v2.0:Named to stay inline with Anima version. Dataset is the same, except with all natural language removed, and should be accessible for download. Has stable concept of "intestines filled" and wague concept "colon filled" (more grounded version, but works much worse due to tag not being in default checkpoints). Illustrious version specifically has strange tendency to generate multiple views of a character, like in reference picture / concept art. Even more so, than Anima. Reason is unknown, but if anyone has any ideas, feel free to comment. I will try to fix it later.Base: Twin for my existing model. (<a target="_blank" rel="ugc" href="https://civarchive.com/models/2907727/intestines-filled-inflatedstuffedbulging">https://civarchive.com/models/2907727/intestines-filled-inflatedstuffedbulging</a>) Intended to produce images, where character has visible relief of intestines bulging on their stomach (booru tag "<a target="_blank" rel="ugc" href="https://danbooru.donmai.us/posts?tags=intestines_filled">intestines filled</a>"). Dataset is mostly synthetic, as it is surprisingly hard to find good images with this concept targetly.

Files with SHA256: bff5d4b4e4319d8f9cfab7748ca7cfe87f2122d003c5b005adfb12deb9dafa7a