Reserve private GPU comparison deployment; explicitly blocked pending compatible backend

Dockerfile

@@ -0,0 +1,5 @@
+FROM python:3.12-slim
+WORKDIR /app
+COPY index.html .
+EXPOSE 7860
+CMD ["python", "-m", "http.server", "7860"]

README.md

@@ -1,10 +1,11 @@
 ---
-title: Ald Tao Gpu
-emoji: 🏢
-colorFrom: indigo
-colorTo: blue
+title: TAO GPU Training Lab · backend pending
+emoji: 🧪
+colorFrom: blue
+colorTo: purple
 sdk: docker
+app_port: 7860
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+Private GPU comparison deployment reserved. Not a working GPU trainer yet. Native comparison backend currently uses AMD HIP/ROCm; Hugging Face offers NVIDIA GPU hardware. A verified CUDA/NVIDIA backend or independent compatible implementation is required. No paid hardware has been requested. CPU runs belong in ald-tao-cpu. This app must not silently fall back to CPU and label it GPU.

index.html

@@ -0,0 +1 @@
+<!doctype html><html lang="en"><meta charset="utf-8"><title>TAO GPU Lab · backend pending</title><style>body{background:#0c1420;color:#e5edf5;font:18px system-ui;max-width:800px;margin:80px auto;padding:20px}a{color:#62dfb1}</style><h1>TAO GPU comparison lab</h1><h2>GPU backend pending</h2><p>This reserved private Space does not run GPU training yet.</p><p>The current native simulator supports AMD HIP/ROCm. Hugging Face GPU hardware is NVIDIA. A compatible backend and matched CPU/GPU validation are required before enabling training.</p><p>No paid accelerator has been allocated. CPU training uses the separate <a href="https://huggingface.co/spaces/TiGa-RCE/ald-tao-cpu">TAO CPU Space</a>.</p></html>