Spaces:
Running
Running
File size: 39,262 Bytes
543c45d | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 | <!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8"/>
<meta name="viewport" content="width=device-width, initial-scale=1.0"/>
<title>Matrix.Corp — AI Model Registry</title>
<script src="https://cdnjs.cloudflare.com/ajax/libs/react/18.2.0/umd/react.production.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/react-dom/18.2.0/umd/react-dom.production.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/babel-standalone/7.23.2/babel.min.js"></script>
<style>
* { box-sizing: border-box; margin: 0; padding: 0; }
body { background: #080c10; font-family: 'IBM Plex Mono', 'Courier New', monospace; color: #e2e8f0; }
::-webkit-scrollbar { width: 6px; height: 6px; }
::-webkit-scrollbar-track { background: #0d1117; }
::-webkit-scrollbar-thumb { background: #1e2936; border-radius: 3px; }
::-webkit-scrollbar-thumb:hover { background: #2d3f52; }
button { font-family: 'IBM Plex Mono', 'Courier New', monospace; }
pre { font-family: 'IBM Plex Mono', 'Courier New', monospace; }
</style>
</head>
<body>
<div id="root"></div>
<script type="text/babel">
const { useState } = React;
const C = {
zenith: "#6ee7f7",
vortex: "#a78bfa",
tg: "#4ade80",
bg: "#080c10",
card: "#0d1117",
border: "#1e2936",
text: "#e2e8f0",
muted: "#64748b",
};
const MODELS = {
zenith: {
color: C.zenith, glow: "rgba(110,231,247,0.12)", icon: "🌌",
name: "Zenith", tagline: "Reasoning & Emotional Intelligence",
hardware: "Tenstorrent p300a",
models: [
{ name: "Zenith-7B", params: "7B", status: "preview", use: "Code generation & fast inference", vram: "14GB", base: "Qwen2.5-Coder-7B" },
{ name: "Zenith-28B", params: "28B", status: "preview", use: "Nuanced reasoning + EQ awareness", vram: "56GB", base: "Qwen3.5-27B (Claude Opus 4.6 distill)" },
{ name: "Zenith-32B", params: "32B", status: "preview", use: "Mathematical & structured reasoning", vram: "64GB", base: "DeepSeek-R1-Distill-Qwen-32B" },
{ name: "Zenith-70B", params: "70B", status: "preview", use: "Max capability — multi-card", vram: "140GB / INT4: 35GB", base: "DeepSeek-R1-Distill-Llama-70B" },
],
features: ["Ring Attention 32K", "MoE 12 experts top-2", "EQ Engine (V1)", "TP=8 / PP=4", "NoC optimized", "Ollama + vLLM"],
benchmarks: [
{ name: "GSM8K", scores: { "28B": 75, "32B": 82, "70B": 89 } },
{ name: "HumanEval", scores: { "28B": 45, "32B": 52, "70B": 68 } },
{ name: "MATH", scores: { "28B": 35, "32B": 42, "70B": 55 } },
{ name: "EQ Bench", scores: { "28B": 78, "32B": 80, "70B": 84 } },
{ name: "TruthfulQA",scores: { "28B": 65, "32B": 70, "70B": 76 } },
],
arch: [
{ label: "Input Embedding + Position", color: "#6ee7f7" },
{ label: "Ring Attention (32K, 8K chunks)", color: "#38bdf8" },
{ label: "MoE Layer — 12 experts, top-2 routing", color: "#6ee7f7" },
{ label: "EQ Attention Bias (per-head)", color: "#67e8f9" },
{ label: "EQ-Gated FFN (sigmoid gate)", color: "#38bdf8" },
{ label: "EQ State Recurrence (GRU)", color: "#6ee7f7" },
{ label: "× 36 layers (28B) / 40 (32B) / 64 (70B)", color: "#475569" },
{ label: "LM Head", color: "#6ee7f7" },
],
trainSteps: [
{ title: "1. Install Dependencies", cmd: "pip install torch transformers accelerate peft bitsandbytes\npip install flash-attn # CUDA only" },
{ title: "2. Prepare Training Data", cmd: "python data/openthoughts_processor.py \\\n --dataset open-thoughts/OpenThoughts3-1.2M \\\n --output_dir ./data/processed \\\n --use_curriculum \\\n --use_quality_filter" },
{ title: "3. Run LoRA Fine-tuning", cmd: "python train.py \\\n --base_model Qwen/Qwen3.5-7B \\\n --train_data ./data/processed \\\n --use_lora --lora_r 16 --lora_alpha 32 \\\n --epochs 3 --batch_size 4 \\\n --use_ring_attention \\\n --max_seq_length 32768 \\\n --mixed_precision bf16 \\\n --use_eq_adapter \\\n --use_tenstorrent" },
{ title: "4. Evaluate", cmd: "python -m evaluation.benchmark \\\n --model_path ./outputs/checkpoint-final \\\n --benchmarks gsm8k humaneval math truthfulqa \\\n --output_dir ./eval_results" },
