Spaces:
Running
Running
|
Download README.md from hugging-apps/pixelgen: direct link, hf CLI and curl.
- Browser
- Download file 842 Bytes
-
https://huggingface.co/spaces/hugging-apps/pixelgen/resolve/main/README.md
- Command line
-
hf download hf://spaces/hugging-apps/pixelgen/README.md
-
curl -L -o README.md https://huggingface.co/spaces/hugging-apps/pixelgen/resolve/main/README.md
842 Bytes
| title: PixelGen | |
| emoji: 🕹️ | |
| colorFrom: green | |
| colorTo: yellow | |
| sdk: gradio | |
| sdk_version: 6.31.0 | |
| app_file: app.py | |
| pinned: false | |
| license: mit | |
| short_description: 1,257 hand-written params, 0 training steps | |
| models: | |
| - DJLougen/pixelgen | |
| python_version: "3.10" | |
| # PixelGen | |
| Text-to-pixel-art with **1,257 hand-written parameters and zero training steps**. Prompts like `"red potion"` or `"blue invader"` become 16×16 PICO-8 sprites, upscaled to crisp pixel art. Every weight was typed in by hand — a real `nn.Module` whose "training program" is its source code. | |
| Model card, weights and experiments: [DJLougen/pixelgen](https://huggingface.co/DJLougen/pixelgen). | |
| The `pixelgen/` package, `weights/pixelgen.pt` checkpoint and `LICENSE` are vendored verbatim from the [MIT-licensed model repo](https://huggingface.co/DJLougen/pixelgen). | |