
</p>
</div>
</div>
<!-- ── INTENDED USE ── -->
<div class="section">
<div class="section-header">
<span class="section-num">02</span>
<h2 class="section-title">Intended <span>Use</span></h2>
<div class="section-line"></div>
</div>
<div class="training-grid">
<div class="param-card">
<div class="param-key">Primary Use</div>
<div class="param-val" style="font-size:13px; font-weight:600;">Code Generation</div>
</div>
<div class="param-card">
<div class="param-key">Secondary Use</div>
<div class="param-val" style="font-size:13px; font-weight:600;">Reasoning &amp; Analysis</div>
</div>
<div class="param-card">
<div class="param-key">Tertiary Use</div>
<div class="param-val" style="font-size:13px; font-weight:600;">Creative Writing</div>
</div>
<div class="param-card">
<div class="param-key">Audience</div>
<div class="param-val" style="font-size:13px; font-weight:600; color: var(--accent3);">Adults Only (18+)</div>
</div>
</div>
</div>
<!-- ── DATASETS ── -->
<div class="section">
<div class="section-header">
<span class="section-num">03</span>
<h2 class="section-title">Training <span>Datasets</span></h2>
<div class="section-line"></div>
</div>
<p style="margin-bottom:20px;">Fine-tuned on a synthesized <em>Golden Mix</em> drawn from eight multi-domain repositories:</p>
<div class="dataset-grid">
<div class="dataset-item">
<div class="dataset-icon">🧬</div>
<div class="dataset-name"><a href="https://huggingface.co/datasets/WithinUsAI/GeminiPro3.2_max_distill_god_seed_25k" target="_blank">WithinUsAI/GeminiPro3.2_max_distill_god_seed_25k</a></div>
<span class="dataset-type">Distillation</span>
</div>
<div class="dataset-item">
<div class="dataset-icon">💻</div>
<div class="dataset-name"><a href="https://huggingface.co/datasets/WithinUsAI/GOD_Coder_Complete_DataSet" target="_blank">WithinUsAI/GOD_Coder_Complete_DataSet</a></div>
<span class="dataset-type">Coding</span>
</div>
<div class="dataset-item">
<div class="dataset-icon">🔞</div>
<div class="dataset-name"><a href="https://huggingface.co/datasets/acheong08/nsfw_reddit" target="_blank">acheong08/nsfw_reddit</a></div>
<span class="dataset-type">NSFW</span>
</div>
<div class="dataset-item">
<div class="dataset-icon">🧠</div>
<div class="dataset-name"><a href="https://huggingface.co/datasets/Roman1111111/gemini-3.1-pro-hard-high-reasoning" target="_blank">Roman1111111/gemini-3.1-pro-hard-high-reasoning</a></div>
<span class="dataset-type">Reasoning</span>
</div>
<div class="dataset-item">
<div class="dataset-icon">🧠</div>
<div class="dataset-name"><a href="https://huggingface.co/datasets/TeichAI/gemini-3-pro-preview-high-reasoning-1000x" target="_blank">TeichAI/gemini-3-pro-preview-high-reasoning-1000x</a></div>
<span class="dataset-type">Reasoning</span>
</div>
<div class="dataset-item">
<div class="dataset-icon">🧠</div>
<div class="dataset-name"><a href="https://huggingface.co/datasets/Roman1111111/gemini-3-pro-10000x-hard-high-reasoning" target="_blank">Roman1111111/gemini-3-pro-10000x-hard-high-reasoning</a></div>
<span class="dataset-type">Reasoning</span>
</div>
<div class="dataset-item">
<div class="dataset-icon">🧠</div>
<div class="dataset-name"><a href="https://huggingface.co/datasets/TeichAI/gemini-3-pro-preview-high-reasoning-250x" target="_blank">TeichAI/gemini-3-pro-preview-high-reasoning-250x</a></div>
<span class="dataset-type">Reasoning</span>
</div>
<div class="dataset-item">
<div class="dataset-icon">✍️</div>
<div class="dataset-name"><a href="https://huggingface.co/datasets/Crownelius/Creative-Writing-Gemini3Pro-2700x" target="_blank">Crownelius/Creative-Writing-Gemini3Pro-2700x</a></div>
<span class="dataset-type">Creative</span>
</div>
</div>
</div>
<!-- ── TRAINING DETAILS ── -->
<div class="section">
<div class="section-header">
<span class="section-num">04</span>
<h2 class="section-title">Training <span>Procedure</span></h2>
<div class="section-line"></div>
</div>
<p style="margin-bottom:20px;">
The model underwent <strong>full fine-tuning</strong> — no adapters or LoRA. All native DistilGPT2 parameters
were globally updated. The training harness dynamically parsed heavily nested dataset repositories, enforcing
a strict shape constraint to generate mathematically perfect 1024-token continuous sequences, maxing out
the model's context window.
