Upload 4 files

Dockerfile

@@ -0,0 +1,17 @@
+FROM python:3.10-slim
+
+# Set working directory
+WORKDIR /app
+
+# Copy requirements and install dependencies
+COPY requirements.txt .
+RUN pip install --upgrade pip && pip install -r requirements.txt
+
+# Copy application code
+COPY app.py .
+
+# Expose port for Gradio
+EXPOSE 7860
+
+# Run the app
+CMD ["python", "app.py"]

README.md

@@ -1,34 +1,32 @@
----
-title: "WorldGen Text-to-3D"
-colorFrom: "blue"
-colorTo: "purple"
-sdk: "gradio"
-sdk_version: "5.44.0"
-python_version: 3.11
-app_file: "app.py"
-pinned: false
----
-
-# WorldGen — Text -> 3D Scene (Hugging Face Spaces)
-
-This Space provides a minimal Text -> 3D demo using the open-source WorldGen project.
-
-## Quick start
-1) Open the Space and enter an English scene description.
-2) Click "Generate". A ZIP file with 3D outputs will appear for download.
-
-> Tip: On free CPU it will be very slow and may timeout. Switch the Space hardware to a GPU (T4/A10G) for practical generation.
-
-## Environment
-- Python 3.11 (set via the YAML block above)
-- Gradio 5.x (set via sdk_version)
-
-## Dependencies
-We keep requirements minimal so pip can resolve versions:
-- torch>=2.7 (required by WorldGen)
-- websockets>=13.1 (required by viser used inside WorldGen)
-- WorldGen from GitHub
-
-## Notes
-- This app disables CUDA and evaluation at runtime to avoid GPU-only/compiled ops when running on CPU.
-- Temporary outputs are cleaned a few minutes after each run to avoid hitting the 50GB ephemeral disk limit.
+# WorldGen • Text-to-3D Scene Generator
+
+This Hugging Face Space allows you to generate 3D scenes from English text prompts using the WorldGen framework.
+
+## 🚀 How to Use
+
+1. Enter a scene description in English (e.g., `a neon-lit cyberpunk street with rain reflections, highly detailed`)
+2. Click **Generate 3D Scene**
+3. Download the ZIP file containing the generated 3D assets
+
+## 🔐 Hugging Face Token
+
+This app downloads model files from Hugging Face Hub. If you encounter `401 Unauthorized` errors, you may need to provide a Hugging Face token.
+
+### Setting Token in Space
+
+1. Go to your Space → **Settings** → **Secrets**
+2. Add a new secret:
+   - Name: `HF_TOKEN`
+   - Value: your Hugging Face access token
+
+The app will automatically read this token via `os.getenv("HF_TOKEN")`.
+
+## ⚙️ Runtime Notes
+
+- CPU-only mode is enabled by default to avoid CUDA errors
+- Temporary files are cleaned automatically after each run
+- Model files are downloaded from `mit-han-lab/svdq-int4-flux.1-dev`
+
+## 📦 Dependencies
+
+See `requirements.txt` for full dependency list.