<h2>IronCatFateToon</h2>Making models is a thing I do for fun. So if you like what I do? You can help me out by just using the model and leaving a review!IronCatFateToon's goal is to be able to create nice looking anime style images with vibrant colours, more simplistic anime (cartoony) colour/style look and with influence of Fate/Stay Night: Unlimited Blade Works styling.<s>This checkpoint </s><s>requires </s><s>a VAE, I recommend </s><a target="_blank" rel="ugc" href="https://huggingface.co/hakurei/waifu-diffusion-v1-4/tree/main/vae"><s>kl-f8-anime2</s></a><s>, I've also used it for the examples.</s> Edit: Since the original upload of this checkpoint, I've updated <code>v1.0</code> itself and reuploaded the checkpoint with the VAE baked in. If you prefer the checkpoint without it baked in, download the <code>v1.0-no-vae</code> version.In order to achieve what's just been promised I used <a target="_blank" rel="ugc" href="https://civarchive.com/models/7240/meinamix">MeinaMix V8 </a>as the base checkpoint, I then merged two of my style LoRA's into MeinaMix V8; <a target="_blank" rel="ugc" href="https://civarchive.com/models/32664/ironcatlora-3-anime-vibrancy-and-simplifier-style">Anime vibrancy &amp; simplifier Style</a> (Merge ratio of <code>0.4</code>`) &amp; <a target="_blank" rel="ugc" href="https://civarchive.com/models/35050/ironcatlora-4-fatestay-night-ubw-style">Fate/stay night: UBW Style</a> (Merge ratio of <code>0.6</code>`) . And finally I merged the result of that with the checkpoint <a target="_blank" rel="ugc" href="https://civarchive.com/models/24149/mistoonanime">Mistoon_Anime</a>. Where I did a weighted merge, 70% was used from the results of me merging the LoRA's into MeinaMix V8. And 30% was used for Mistoon_Anime. (Note: I didn't use the exact same Anime vibrancy &amp; simplifier Style LoRA when merging, I retrained on the same dataset from that LoRA for a more nuanced look that complimented MeinaMix V8 better i.m.o.)Some things you might recognize from Fate/stay night: UBW in this checkpoint, is in the way how eyes, scenery, city, cityscape, sky, grass, trees, certain 2.5D angles et cetera are generated. The anime/cartoony look is a combination from my Anime vibrancy &amp; simplifier Style LoRA and Mistoon_Anime. While MeinaMix V8 as a base gives in my own opinion a really sollid base for good looking art.I haven't used any embeddings in the example images so you should get a good idea, in what the checkpoint can create on its own. I generated the images using some wildcard libraries, so the prompts might sometimes seem random.Advice:<ul><li>I would recommended a clip skip of 2</li><li>I personally like <a target="_blank" rel="ugc" href="https://upscale.wiki/wiki/Model_Database">lollypop </a>as an upscaler a lot, it gives anime images a nice soft upscaled touch.</li><li>I am personally a fan of using DPM++ SDE Karras with 15 steps and when Hires fixing, I like to use 10 for Hires steps and a denoise of about 0.35 or 0.4. Most example images use 512x768 as base resolution.</li><li>If you want the same vibrant, high contrast style look from my images you've seen, mostly in the images with girls, I suggest the following starter prompt: <code>(masterpiece, best quality:1.4), cinematic light, colorful, high contrast</code>, the prompt <code>cinematic light</code> is optional but I've used it in most of the example images since I liked the look. The <code>high contrast</code> prompt may sometimes do bad in certain NSFW generations, if you notice it in a seed you can try to remove it.<ul><li>The starter prompt is not necessarily needed, I also included 2 examples where I just typed out a story based prompt and only used the negative prompt like below. And those images also came out great i.m.o.</li></ul></li><li>I would advice the following starter negative prompt in general to get the look you see in most of my example images: <code>(worst quality:1.2), (low quality:1.2), (lowres:1.1), (monochrome:1.1), (greyscale), multiple views, comic, sketch</code>. You can also use my full negative prompt from my sample images, if you liked the look of them.<ul><li>The <a target="_blank" rel="ugc" href="https://huggingface.co/nick-x-hacker/bad-artist">bad-artist negative embedding</a> seems to work pretty nice too, which you can try instead of my suggested negative prompt.</li></ul></li></ul>This model wouldn't have been possible without these awesome creators:<ul><li>@Meina for <a target="_blank" rel="ugc" href="https://civarchive.com/models/7240/meinamix">MeinaMix V8</a></li><li>@Inzaniak for <a target="_blank" rel="ugc" href="https://civarchive.com/models/24149/mistoonanime">Mistoon_Anime</a></li><li><a target="_blank" rel="ugc" href="https://huggingface.co/hakurei">hakurei</a> for <a target="_blank" rel="ugc" href="https://huggingface.co/hakurei/waifu-diffusion-v1-4/tree/main/vae">kl-f8-anime2</a></li></ul>

IronCatFateToon

This version of <code>v1.0</code> doesn't have a VAE baked in. The version I originally uploaded in <code>v1.0</code>, has been reuploaded with the VAE baked in.

Files with SHA256: 9c6ae4a76c21359f914a0810af2b1796b181812d93c03391e9889dfa5530036d