{ title: "5. Push to HuggingFace", cmd: "from huggingface_hub import HfApi\napi = HfApi()\napi.upload_folder(\n folder_path='./outputs/checkpoint-final',\n repo_id='Matrix-Corp/Zenith-28b-p300-V1',\n repo_type='model'\n)" },
],
ollama: "ollama create zenith-28b -f Modelfile\nollama run zenith-28b \"Solve this step by step: ...\"",
vllm: "python -m vllm.entrypoints.openai.api_server \\\n --model ./outputs/checkpoint-final \\\n --tensor-parallel-size 2 \\\n --max-model-len 32768 \\\n --port 8000",
hf: "from transformers import AutoModelForCausalLM, AutoTokenizer\n\nmodel = AutoModelForCausalLM.from_pretrained(\n \"Matrix-Corp/Zenith-28b-p300-V1\",\n torch_dtype=\"auto\",\n device_map=\"auto\"\n)\ntokenizer = AutoTokenizer.from_pretrained(\n \"Matrix-Corp/Zenith-28b-p300-V1\"\n)\ninputs = tokenizer(\"Your prompt\", return_tensors=\"pt\")\noutputs = model.generate(**inputs, max_new_tokens=512)\nprint(tokenizer.decode(outputs[0]))",
},
vortex: {
color: C.vortex, glow: "rgba(167,139,250,0.12)", icon: "🔬",
name: "Vortex Scientific", tagline: "Deep Science Reasoning",
hardware: "MacBook M2/M3 + Nvidia 4060",
models: [
{ name: "Vortex-7B", params: "7B", status: "preview", use: "Science reasoning on consumer hardware", vram: "8GB", base: "From scratch — custom architecture" },
{ name: "Vortex-13B", params: "13B", status: "preview", use: "Advanced multi-domain science", vram: "16GB / INT8", base: "From scratch — custom architecture" },
],
features: ["Hybrid SSM + Attention", "60% SSM / 40% Attn (7B)", "4 Science Modules", "Custom 50K Sci Tokenizer", "LaTeX & Equations", "Molecular Module"],
benchmarks: [
{ name: "Physics", scores: { "7B": 62, "13B": 74 } },
{ name: "Math", scores: { "7B": 58, "13B": 71 } },
{ name: "Chemistry", scores: { "7B": 65, "13B": 78 } },
{ name: "Biology", scores: { "7B": 60, "13B": 72 } },
{ name: "Zoology", scores: { "7B": 55, "13B": 68 } },
],
arch: [
{ label: "Science Tokenizer (50K vocab)", color: "#a78bfa" },
{ label: "SSM Block — O(n) complexity (Mamba-style)", color: "#8b5cf6" },
{ label: "Local Attention — 512 token window", color: "#a78bfa" },
{ label: "Equation Module (LaTeX parser)", color: "#c4b5fd" },
{ label: "Numerical Reasoning Module", color: "#8b5cf6" },
{ label: "Citation Awareness Module", color: "#a78bfa" },
{ label: "Molecular Module (118 elements)", color: "#c4b5fd" },
{ label: "SciGate FFN — domain routing", color: "#8b5cf6" },
{ label: "× 32 layers (7B) / 40 (13B)", color: "#475569" },
{ label: "LM Head", color: "#a78bfa" },
],
trainSteps: [
{ title: "1. Install Dependencies", cmd: "pip install torch transformers accelerate datasets\npip install rdkit pubchempy pylatexenc # science deps\npip install flash-attn # CUDA only" },
{ title: "2. Build Science Tokenizer", cmd: "python tokenizer/vortex_tokenizer.py \\\n --corpus_dir ./data/raw \\\n --vocab_size 50000 \\\n --output_dir ./tokenizer/trained" },
{ title: "3. Prepare Science Data", cmd: "python data/dataset_loader.py \\\n --datasets pile_scientific s2orc automath pubmed_qa \\\n --output_dir ./data/processed\n\n# Optional: scrape additional sources\npython data/scraper.py \\\n --sources arxiv pubmed wikipedia nist nasa \\\n --max_docs 100000" },
{ title: "4. Train from Scratch", cmd: "python train.py \\\n --model_size 7b \\\n --device cuda \\\n --data_dir ./data/processed \\\n --output_dir ./checkpoints \\\n --max_steps 100000 \\\n --use_curriculum" },
{ title: "5. Push to HuggingFace", cmd: "from huggingface_hub import HfApi\napi = HfApi()\napi.upload_folder(\n folder_path='./checkpoints/latest',\n repo_id='Matrix-Corp/Vortex-7b-V1',\n repo_type='model'\n)" },
],
ollama: "ollama create vortex-7b -f ollama_7b_modelfile\nollama run vortex-7b \"Explain quantum entanglement\"",
vllm: "python inference/inference.py \\\n --model_path ./checkpoints/latest \\\n --model_size 7b \\\n --device cuda \\\n --interactive",
hf: "from transformers import AutoModelForCausalLM, AutoTokenizer\n\nmodel = AutoModelForCausalLM.from_pretrained(\n \"Matrix-Corp/Vortex-7b-V1\",\n torch_dtype=\"auto\",\n device_map=\"auto\"\n)\ntokenizer = AutoTokenizer.from_pretrained(\n \"Matrix-Corp/Vortex-7b-V1\"\n)\ninputs = tokenizer(\"Explain photosynthesis\", return_tensors=\"pt\")\noutputs = model.generate(**inputs, max_new_tokens=512)\nprint(tokenizer.decode(outputs[0]))",
