Updated for use on PDXL! Same dataset, except I changed the tags from danbooru-style to e621-style, since those tags are far more robust. Pony is mostly trained on digital art, however most of the images that I found of the Stealth Suit were the 3D SFM model... So this model has a considerable bias towards 3D. It learned the details of the suit and helmet a lot better, though. So, it's got a big bias towards 3d, so just prompt accordingly to whether or not you want to embrace it. Hope you enjoy!
Updated for use on PDXL! Same dataset, except I changed the tags from danbooru-style to e621-style, since those tags are far more robust. Pony is mostly trained on digital art, however most of the images that I found of the Stealth Suit were the 3D SFM model... So this model has a considerable bias towards 3D. It learned the details of the suit and helmet a lot better, though. So, it's got a big bias towards 3d, so just prompt accordingly to whether or not you want to embrace it. Hope you enjoy!