Update README.md
Browse filesA comprehensive Stable Diffusion model for generating fantasy trading card style art, trained on all currently available Magic: the Gathering card art (~35k unique pieces of art).
Training was done on a dataset consisting of cropped, 512x512 versions of the art for every MTG card, each of which was tagged using a custom python script, from data pulled from Scryfall. Training was done with the Dreambooth extension for Automatic1111's wonderful UI.
The result is a comprehensive model that has a good understanding of MTG artists, sets, planes, card types, creature types, years, colors, and more. If you had ever wondered what a Merfolk, drawn by Ron Spencer, would have looked like on Tarkir, as part of the Mardu clan, with dash, haste, and trample - this model can deliver what you want.
Prior information has been preserved well, allowing you to incorporate non-MTG artists, people, characters, places, etc. A few examples are below.
I've done only limited model testing - this was originally meant to be a second test run, but it turned out far better than I had anticipated, so I am releasing it, and hope others enjoy it as much as I have. I've added a small guide "using the model," based on what I've determined from my limited testing.