},
touchgrass: {
color: C.tg, glow: "rgba(74,222,128,0.12)", icon: "🌿",
name: "Touch Grass", tagline: "Music AI Assistant",
hardware: "Any hardware — ultra lightweight",
models: [
{ name: "TouchGrass-3B", params: "3B", status: "planned", use: "Ultra lightweight music assistant", vram: "6GB", base: "Qwen3.5-3B-Instruct" },
{ name: "TouchGrass-7B", params: "7B", status: "planned", use: "Full-featured music assistant", vram: "14GB", base: "Qwen3.5-7B-Instruct" },
],
features: ["Tab & Chord Generation", "Music Theory Engine", "Ear Training Module", "Music EQ Adapter", "Songwriting Assistant", "All Instruments"],
benchmarks: [
{ name: "Music Theory", scores: { "3B": 70, "7B": 82 } },
{ name: "Tab Gen", scores: { "3B": 65, "7B": 78 } },
{ name: "Ear Training", scores: { "3B": 68, "7B": 80 } },
{ name: "Songwriting", scores: { "3B": 72, "7B": 85 } },
{ name: "EQ Detect", scores: { "3B": 74, "7B": 83 } },
],
arch: [
{ label: "Qwen3.5 Base + Music Token Extensions", color: "#4ade80" },
{ label: "LoRA Fine-tune (r=16, music data)", color: "#22c55e" },
{ label: "Tab & Chord Module (validated output)", color: "#4ade80" },
{ label: "Music Theory Engine (circle of fifths etc)", color: "#86efac" },
{ label: "Ear Training Module (interval/quality)", color: "#22c55e" },
{ label: "Music EQ Adapter (4 emotions)", color: "#4ade80" },
{ label: "Songwriting Module (structure + hooks)", color: "#86efac" },
{ label: "Instrument Context Switch", color: "#22c55e" },
{ label: "LM Head", color: "#4ade80" },
],
trainSteps: [
{ title: "1. Install Dependencies", cmd: "pip install torch transformers accelerate peft bitsandbytes trl\npip install datasets music21 # music processing" },
{ title: "2. Extend Tokenizer", cmd: "python tokenizer/music_token_extension.py \\\n --base_model Qwen/Qwen3.5-7B-Instruct \\\n --output_dir ./tokenizer/touchgrass" },
{ title: "3. Generate Music QA Dataset", cmd: "python data/music_qa_generator.py \\\n --categories all \\\n --instruments all \\\n --skill_levels beginner intermediate advanced \\\n --output ./data/music_qa.json" },
{ title: "4. LoRA Fine-tune", cmd: "python training/finetune.py \\\n --base_model Qwen/Qwen3.5-7B-Instruct \\\n --train_data ./data/music_qa.json \\\n --use_lora --lora_r 16 --lora_alpha 32 \\\n --epochs 3 --batch_size 8 \\\n --learning_rate 1e-4 \\\n --mixed_precision bf16" },
{ title: "5. Push to HuggingFace", cmd: "from huggingface_hub import HfApi\napi = HfApi()\napi.upload_folder(\n folder_path='./outputs/checkpoint-final',\n repo_id='Matrix-Corp/TouchGrass-7B',\n repo_type='model'\n)" },
],
ollama: "ollama create touchgrass-7b -f ollama_7b_modelfile\nollama run touchgrass-7b \"Teach me how to play a G chord\"",
vllm: "python inference/inference.py \\\n --model Matrix-Corp/TouchGrass-7B \\\n --instrument guitar \\\n --skill_level beginner \\\n --interactive",
hf: "from transformers import AutoModelForCausalLM, AutoTokenizer\n\nmodel = AutoModelForCausalLM.from_pretrained(\n \"Matrix-Corp/TouchGrass-7B\",\n torch_dtype=\"auto\",\n device_map=\"auto\"\n)\ntokenizer = AutoTokenizer.from_pretrained(\n \"Matrix-Corp/TouchGrass-7B\"\n)\ninputs = tokenizer(\n \"[GUITAR][BEGINNER] How do I play a G chord?\",\n return_tensors=\"pt\"\n)\noutputs = model.generate(**inputs, max_new_tokens=512)\nprint(tokenizer.decode(outputs[0]))",
},
};
const HARDWARE = [
{ name: "MacBook Pro M2 16GB", icon: "💻", vram: 16, type: "MPS", canRun: ["Zenith-7B (INT4)", "Vortex-7B", "TouchGrass-3B", "TouchGrass-7B"] },
{ name: "MacBook Pro M3 Max 48GB",icon: "💻", vram: 48, type: "MPS", canRun: ["Zenith-7B", "Zenith-28B (INT4)", "Vortex-7B", "Vortex-13B", "TouchGrass-3B", "TouchGrass-7B"] },
{ name: "Nvidia 4060 Laptop 8GB", icon: "🖥️", vram: 8, type: "CUDA", canRun: ["Zenith-7B (INT4)", "Vortex-7B", "TouchGrass-3B"] },
{ name: "Nvidia 4060 Ti 16GB", icon: "🖥️", vram: 16, type: "CUDA", canRun: ["Zenith-7B", "Vortex-7B", "Vortex-13B (INT8)", "TouchGrass-3B", "TouchGrass-7B"] },
