Hey everyone!
Some of you may have noticed that I haven't released a new LoRA in quite a while. I also haven't been spending as much time exploring and converting newer DiT-based models like Anima, Krea, Z-Image, and others.
The reason is pretty simple: the AI image generation scene has been evolving incredibly fast. There are so many new architectures, checkpoints, workflows, and tools coming out that keeping up with everything has become difficult. By the time I get comfortable with one model, there are already several new things to explore.
So, instead of constantly chasing the next model, I've decided to spend more time working on something else: extensions for the Stable Diffusion WebUI family, especially Forge, reForge, and Forge Neo.
What have I been working on?
Most of these projects started with small problems or ideas from my own image-generation workflow. Sometimes I wanted a more convenient way to manage prompts. Sometimes I wanted more control over sampling, or better tools for fixing and refining an image after generation.
A few projects I've been working on include:
Prompt Vault: A place to organize prompts, tags, and saved ideas, with additional tools to make prompt management easier.
Kitchen Samplers: Additional samplers and scheduler options for Forge-family WebUIs.
Optical Realism: Experiments with optical effects, lens characteristics, depth of field, and other photographic touches.
Auto Color Corrector: Tools to help correct colour casts, contrast, exposure, and other issues in generated images.
Digital Mastering: Post-processing tools for colour grading and putting the finishing touches on an image.
Forge Notebook: A simple way to keep notes inside the WebUI.
Lobe Theme Redux: An adapted interface theme for the Forge family.
Photopea Embed for ReForge / Neo: Bringing Photopea into the WebUI workflow.
Impact ADetailer Neo: An alternative ADetailer workflow with additional image inspection and comparison tools.
Some are forks or adaptations of existing projects, while others are experiments and tools I've put together myself. They're all things I've wanted to use in my own workflow, and I figured they might be useful to someone else too.
Why focus on extensions?
For me, generating an image is only part of the process. There's also prompt management, choosing the right sampling method, correcting colours, refining details, and deciding when an image is actually finished.
A good checkpoint can make a huge difference, of course. But I also think the tools we use around it deserve some attention. Sometimes a small quality-of-life improvement can make the whole process more enjoyable, and a little extra care in post-processing can help an image get closer to what I had in mind.
I'm not abandoning LoRA creation or new models forever. I simply don't want to turn this hobby into a race to keep up with every new release. For now, I'd rather explore these smaller projects at my own pace and see where they lead.
Give them a try!
If you use Forge, reForge, or Forge Neo, feel free to check out my repositories:
I'm still learning and improving these projects, so feedback, bug reports, and suggestions are always welcome. If something is useful to your workflow, I'd be happy to hear about it.
Thanks for following my work, and happy generating! ❤️