A merge of Autismmix and couple others, that turned out so well that I decided to share it with y'all. Has a VAE baked in.Resolutions: (it's the same with all xl models, but..)<ul><li>1024 x 1024</li><li>1152 x 896</li><li>896 x 1152</li><li>1216 x 832</li><li>832 x 1216</li><li>1344 x 768</li><li>768 x 1344</li><li>1536 x 640</li><li>640 x 1536</li></ul>Recommended settings: (use whatever you like, these work for me)<ul><li>Sampling method: Euler Ancestral</li><li>Sampling steps: 30</li><li>Schedule Type: Automatic</li><li>CFG Scale 5 - 8</li></ul>Recommended positive prompt:score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_anime, (other sources also work) rating_explicit, (if sfw then rating_safe, or rating_questionable) BREAK, (all the rest of the stuff), also recommend using "detailed background" and then describing what the background is/has.recommended negative prompt: DO NOT USE SCHIZO PROMPTS! all it does is tank the quality! you do not NEED to put anything in here! If you want more cartoonish images, use "realistic, 3d". I also like to use "simple background" just to reinforce the positive prompt, but again, not needed.I have generated the sample images using plain txt2img, and then giving them couple rounds in img2img. The resizing is done in img2img also, with denoising set to 0.1 -0.3.

Files with SHA256: 83145996eceebf0498ec3820202653077d8fc5716da232d94a710c5402d7fd7c