Spaces:
Running
Running
|
Download README.md from Jomify/MangaTrans-Diffusion: direct link, hf CLI and curl.
- Browser
- Download file 1.22 kB
-
https://huggingface.co/spaces/Jomify/MangaTrans-Diffusion/resolve/main/README.md
- Command line
-
hf download hf://spaces/Jomify/MangaTrans-Diffusion/README.md
-
curl -L -o README.md https://huggingface.co/spaces/Jomify/MangaTrans-Diffusion/resolve/main/README.md
1.22 kB
| title: MangaTrans Diffusion | |
| emoji: 🖋️ | |
| colorFrom: red | |
| colorTo: gray | |
| sdk: static | |
| app_file: index.html | |
| fullWidth: true | |
| license: mit | |
| short_description: A diffusion transformer for Chinese manga lettering | |
| tags: | |
| - diffusion | |
| - manga | |
| - image-editing | |
| - research | |
| # MangaTrans-Diffusion · Project showcase | |
| A bilingual visual explanation of a 119.3M-parameter diffusion transformer, | |
| with the repository's all-ages before/after sample and implementation diagram. | |
| [Source code](https://github.com/jomify/MangaTrans-Diffusion) · | |
| [Model weights](https://huggingface.co/Jomify/MangaTrans-Diffusion-v3) | |
| The page shows the Japanese `おはよう` source beside the v3 `早上好` render, | |
| links to the full-resolution files and `example/regions.json`, and then explains | |
| the inference architecture. The paired images are a repository sample and the | |
| Chinese string is supplied by the caller. Run the Python CLI on a CUDA machine | |
| for additional inference. No GPU service or paid hosting is required for this page. | |
| The source code and released DiT weights use MIT. Third-party VAE and font terms | |
| remain separate. See the model card for provenance, evaluation limitations, | |
| and details of the FLUX.1-dev VAE dependency. | |