A LoRA for Stable Diffusion 1.5 checkpoints, trained to mimic the art style of Hideyuki Tanaka. Hideyuki Tanaka is a 3D designer, animator, character designer and founder of FrameGraphics. He is known for his surreal, colorful, and contemporary illustrations; he is the character designer for Metro's Bust-A-Groove/Bust-A-Move video game series and the designer and director of Super Milk Chan. Trained on about a hundred pictures of Tanaka's work, both 2D illustrations and 3D models.<h2 id="usage-tips-pch7x2os4">Usage Tips</h2><ul><li>The trigger word is "hwhideyukitanaka".</li><li>Use <a target="_blank" rel="ugc" href="https://civarchive.com/models/9942/abyssorangemix3-aom3">AbyssOrangeMix3</a> for best results. Most standard NAI-based checkpoints should be compatible with the LoRA, but this checkpoint seems to render the style with the most accuracy.</li><li>Use Booru language when prompting.</li><li>LoRA strength should be 0.8 to 1.0.</li><li>DPM++ SDE at 512x768 base resolution. Hires fix is not required.</li><li>Optionally use certain tags to emphasize the style; try "solid oval eyes" and/or "flat color" in your prompt.</li></ul><h2 id="remarks-5bm8zjxom">Remarks</h2><ul><li>Stable Diffusion XL/Pony versions are on my bucket list; I don't have enough VRAM to train these LoRA types yet. Be patient!</li><li>It's ideal to use img2img, inpainting and controlnet techniques to get good images, you won't get perfect images from generations alone.</li></ul><div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/VUFORh6Ps8Y" start="0"></iframe></div>w e l o v e m u s i c ~ !!

Files with SHA256: 41e5e3bce89ae28a506a045173267430ee96d724b1e7b1cc9eb27cddea1302fb