Yet another Stable Diffusion image generator (Single executable UI)

[Bastardly]

Thought that it was time to share my little project with the community.
It is yet another open source client running on stable-diffusion.cpp - Hence the name "Yasdig" (Yet another stable diffusion image generator).

It is mainly written in Rust, is should be very beginner friendly and easy to install. (No Python, Cuda, Rocm) Single file to get it up and running.
It is built with the philosophy: Generate cheap, and then escalate the keepers.

Features:

• Image generation: SD, SDXL, Moss, Qwen, Flux, Z-image... it's in the repo's readme.
• Image refinement: Refine, edit, expand, post-processing, upscaling.
• Image to image
• Video generation: Wan, H3, LTX2.3 and 2.5
• Video upscaling with FSR or LTX2.3
• Cast for image and video generation.
• Narrated stills using chatterbox.
• Models manager with download from Huggingface.
• Agent assistant for image and video generation.
• Convert tool to easily change the weight type on heavier models.

Currently, I'm only using Vulkan as a backend. Running on Linux and Windows.

https://github.com/Bastardly/yasdig
https://yasdig.com/ has some pretty screenshots.

Feedback is welcome.

Hope you like it.