Image Generation
This section is dedicated to Image Generation.
The initial plan to consolidate images generated by the different Image Generators that were available a year ago is obsolete, because it is simply not possible to keep up with the pace at which updates and new models arrive. Therefore I shifted the focus, and will use this section to just share images I generated myself, or nice ones I come across during my journey.
About this section
The initial plan to consolidate images generated by the different Image Generators that were available a year ago is obsolete, because it is simply not possible to keep up with the pace at which updates and new models arrive. Therefore I shifted the focus, and will use this section to just share images I generated myself, or occasionally if I come across nice ones from others during my journey.
Also, since I hardly use any other Image Generators than those from Stability.ai I will talk a bit about my experiences heras I move forward .
Automatic 1111 (A1111), comfyUI and fooocus
Automatic 1111 was the first generator I installed using the guidance of one of Furkan Gözükaran's excellent tuorials on SECourses. Furkan's tutorials are very detailed, and it took a while for me to learn as I was a total newbee, but that was part of my plan anyhow. Today there are many many 1-click-installers availablen, for those who like me start from scratch.
Fooocus was the second generator I installed. It has a very simple UI interface, with lots of preset styles to make beautiful images in no time. The installation is very easy too.
ComfyUI is the 3rd application I use, and it is just brilliant. Not only is it extremely simple to install on Windows (use the portable install) but it is by far the most supported Open Source application I believe. I recommend to invest some time to "study" the many tutorials on YouTub. A good place to start would be Olivio Sarakas' "Comfy Academy" lessons here .
One of the awesome things about comfyui are the embedded workflows in the generated images. It means that you can just drag and drop any comfyUI image into you workspace and you get the full workflow. In my section Gallery, you will find comfyUI images I generated with their embedded workflows.
Please note two of the most common errors one gets are "missing nodes", which most of the time can easily be installed through the comfyUI manager noded, as well as errors that the needed models cannot be found. This is because every one has its own way of saving them into their directories.
Models, Checkpoints, Loras, Workflows, ...
Whether you use A111, fooocus or comfyUI, you need to have models, checkpoint loras and the like. There are a couple of excellent place to find them. Amongst them Hugging Face and Civitai.com. On Civitai you will also find a great collection of comfyui workflows, as well as on Openart.ai . The workflows on Openart are often very well documented.