{ name: "Tenstorrent p300a", icon: "⚡", vram: 64, type: "TENSTORRENT", canRun: ["Zenith-7B", "Zenith-28B", "Zenith-32B", "Zenith-70B (INT4)"] },
{ name: "2× Tenstorrent p300a", icon: "⚡", vram: 128, type: "TENSTORRENT", canRun: ["Zenith-7B", "Zenith-28B", "Zenith-32B", "Zenith-70B (INT8)"] },
{ name: "A100 80GB", icon: "🔥", vram: 80, type: "CUDA", canRun: ["Zenith-7B", "Zenith-28B", "Zenith-32B", "Zenith-70B (INT4)", "Vortex-7B", "Vortex-13B", "TouchGrass-3B", "TouchGrass-7B"] },
];
const STATUS_STYLE = {
preview: { label: "PREVIEW", color: "#f59e0b", bg: "rgba(245,158,11,0.1)" },
planned: { label: "PLANNED", color: "#64748b", bg: "rgba(100,116,139,0.1)" },
released: { label: "RELEASED", color: "#4ade80", bg: "rgba(74,222,128,0.1)" },
};
const NAV = [
{ id: "explorer", label: "Models", icon: "◈" },
{ id: "architecture", label: "Architecture", icon: "⬡" },
{ id: "benchmarks", label: "Benchmarks", icon: "▲" },
{ id: "training", label: "Training", icon: "◎" },
{ id: "deploy", label: "Deploy", icon: "▶" },
{ id: "hardware", label: "Hardware", icon: "⬢" },
];
function CodeBlock({ code, color }) {
const [copied, setCopied] = useState(false);
const copy = () => {
navigator.clipboard.writeText(code).then(() => { setCopied(true); setTimeout(() => setCopied(false), 1800); });
};
return (
<div style={{ position: "relative" }}>
<pre style={{ background: "#060a0e", border: `1px solid ${C.border}`, borderRadius: 8, padding: "18px 20px", fontSize: 11, color: "#7dd3fc", overflowX: "auto", lineHeight: 1.9, margin: 0, whiteSpace: "pre-wrap", wordBreak: "break-all" }}>
{code}
</pre>
<button onClick={copy} style={{ position: "absolute", top: 10, right: 10, background: copied ? `${color}22` : "transparent", border: `1px solid ${copied ? color : C.border}`, borderRadius: 5, padding: "4px 10px", color: copied ? color : C.muted, cursor: "pointer", fontSize: 9, letterSpacing: 1 }}>
{copied ? "✓ COPIED" : "COPY"}
</button>
</div>
);
}
function App() {
const [family, setFamily] = useState("zenith");
const [section, setSection] = useState("explorer");
const [hw, setHw] = useState(null);
const [deployTab, setDeployTab] = useState("ollama");
const [trainStep, setTrainStep] = useState(0);
const F = MODELS[family];
const col = F.color;
const glow = F.glow;
const handleFamilyChange = (key) => { setFamily(key); setTrainStep(0); setHw(null); };
return (
<div style={{ background: C.bg, minHeight: "100vh", display: "flex", flexDirection: "column", color: C.text }}>
{/* ── Header ── */}
<div style={{ borderBottom: `1px solid ${C.border}`, padding: "14px 24px", display: "flex", alignItems: "center", justifyContent: "space-between", background: "rgba(13,17,23,0.98)", position: "sticky", top: 0, zIndex: 99 }}>
<div style={{ display: "flex", alignItems: "center", gap: 12 }}>
<div style={{ width: 34, height: 34, background: `linear-gradient(135deg, ${C.zenith}, ${C.vortex})`, borderRadius: 7, display: "flex", alignItems: "center", justifyContent: "center", fontSize: 17, fontWeight: 800, color: "#000" }}>M</div>
<div>
<div style={{ fontSize: 13, fontWeight: 700, letterSpacing: 3, color: "#fff" }}>MATRIX.CORP</div>
<div style={{ fontSize: 9, color: C.muted, letterSpacing: 2 }}>AI MODEL REGISTRY</div>
</div>
</div>
<div style={{ display: "flex", gap: 8 }}>
{Object.entries(MODELS).map(([key, m]) => (
<button key={key} onClick={() => handleFamilyChange(key)} style={{ background: family === key ? m.glow : "transparent", border: `1px solid ${family === key ? m.color : C.border}`, borderRadius: 6, padding: "6px 14px", color: family === key ? m.color : C.muted, cursor: "pointer", fontSize: 11, letterSpacing: 1, transition: "all 0.15s" }}>
{m.icon} {key === "touchgrass" ? "TOUCH GRASS" : m.name.toUpperCase().split(" ")[0]}
</button>
))}
</div>
</div>
<div style={{ display: "flex", flex: 1, minHeight: 0 }}>
{/* ── Sidebar ── */}
<div style={{ width: 188, borderRight: `1px solid ${C.border}`, padding: "20px 0", display: "flex", flexDirection: "column", gap: 2, flexShrink: 0, background: "rgba(13,17,23,0.6)" }}>
{NAV.map(item => (