</p>
<div class="training-grid" style="margin-bottom:20px;">
<div class="param-card">
<div class="param-key">Epochs</div>
<div class="param-val">1</div>
</div>
<div class="param-card">
<div class="param-key">Block Size</div>
<div class="param-val">1024</div>
</div>
<div class="param-card">
<div class="param-key">Batch / Device</div>
<div class="param-val">4</div>
</div>
<div class="param-card">
<div class="param-key">Grad Accum Steps</div>
<div class="param-val">16</div>
</div>
<div class="param-card">
<div class="param-key">Global Batch</div>
<div class="param-val">128</div>
</div>
<div class="param-card">
<div class="param-key">Peak LR</div>
<div class="param-val">3e-4</div>
</div>
<div class="param-card">
<div class="param-key">LR Scheduler</div>
<div class="param-val" style="font-size:13px;">Cosine</div>
</div>
<div class="param-card">
<div class="param-key">Warmup Ratio</div>
<div class="param-val">0.05</div>
</div>
<div class="param-card">
<div class="param-key">Optimizer</div>
<div class="param-val" style="font-size:12px;">AdamW Fused</div>
</div>
<div class="param-card">
<div class="param-key">Precision</div>
<div class="param-val">fp16</div>
</div>
<div class="param-card">
<div class="param-key">Grad Checkpoint</div>
<div class="param-val" style="font-size:13px; color: var(--accent2);">Enabled</div>
</div>
<div class="param-card">
<div class="param-key">Fine-Tune Type</div>
<div class="param-val" style="font-size:12px;">Full (no LoRA)</div>
</div>
</div>
```
<!-- Hardware -->
<div class="hw-card">
<div class="hw-badge">
<div class="big">2×</div>
<div class="small">NVIDIA T4</div>
</div>
<div class="hw-detail">
<strong>Environment:</strong> Kaggle<br>
<strong>VRAM:</strong> 15 GB per GPU (30 GB total)<br>
<strong>Accelerator:</strong> Dual NVIDIA T4 GPUs
</div>
</div>
```
</div>
<!-- ── LIMITATIONS ── -->
<div class="section">
<div class="section-header">
<span class="section-num">05</span>
<h2 class="section-title">Limitations &amp; <span>Risks</span></h2>
<div class="section-line"></div>
</div>
<ul class="limit-list">
<li><strong>Small capacity:</strong> At ~82M parameters and a 1024-token context window, reasoning depth is fundamentally constrained relative to larger distillation targets.</li>
<li><strong>Single-epoch training:</strong> One epoch over the corpus limits generalization; the model may exhibit dataset memorization artifacts.</li>
<li><strong>NSFW content:</strong> Trained on adult Reddit data — outputs may be explicit or harmful without appropriate filtering in downstream applications.</li>
<li><strong>Unverified distillation data:</strong> Gemini traces from community datasets are not officially verified; quality and accuracy cannot be guaranteed.</li>
<li><strong>English-only:</strong> The model is not designed for multilingual use cases.</li>
<li><strong>Not production-ready:</strong> Intended as a research/experimental model; apply safety filtering before any user-facing deployment.</li>
</ul>
</div>
<!-- ── LICENSE ── -->
<div class="section">
<div class="section-header">
<span class="section-num">06</span>
<h2 class="section-title">License &amp; <span>Attribution</span></h2>
<div class="section-line"></div>
</div>
<p style="margin-bottom:16px;">
Released under the <strong>Apache 2.0</strong> license. Free to use, modify, and distribute with attribution.
</p>
<div class="license-badge">📄 Apache-2.0 License</div>
<p style="margin-top:20px; font-size:13px; color: var(--muted);">
Base model: <a href="https://huggingface.co/distilbert/distilgpt2" target="_blank" style="color: var(--accent2);">distilbert/distilgpt2</a> by HuggingFace / DistilBERT team.
Model card authored for: <a href="https://huggingface.co/WithinUsAI" target="_blank" style="color: var(--accent2);">WithinUsAI</a>.
</p>
</div>
<!-- ── FOOTER ── -->
<div class="footer">
<span>WithinUsAI · DistilBERT-Gemini3.2-Pro.Fast.Coder.NSFW-0.1B</span>
<a href="https://huggingface.co/WithinUsAI/DistilBERT-Gemini3.2-Pro.Fast.Coder.NSFW-0.1B" target="_blank">🤗 View on Hub →</a>
</div>
</div>
</body>
</html>