This is a demonstration link of typical model and style of lora using model KBlueLeaf/kohaku-v2.1 - https://civitai.com/models/136268/kohaku-v2
As your instructors illustrated in our class, in the next two weeks(by October 24, 2024), you would explore the text-to-image tools illustrated in the class(text-to-image), image-to-image, faceswap, controlnets, LORAs etc) and use them to create an art workflow to post on our shared website.
For the assignment two part one, you have already uploaded some “concepts” that you think weren’t available in popular models. We have selectively trained some of them to be used in this assignment part two.
You can find them in - LORA and Embeddings files
Explore the trained models and select the one you like. We haven’t trained all the concepts as there were concepts that do exist in popular models.
In previous class, we have shortly introduced how to use ComfyUI. In this assignment, you would need to create a ComfyUI workflow that uses the selected Lora. The workflows can be downloaded from the above LORA and Embeddings links.
Be sure to check the following videos to get started with ComfyUI.
After you have gone through the above video, you then go on to create a workflow that does something combining the various tools showcased in the videos.
You submit the files(PNG) created by your ComfyUI workflow in the following link.