SHA256 hash: 2EA0C97995983B822CB8AFF049C26CE144493AA374FEFEEFC3B432E8C48FCF5E
Tags: small breast, skinny, woman, cute, pretty, nsfw, 2.5d
SHA256 hash: 35841AF675320265BDA2F9565D0A0F0415C3112638B25CFD53FDFE8F03470B39
Tags: anime, character, female, woman, hentai
虚拟现实waifu计划的第一个版本，目前主要问题是角色形象太单一，基本上完全来自anmi的风格。注意：1、目前版本的出图品质并不稳定，尽量参考例子填写提示词并设置参数以获得最佳效果。2、无需vae，模型已经内置了我最爱的vae。The main issue with the first version of the virtual reality Waifu project is that the character image is too single, basically entirely based on the style of anmi.be careful:1. The current version of the image quality is not stable. Please refer to examples to fill in prompt words and set parameters to achieve the best results.2. No need for vae, the model already has my favorite vae built-in.
SHA256 hash: 71BFCC838E9CAB7019704A8CD9787BF4378A3A54DA241456B1E451436BFE7B7D
Tags: anime, landscapes, photography, ilustrattion
This mix model is a merge model!Merged Camellia Mix, Pastel Mix, etc.Please do not use this model commercially or as a means of making personal money.
SHA256 hash: DBC0247D5DBF1216E9DEF6C5989A7C8BDEEAD80548FC3A2B526116DF4893F2AF
Tags:
55% lyriel_v13~35% chilloutmix_NiPrunedFp32Fix, deliberate_v2, lofi_V2, realisticVisionV20~10% sd-v1-4, arexMix_v01, kotosmix_v10, libertymain, protogenX34Phottorealism, qgo10bIf you use my model to make or post anything just mention this model in the Credits
SHA256 hash: 588838C1E245C7EC8C4D3DC194572650909236C82C7C4A0D366B6E62E8F3ED25
Trained words: ['Macro']
Tags: photorealistic, sexy, female, woman, portraits, realistic, 8k, 4k
Maybe I don't know where to look. But I haven't found many good futa checkpoints.I don't know anything about merging checkpoints but I used the merging tool on Automatic1111.This isA =abyssorangemix2_Hard [e714ee20aa]+B =monstercock_v15 [6d86282a0]withM = 0.45, Weighted SumVAEclearvae_main1futa, futanari, penis work well. It outputs some weird styles (presumably from monstercock) at times so I add "anime style".It works suprisingly well. Check my samples for ideas if you can't get consistent results.Feel free to improve on this or its roots.
SHA256 hash: 62B68F1181903DFB56FF2E65832F24F3E7387FF5BFFB60DEDDA3B66CC5294110
Trained words: ['what']
Tags: anime, female, futanari, cock, penis, futa, hentai
A mix of merged models, my frankenstein. Hopefully it isn't a complete mess, (still new to this). Hands kinda ew so use at your own risk.I made this by merging a realistic model I was working on to a bunch of anime models and I really enjoyed the results, thought I would share. This Model has quite a bit of anime models in it….too much to list. (if I remember I will list which ones)I recomend to use (not necessary, but I like how they make things look):https://github.com/dustysys/ddetailerhttps://civitai.com/models/7808/easynegativehttps://civitai.com/models/4629/deep-negative-v1x -- (I noticed this makes images less vibrant, kind of cool to tone down how bright colors can get)(vae I use, Necceary if you want vibrant images and not washed out ones)https://huggingface.co/hakurei/waifu-diffusion-v1-4/blob/main/vae/kl-f8-anime2.ckptDPM++ 2M Karras works good (sampling steps: 25-40).Hiresfix (step:15, denoise strength: 0.45-0.5)if faces still look ew don't forget to use inpainting to fix them, and to also fix other details.