<button key={item.id} onClick={() => setSection(item.id)} style={{ background: section === item.id ? `linear-gradient(90deg, ${glow}, transparent)` : "transparent", border: "none", borderLeft: `2px solid ${section === item.id ? col : "transparent"}`, padding: "11px 20px", color: section === item.id ? col : C.muted, cursor: "pointer", fontSize: 11, letterSpacing: 1, textAlign: "left", display: "flex", alignItems: "center", gap: 10, transition: "all 0.15s" }}>
<span style={{ fontSize: 13, opacity: section === item.id ? 1 : 0.6 }}>{item.icon}</span>
{item.label.toUpperCase()}
</button>
))}
{/* Family pill */}
<div style={{ margin: "20px 12px 0", padding: "14px 12px", background: glow, border: `1px solid ${col}22`, borderRadius: 9 }}>
<div style={{ fontSize: 22, marginBottom: 7 }}>{F.icon}</div>
<div style={{ fontSize: 10, color: col, letterSpacing: 1, marginBottom: 5, fontWeight: 700 }}>{F.name.toUpperCase()}</div>
<div style={{ fontSize: 9, color: C.muted, lineHeight: 1.6 }}>{F.hardware}</div>
<div style={{ marginTop: 10, fontSize: 9, color: C.muted }}>
{F.models.length} model{F.models.length > 1 ? "s" : ""}
</div>
</div>
</div>
{/* ── Main Content ── */}
<div style={{ flex: 1, padding: "28px 32px", overflowY: "auto" }}>
{/* ────── EXPLORER ────── */}
{section === "explorer" && (
<div>
<div style={{ marginBottom: 26 }}>
<div style={{ fontSize: 10, color: col, letterSpacing: 3, marginBottom: 7 }}>MODEL EXPLORER</div>
<div style={{ fontSize: 24, fontWeight: 800, color: "#fff", marginBottom: 5 }}>{F.icon} {F.name}</div>
<div style={{ fontSize: 12, color: C.muted }}>{F.tagline}</div>
</div>
<div style={{ display: "flex", flexWrap: "wrap", gap: 7, marginBottom: 28 }}>
{F.features.map(f => (
<span key={f} style={{ padding: "4px 13px", background: glow, border: `1px solid ${col}30`, borderRadius: 20, fontSize: 10, color: col }}>{f}</span>
))}
</div>
<div style={{ display: "grid", gridTemplateColumns: "repeat(auto-fill, minmax(270px, 1fr))", gap: 16 }}>
{F.models.map(m => (
<div key={m.name} style={{ background: C.card, border: `1px solid ${C.border}`, borderRadius: 11, padding: 22, transition: "border-color 0.15s, transform 0.15s", cursor: "default" }}
onMouseEnter={e => { e.currentTarget.style.borderColor = col + "55"; e.currentTarget.style.transform = "translateY(-2px)"; }}
onMouseLeave={e => { e.currentTarget.style.borderColor = C.border; e.currentTarget.style.transform = "none"; }}>
<div style={{ display: "flex", justifyContent: "space-between", alignItems: "flex-start", marginBottom: 14 }}>
<div>
<div style={{ fontSize: 15, fontWeight: 700, color: "#fff" }}>{m.name}</div>
<div style={{ fontSize: 11, color: col, marginTop: 3 }}>{m.params} parameters</div>
</div>
<span style={{ padding: "3px 9px", background: STATUS_STYLE[m.status].bg, color: STATUS_STYLE[m.status].color, borderRadius: 5, fontSize: 9, letterSpacing: 1 }}>{STATUS_STYLE[m.status].label}</span>
</div>
<div style={{ fontSize: 11, color: C.muted, marginBottom: 14, lineHeight: 1.7 }}>{m.use}</div>
<div style={{ borderTop: `1px solid ${C.border}`, paddingTop: 13, display: "flex", flexDirection: "column", gap: 7 }}>
<div style={{ display: "flex", justifyContent: "space-between", fontSize: 10 }}>
<span style={{ color: C.muted }}>VRAM REQUIRED</span>
<span style={{ color: "#fff", fontWeight: 700 }}>{m.vram}</span>
</div>
<div style={{ display: "flex", justifyContent: "space-between", fontSize: 10, gap: 10 }}>
<span style={{ color: C.muted, flexShrink: 0 }}>BASE MODEL</span>
<span style={{ color: "#94a3b8", textAlign: "right" }}>{m.base}</span>
</div>
</div>
</div>
))}
</div>
</div>
)}
{/* ────── ARCHITECTURE ────── */}
{section === "architecture" && (
<div>
<div style={{ marginBottom: 26 }}>
<div style={{ fontSize: 10, color: col, letterSpacing: 3, marginBottom: 7 }}>ARCHITECTURE VISUALIZER</div>
<div style={{ fontSize: 24, fontWeight: 800, color: "#fff" }}>{F.icon} {F.name} — Layer Stack</div>
</div>
<div style={{ display: "flex", gap: 28, flexWrap: "wrap" }}>
<div style={{ flex: "1 1 380px", maxWidth: 480 }}>
{F.arch.map((layer, i) => (
