Stable Diffusion CoreML Model Multi-resolution Generation Tool. No conversion, No more disk space.
This project breaks the limitation that Stable Diffusion CoreML Model only supports a single resolution.
No need for complicated model conversion, no need for duplicate and huge models to waste disk space,
Compatible with the official apple/ml-stable-diffusion project and apps using this project, such as Mochi Diffusion.
You just need to have a model that has been converted to coreml format, 512x512 or 768x768 resolution,
Drag and drop the model to the interface and press Enter...
Instantly you will have additional 512x768 and 768x512 resolution models, which only take up a few hundred KB of space.
-Only supports SD 1.5, SDXL can be added by yourself
-Only compatible with models converted using coremltools 7.0
t.mp4
此项目打破了Stable Diffusion CoreML Model只支持单一分辨率的限制，
无需复杂的模型转换，无需重复且巨大的模型浪费磁盘空间，
兼容官方apple/ml-stable-diffusion项目，以及使用此项目的app,例如：Mochi Diffusion.
您只需拥有一个已经转换成coreml格式的模型，512x512或768x768分辨率，
将模型拖放至界面，按下回车...
一瞬间你将立马拥有另外的512x768和768x512分辨率的模型，这仅仅占用几百KB的空间。
-仅支持SD 1.5，SDXL可自行添加
-仅兼容使用coremltools 7.0转换的模型