All images in this folder, except example-2-quality-settings.png, have the workflow embedded and can be loaded into
ComfyUI to re-create them.
The workflow can also save a separate text file with the parameters that were used for generating an image.
To do that either load the images via the Load button or just drag one of the images
on the browser window that has ComfyUI loaded.
This image uses the default settings from the .json file and should generate quickly. These settings are a great
starting point to quickly test a prompt or generate batches of images to pick from for further refinement.
This is a variation of example 1, but with some settings changed to produce a higher quality and higher resolution result in the end.
Using a prompt from one of the signature images that were used in an older version's release on CivitAI.
This was an idea for some test images generated with an older version of this workflow during development.
Who doesn't like dragons? We had to have one in these examples.
Based on a prompt idea I found on one of the Discord servers. Re-creating a similar effect without the use of a LoRA, but using one of the 5 included controlnet/revision stages instead.
Using one of the 5 included controlnet/revision stages to create a realistic looking image of a warrior.
This is the same image, but without using controlnet
