<a target="_blank" rel="ugc" href="https://huggingface.co/Crowlley/Krea2Neutrino"><img src="https://img.shields.io/badge/Hugging%20Face-Model%20Page-FFD21E?style=flat&amp;logo=huggingface&amp;logoColor=black" alt="Hugging Face" /></a> <a target="_blank" rel="ugc" href="https://www.fiverr.com/partai/develop-your-custom-lora-or-loha-model-for-stable-diffusion?utm_campaign=gigs_show&amp;utm_medium=shared&amp;utm_source=copy_link&amp;utm_term=q74apzy"><img src="https://img.shields.io/badge/Fiverr-Hire%20Me-1DBF73?style=flat&amp;logo=fiverr&amp;logoColor=white" alt="Fiverr" /></a> <a target="_blank" rel="ugc" href="https://ko-fi.com/part_loras"><img src="https://img.shields.io/badge/Ko--fi-Donate-FF5E5B?style=flat&amp;logo=ko-fi&amp;logoColor=white" alt="Ko-fi" /></a><h3 id="clip-skip-1.-model-was-trained-using-clip-skip-1.">CLIP SKIP = 1. Model was trained using CLIP SKIP 1.</h3><h3 id="i-suggest-using-this-vae.">I suggest using this <a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/sd-vae-ft-mse-original/tree/main">VAE</a>.</h3><h3 id="this-model-was-trained-and-never-merged.-i-suggest-you-merge-it-yourself.-i-did-not-train-it-as-much-as-i-wanted-but-146-hours-of-training-using-free-colab-is-rough.-for-comparison-sake-a-person-i-know-trained-for-2x-more-steps-than-me-in-12-hours.">This model was trained and never merged. I suggest you merge it yourself. I did not train it as much as I wanted; but 146 hours of training using Free Colab is rough. For comparison sake, a person I know trained for 2x more steps than me in 12 hours.</h3>It struggles a bit with img2img; you can and you should merge it. Art style is a bit random but you can prompt for a lot of styles since is trained from a base model.The main objective of this model is to train things popular anime models struggle with and experiment with highly curated captions/tags. Took me 100 hours to fix and tag the 5610 images used to train; I had to stop and code multiple features into <a target="_blank" rel="ugc" href="https://civarchive.com/articles/2079/dataset-all-in-one-tools">my tools</a> to make my life easier.Other than that, just use normal Booru-style prompting.If you want to use a LoRA, it's a good idea to use the LoRA at low weights. Some of the tags I trained:<ul><li>Adult characters with small breasts and flat chest; tagged breasts size correctly (in my own opinion, of course)</li><li>Also trained fishnets like fishnet tighhighs, fishnet armwear, fishnet top, fishnet pantyhose; used a very conservative type of textures since SD can't really deal with tiny squares that well cus VAE destroy it</li><li>Pearky breasts</li><li>Futanari, penis, testicles, bulge, erection under clothes, condom on penis, erection, flaccid, penis sizes, ball bra, cock ring, tucked penis, foreskin, phimosis, chastity, (these last 3 don't work well yet; needs more training)</li><li>Woman, pussy, cameltoe</li><li>Crotchless panties and cupless bra</li><li>Fantasy skin colors like blue, red, etc</li><li>Monster characters</li><li>Trained piercing and tattoos with correct tagging but doesn't seem to be working yet</li><li>Butt plug, heart butt plug, anal tail</li><li>Tentacle sex, consensual tentacles</li><li>Irrumatio, fellatio, cunnilingus</li><li>Also trained a lot of cum related tags, like projectile cum, and others</li><li>Pubic hair peek, female pubic hair, heart-shaped pubic hair (last one works but its kinda meh)</li><li>Clothing aside like panties aside, leotard aside and others; but it seems to be confused with crotchless panties most of the time</li><li>Doublouse stuff, kinda hard to pull off still</li></ul>I'm sure I'm forgetting a lot of things but thats most of it.Model is heavily biased towards MY own taste in artstyles and I avoided training young characters in.

[WIP] Neutrino - TRAINED MODEL

Struggles a bit with img2img and talking with some people I know, it seems like it needs more training to fix that.

Files with SHA256: 0e98b98d356b0acaad89ac079cb832dc9a0ca8c8b2a94fa1d997aea84ece844a