This Model work better with <a target="_blank" rel="ugc" href="https://github.com/comfyanonymous/ComfyUI">ComfyUI</a>. here some workflow <a target="_blank" rel="ugc" href="https://huggingface.co/BoRnNo0b/files-mirror/tree/main">My-ComfyUI-workflow.</a>You can use this <a target="_blank" rel="ugc" href="https://civarchive.com/models/23906/kl-f8-anime2-vae">kl-f8-anime2 VAE,</a> <a target="_blank" rel="ugc" href="https://civarchive.com/models/106225/matrix-vae">Matrix-VAE, </a><a target="_blank" rel="ugc" href="https://civarchive.com/models/110630/anything-model-vae-v40">Anything model VAE</a> or any VAE you want.Steps: 8~30Sampler: lcm/sgm_uniform or dpmpp_2m/karrasCFG scale: 1.5~5I use Hires upscaler: <a target="_blank" rel="ugc" href="https://huggingface.co/BoRnNo0b/files-mirror/tree/main/ESRGAN">4x_fatal_Anime_500000_G</a>but you use what you see fit.Q. What hard I use ?A. <a target="_blank" rel="ugc" href="https://civarchive.com/models/47085/envybetterhands-locon">EnvyBetterHands LoCon</a>

Matrix-Anime

matrixAnime_v18VAE.safetensors

v1.8-VAE

v2.8

v2.6

v2.0

v1.6-VAE

DescriptionTypical anime mix that got purged.. One of my favorites - didn't want it to disappear.Prompting Info<ul><li>Works really effortlessly with out too much hand holding or complexity.</li><li>Mixes amazing with LORA</li></ul>Note on V18:<ul><li>The model has a baked in VAE (In A111 select Automatic for your VAE) it's pretty good, I believe it's just "matrixVAE_v30" - so if you already use that then you could probably ignore / prune it.<ul><li>I try to preserve these models identically to how they were.</li></ul></li></ul>Negative Prompt<pre><code>(worst quality, low quality:1.4), 
EasyNegative</code></pre>CreditsDataset / Trainer:<ul><li>@Lawsuit<ul><li>Unsure on what mixes went into this</li></ul></li></ul>

Files with SHA256: 2f256677f35a6e95ae2348b2eded3b46d2f01e17330202cc9963b230548edeef