<div key={i} style={{ padding: "13px 20px", background: i % 2 === 0 ? `${layer.color}10` : `${layer.color}07`, border: `1px solid ${layer.color}30`, borderBottom: i < F.arch.length - 1 ? "none" : `1px solid ${layer.color}30`, borderRadius: i === 0 ? "9px 9px 0 0" : i === F.arch.length - 1 ? "0 0 9px 9px" : 0, display: "flex", justifyContent: "space-between", alignItems: "center" }}>
<span style={{ fontSize: 12, color: layer.color }}>{layer.label}</span>
<span style={{ fontSize: 9, color: C.muted, flexShrink: 0, marginLeft: 10 }}>L{i + 1}</span>
</div>
))}
</div>
{family === "zenith" && (
<div style={{ flex: "1 1 280px" }}>
<div style={{ padding: 20, background: C.card, border: `1px solid ${C.border}`, borderRadius: 11, marginBottom: 16 }}>
<div style={{ fontSize: 10, color: col, letterSpacing: 2, marginBottom: 14 }}>EQ ENGINE DEEP DIVE</div>
{[
["EQ state vector", "frustration (0–1) + 8 emotion logits = 9 values"],
["Attention bias", "per-head scalar → shifts attention distribution"],
["FFN gate", "sigmoid over hidden dim → suppresses/amplifies neurons"],
["GRU recurrence", "state passed layer-to-layer → emotional memory"],
["Consistency loss", "MSE between adjacent layer states → smooth arc"],
["Always FP16", "EQ modules never quantized — quality critical"],
].map(([k, v], i) => (
<div key={i} style={{ marginBottom: 10, paddingBottom: 10, borderBottom: i < 5 ? `1px solid ${C.border}` : "none" }}>
<div style={{ fontSize: 10, color: col, marginBottom: 3 }}>{k}</div>
<div style={{ fontSize: 10, color: C.muted, lineHeight: 1.5 }}>{v}</div>
</div>
))}
</div>
</div>
)}
{family === "vortex" && (
<div style={{ flex: "1 1 280px" }}>
<div style={{ padding: 20, background: C.card, border: `1px solid ${C.border}`, borderRadius: 11 }}>
<div style={{ fontSize: 10, color: col, letterSpacing: 2, marginBottom: 14 }}>SCIENCE MODULES</div>
{[
["Equation Module", "LaTeX span detection + structural tree encoding"],
["Numerical Module", "Digit-level encoding, sci notation, unit awareness"],
["Citation Module", "Citation span detection + provenance tracking"],
["Molecular Module", "118 element embeddings + SMILES + amino acids"],
["SciGate FFN", "7-domain routing — different FFN path per science area"],
].map(([k, v], i) => (
<div key={i} style={{ marginBottom: 10, paddingBottom: 10, borderBottom: i < 4 ? `1px solid ${C.border}` : "none" }}>
<div style={{ fontSize: 10, color: col, marginBottom: 3 }}>{k}</div>
<div style={{ fontSize: 10, color: C.muted, lineHeight: 1.5 }}>{v}</div>
</div>
))}
</div>
</div>
)}
{family === "touchgrass" && (
<div style={{ flex: "1 1 280px" }}>
<div style={{ padding: 20, background: C.card, border: `1px solid ${C.border}`, borderRadius: 11 }}>
<div style={{ fontSize: 10, color: col, letterSpacing: 2, marginBottom: 14 }}>MUSIC MODULES</div>
{[
["Tab & Chord Module", "Validated ASCII tab + chord diagram generation"],
["Theory Engine", "Circle of fifths, modes, progressions, voice leading"],
["Ear Training", "Interval/quality description, solfege exercises"],
["Music EQ Adapter", "4 emotions: frustrated / confused / excited / confident"],
["Songwriting Module", "Structure, lyrics, hooks, genre-aware progressions"],
].map(([k, v], i) => (
<div key={i} style={{ marginBottom: 10, paddingBottom: 10, borderBottom: i < 4 ? `1px solid ${C.border}` : "none" }}>
<div style={{ fontSize: 10, color: col, marginBottom: 3 }}>{k}</div>
<div style={{ fontSize: 10, color: C.muted, lineHeight: 1.5 }}>{v}</div>
</div>
))}
</div>
</div>
)}
</div>
</div>
)}
{/* ────── BENCHMARKS ────── */}
{section === "benchmarks" && (
<div>
<div style={{ marginBottom: 26 }}>
<div style={{ fontSize: 10, color: col, letterSpacing: 3, marginBottom: 7 }}>BENCHMARK CHARTS</div>
<div style={{ fontSize: 24, fontWeight: 800, color: "#fff" }}>{F.icon} {F.name} — Estimated Performance</div>
<div style={{ marginTop: 8, padding: "8px 14px", background: "rgba(245,158,11,0.08)", border: "1px solid rgba(245,158,11,0.2)", borderRadius: 7, fontSize: 10, color: "#f59e0b", display: "inline-block" }}>
⚠ Preliminary estimates based on base model benchmarks + fine-tuning projections. Real benchmarks coming with trained weights.