SHA256 hash: F8DEFBE6EF6992CD5989E2A459B50C6B338137E6379AEC57E63E5C7B8277357F
Tags: anime
Halo semua gw mau Share hasil Merge model, ini bukan model yang sempurna tapi bisa menghasilakan Gambar yg cukup pas-pasan atau rata-rata ok lah hasilnya. Untuk yg tertarik dengan model ini silahkan download & review model ini jika kalian suka. Kalian juga bisa cek 👉 Yosia Kefas - Trakteer Gw bakal share semua hasil Training LoRA LoCon disana, dijamin kalian tidak akan menyesal.みんな様こんにちは、自分がMergeをしたモデルを共有したいと思いますが。このモデルは完全なモデルではありませんが、ごく平均的なモデルのです。ぜひこのモデルが良いと思っていれば、レビューしてください。Hello everyone i want to share the results of the Merge model, this is not a perfect model but can produce images that are quite mediocre or average results. You can try & review this model if you like it.Indonesian Discord Server: https://discord.gg/j8qmYPAGQhChichi pui: ヨシア＠日本語独学者🇮🇩 ✌️ | AIイラスト専用の投稿サイト | Pixiv: Yoshia Kefasu - Pixiv
SHA256 hash: 1CE3B133845EC984515E395B1F8090BD7EC4199F3264220DBBD64A1008F0DFF6
Tags: anime
I got an idea of mixing some art styles lora of furry several days ago. So I created this model. The grapefruit v4.1 model by lkena is its basic model, and this is grapefruit model's homepage https://civitai.com/models/24383/grapefruit-hentai-model.I mixed three furry art style lora model which were created by myself in it and the results were not bad. The three mixed lora models' original artists are DAGASI, shiitakemeshi and nefuraito336 on e621. They could be find on my civitai page.By using this furry modelyou could easily create:1.furry/kemono female safe pictures2.furry/kemono female nsfw pictures3.realistic furry/animal pictures4.anime/realistic human girl's safe/nsfw picturesnot easy to create:1.horizontal pictures2.male creatures3.multiple characters4.correct hand anatomy5.pictures with overly bright colourBecause the model is based on grepefruit hentai model which is focus on female, you should add lots of tags to pledge the validity of male's body anatomy if you want to create some male's pictures. you could copy my prompts of the example pictures, and the nagetive prompts I used just the easiest NovelAI's nagetive prompts. But the model may generates some pictures with wrong creature anatomy because furry character's particularity, you could add some tags to correct it. I use the VAE "vae-ft-mse-840000-ema-pruned" to create the pictures. I haven't try the effect of Hire.fix on this model.clip recommendation:
SHA256 hash: B71363A9C91F5AE4A438BA0B2F992CA1520CCE2317E1A44FA8E6C1CDA794D7F5
Tags: anime, kemono, furry, cartoon, 1girl
THIS IS A 2.1 BASED MODEL!!!! BEFORE COMMENTING TAKE A MOMENT TO CONSIDER AM I USING 2.1 BASED LORAS OR CONTRONET!!!!
If you'd like to share your creations, please join us on Discord -
https://discord.gg/HqdffGgeBa
CONTROLNET NOW SUPPORTED!
https://huggingface.co/thibaud/controlnet-sd21
Please see the instructions here, featureset is reduced but we now have canny and depthmap support. Thank you thibaud🙏
Important
This model is designed to be used with three negative embeds, please download these as well as the main model file, adding them to your AUTOMATIC1111 webui embeddings folder
https://civitai.com/models/13515/nfixer-for-illuminati-diffusion-v11
https://civitai.com/models/13518/nartfixer-for-illuminati-diffusion-v11
https://civitai.com/models/13519/nrealfixer-for-illuminati-diffusion-v11
Please note that the safe range for good gens is between 768-1024 / 1152 for some seeds and prompts / 1280x768 or 768x1280 in rare instances will also work. This seems to be relatively fundamental to SD2.1.
Please see example images to get an idea how to prompt. No specific trigger words are necessary, this is a full finetune.
If you are an advanced user, as always, go ahead and just experiment with negative prompting or other embeds, but please not that this model was trained with offset noise, as such, other embeddings may reduce the quality of the contrast that has been introduced.
Please note that some of the sample gens have SD upscale applied to them, to reproduce, please omit this info when importing into AUTO1111 webui.
License -
The following license agreement governs the use of the Illuminati Diffusion v1.1 model, which is a modified CreativeML OpenRAIL-M license. By accessing, downloading, or using the model or its derivatives, you agree to comply with all terms and conditions set forth in this license agreement.Leoanrdo.ai is the only official generative AI platform licensed for the new model.You are strictly prohibited from hosting or using the model or any of its derivatives on any website, app, or digital platform from which you currently earn, plan to earn, or may potentially earn revenue or donations without obtaining prior written consent from the authors.Furthermore, you may host the model or any of its derivatives on completely non-commercial websites, apps, or other digital platforms, meaning that you are not earning any revenue or accepting any donations.You may upload the model or its derivatives for personal compute usage on services such as Google Colab, provided that such usage is solely for personal and non-commercial purposes, and does not involve any revenue or donations. However, you must include a link to the model card and clearly state the full model name (Illuminati Diffusion v1.1) in the uploaded materials.Note that such personal compute usage does not permit any redistribution of the model or its derivatives. If you wish to use the model for any other purposes, please refer to the other sections of this license agreement.Note that the outputs of the model or its derivatives can be used for commercial purposes by teams of 5 or fewer people.It is strictly prohibited to use the model to deliberately produce or share illegal or harmful content. Violation of this provision may result in legal action against you.The authors of the model claim no rights over the outputs that you generate. However, you are solely responsible for their use, which must not go against the provisions set in this license.Redistribution of the model weights is not permitted.By accessing, downloading, or using the model or its derivatives, you acknowledge that you have read and understood the terms and conditions of this license agreement and agree to comply with all of them.
Illuminati Diffusion v1.1Illuminati Diffusion is a latent text-to-image diffusion model that has been conditioned on high aesthetic synthetic images through fine-tuning. It was trained on 82,000 images locally on my PC with a single 3090ti, taking over 100 hours.
Resumed training with noise offset for dramatically improved depth of contrast and colour. Stable diffusion 2.1 768.
Follow me on Twitter - https://twitter.com/cac0e
SHA256 hash: CAE1BEE30E67339DD931925F98C12D2B86FE9F4786795137040E4956F4BFCFFE
Tags: general purpose, fine-tune, artistic, realistic