</div>
</div>
<div style={{ display: "grid", gridTemplateColumns: "repeat(auto-fill, minmax(300px, 1fr))", gap: 16 }}>
{F.benchmarks.map(bench => (
<div key={bench.name} style={{ background: C.card, border: `1px solid ${C.border}`, borderRadius: 11, padding: 22 }}>
<div style={{ fontSize: 13, color: "#fff", marginBottom: 18, letterSpacing: 1, fontWeight: 700 }}>{bench.name}</div>
{Object.entries(bench.scores).map(([model, score]) => (
<div key={model} style={{ marginBottom: 14 }}>
<div style={{ display: "flex", justifyContent: "space-between", fontSize: 10, marginBottom: 7 }}>
<span style={{ color: C.muted }}>{model}</span>
<span style={{ color: col, fontWeight: 700 }}>{score}%</span>
</div>
<div style={{ height: 7, background: C.border, borderRadius: 4, overflow: "hidden" }}>
<div style={{ height: "100%", width: `${score}%`, background: `linear-gradient(90deg, ${col}, ${col}77)`, borderRadius: 4 }} />
</div>
</div>
))}
</div>
))}
</div>
</div>
)}
{/* ────── TRAINING ────── */}
{section === "training" && (
<div>
<div style={{ marginBottom: 26 }}>
<div style={{ fontSize: 10, color: col, letterSpacing: 3, marginBottom: 7 }}>INTERACTIVE TRAINING GUIDE</div>
<div style={{ fontSize: 24, fontWeight: 800, color: "#fff" }}>{F.icon} How to Train {F.name}</div>
</div>
{/* Step tabs */}
<div style={{ display: "flex", marginBottom: 24, borderRadius: 9, overflow: "hidden", border: `1px solid ${C.border}` }}>
{F.trainSteps.map((s, i) => (
<button key={i} onClick={() => setTrainStep(i)} style={{ flex: 1, padding: "11px 6px", background: trainStep === i ? glow : "transparent", border: "none", borderRight: i < F.trainSteps.length - 1 ? `1px solid ${C.border}` : "none", color: trainStep === i ? col : C.muted, cursor: "pointer", fontSize: 9, letterSpacing: 0.5, transition: "all 0.15s", lineHeight: 1.6 }}>
<div style={{ fontSize: 16, marginBottom: 3 }}>{"①②③④⑤"[i]}</div>
STEP {i + 1}
</button>
))}
</div>
<div style={{ background: C.card, border: `1px solid ${col}33`, borderRadius: 11, padding: 26 }}>
<div style={{ fontSize: 13, color: col, marginBottom: 18, letterSpacing: 1, fontWeight: 700 }}>{F.trainSteps[trainStep].title}</div>
<CodeBlock code={F.trainSteps[trainStep].cmd} color={col} />
<div style={{ display: "flex", justifyContent: "space-between", marginTop: 22 }}>
<button onClick={() => setTrainStep(Math.max(0, trainStep - 1))} disabled={trainStep === 0} style={{ background: "transparent", border: `1px solid ${trainStep === 0 ? C.border + "55" : C.border}`, borderRadius: 7, padding: "9px 22px", color: trainStep === 0 ? C.muted + "55" : C.muted, cursor: trainStep === 0 ? "default" : "pointer", fontSize: 11 }}>← Previous</button>
<span style={{ fontSize: 10, color: C.muted, alignSelf: "center" }}>{trainStep + 1} / {F.trainSteps.length}</span>
<button onClick={() => setTrainStep(Math.min(F.trainSteps.length - 1, trainStep + 1))} disabled={trainStep === F.trainSteps.length - 1} style={{ background: trainStep === F.trainSteps.length - 1 ? "transparent" : glow, border: `1px solid ${trainStep === F.trainSteps.length - 1 ? C.border + "55" : col}`, borderRadius: 7, padding: "9px 22px", color: trainStep === F.trainSteps.length - 1 ? C.muted + "55" : col, cursor: trainStep === F.trainSteps.length - 1 ? "default" : "pointer", fontSize: 11 }}>Next Step →</button>
</div>
</div>
</div>
)}
{/* ────── DEPLOY ────── */}
{section === "deploy" && (
<div>
<div style={{ marginBottom: 26 }}>
<div style={{ fontSize: 10, color: col, letterSpacing: 3, marginBottom: 7 }}>DEPLOYMENT WALKTHROUGH</div>
<div style={{ fontSize: 24, fontWeight: 800, color: "#fff" }}>{F.icon} Deploy {F.name}</div>
</div>
<div style={{ display: "flex", marginBottom: 22, borderRadius: 9, overflow: "hidden", border: `1px solid ${C.border}`, width: "fit-content" }}>
{[["ollama","🦙 Ollama"],["vllm","⚡ vLLM"],["hf","🤗 Transformers"]].map(([id, label]) => (
<button key={id} onClick={() => setDeployTab(id)} style={{ padding: "10px 22px", background: deployTab === id ? glow : "transparent", border: "none", borderRight: id !== "hf" ? `1px solid ${C.border}` : "none", color: deployTab === id ? col : C.muted, cursor: "pointer", fontSize: 11, letterSpacing: 0.5, transition: "all 0.15s" }}>
{label}
</button>
))}
</div>
<div style={{ background: C.card, border: `1px solid ${C.border}`, borderRadius: 11, padding: 26 }}>
<div style={{ fontSize: 11, color: col, marginBottom: 18, letterSpacing: 1 }}>
{deployTab === "ollama" ? "LOCAL DEPLOYMENT — Easiest, works offline" : deployTab === "vllm" ? "HIGH-THROUGHPUT API SERVER" : "HUGGING FACE TRANSFORMERS INTEGRATION"}
</div>
<CodeBlock code={deployTab === "ollama" ? F.ollama : deployTab === "vllm" ? F.vllm : F.hf} color={col} />
<div style={{ marginTop: 18, padding: 14, background: `${col}0e`, border: `1px solid ${col}20`, borderRadius: 8, fontSize: 11, color: C.muted, lineHeight: 1.7 }}>
{deployTab === "ollama" && "💡 Best for local use, privacy, and testing. Install Ollama from ollama.ai — works on Mac, Linux, and Windows. No GPU required for 3B/7B models."}
{deployTab === "vllm" && "💡 Best for production APIs and high concurrency. Requires CUDA GPU. Supports OpenAI-compatible API format — drop-in replacement for existing pipelines."}
{deployTab === "hf" && "💡 Best for research and custom pipelines. Works with any PyTorch-compatible hardware including CPU, CUDA, and MPS (Apple Silicon)."}
</div>
</div>
</div>
)}
{/* ────── HARDWARE ────── */}
{section === "hardware" && (
<div>
<div style={{ marginBottom: 26 }}>
<div style={{ fontSize: 10, color: col, letterSpacing: 3, marginBottom: 7 }}>HARDWARE SELECTOR</div>
<div style={{ fontSize: 24, fontWeight: 800, color: "#fff" }}>What Can I Run?</div>
<div style={{ fontSize: 12, color: C.muted, marginTop: 5 }}>Click your hardware to see compatible models across all families</div>
</div>
<div style={{ display: "grid", gridTemplateColumns: "repeat(auto-fill, minmax(210px, 1fr))", gap: 12, marginBottom: 28 }}>
{HARDWARE.map((h, i) => (
<button key={i} onClick={() => setHw(hw === i ? null : i)} style={{ background: hw === i ? glow : C.card, border: `1px solid ${hw === i ? col : C.border}`, borderRadius: 11, padding: "16px 18px", textAlign: "left", cursor: "pointer", transition: "all 0.15s", transform: hw === i ? "translateY(-2px)" : "none" }}>
<div style={{ fontSize: 22, marginBottom: 9 }}>{h.icon}</div>
<div style={{ fontSize: 12, color: "#fff", marginBottom: 5, fontWeight: 600 }}>{h.name}</div>
<div style={{ display: "flex", gap: 8 }}>
<span style={{ fontSize: 9, color: C.muted, background: C.border, padding: "2px 7px", borderRadius: 4 }}>{h.vram}GB</span>
<span style={{ fontSize: 9, color: C.muted, background: C.border, padding: "2px 7px", borderRadius: 4 }}>{h.type}</span>
</div>
</button>
))}
</div>
{hw !== null && (
<div style={{ background: C.card, border: `1px solid ${col}33`, borderRadius: 11, padding: 26 }}>
<div style={{ fontSize: 12, color: col, marginBottom: 18, letterSpacing: 1, fontWeight: 700 }}>
{HARDWARE[hw].icon} {HARDWARE[hw].name} — {HARDWARE[hw].canRun.length} Compatible Models
</div>
<div style={{ display: "flex", flexWrap: "wrap", gap: 9 }}>
{HARDWARE[hw].canRun.map(m => {
const isZenith = m.startsWith("Zenith");
const isVortex = m.startsWith("Vortex");
const isTG = m.startsWith("Touch");
const mc = isZenith ? C.zenith : isVortex ? C.vortex : C.tg;
return (
<span key={m} style={{ padding: "7px 16px", background: `${mc}11`, border: `1px solid ${mc}33`, borderRadius: 7, fontSize: 11, color: mc, fontWeight: 600 }}>{m}</span>
);
})}
</div>
<div style={{ marginTop: 18, padding: 14, background: "rgba(255,255,255,0.03)", borderRadius: 8, fontSize: 10, color: C.muted, lineHeight: 1.7 }}>
💡 Models marked with (INT4) or (INT8) require quantization. Use bitsandbytes for CUDA or llama.cpp for CPU/MPS inference.
</div>
</div>
)}
</div>
)}
</div>
</div>
{/* ── Footer ── */}
<div style={{ borderTop: `1px solid ${C.border}`, padding: "11px 24px", display: "flex", justifyContent: "space-between", alignItems: "center", fontSize: 9, color: C.muted, letterSpacing: 1, background: "rgba(13,17,23,0.8)" }}>
<span>MATRIX.CORP © 2025 — 3 FAMILIES · 8 MODELS · BUILT BY ZANDY-WANDY</span>
<div style={{ display: "flex", gap: 18 }}>
<span style={{ color: C.zenith }}>● ZENITH</span>
<span style={{ color: C.vortex }}>● VORTEX</span>
<span style={{ color: C.tg }}>● TOUCH GRASS</span>
</div>
</div>
</div>
);
}
ReactDOM.createRoot(document.getElementById("root")).render(<App />);
</script>
</body>
</html> |