You can try any model you want, however I really suggest using <a target="_blank" rel="ugc" href="https://civarchive.com/models/288584?modelVersionId=324524">AutismMix Checkpoint</a>This model was trained using:<ul><li>1024,1024 size images</li><li>400 images</li><li>1 step per image</li><li>20 epochs</li><li>4 batch</li><li>2000 steps</li></ul>LoRA settings:<ul><li>Weight: 0.5 - 0.8 (Might be a bit undertrained, if paired with multiple loras with high weight this gives bad results, might fix in a future release. For consistency try to avoid weights of 0.9 - 1.0 and stick to 0.5)</li></ul>Base model settings:<ul><li>Steps: &lt;30</li><li>Width: 896px</li><li>Height: 1132px</li><li>Sampler: Euler a</li><li>VAE: None</li><li>CFG Scale: 3 - 5</li><li>Clip skip: 2</li></ul>Upscaler settings:<ul><li>Upscaler: Latent (Nearest-Exact)</li><li>Upscale by: 1.5x</li><li>Denoising strength: 0.55</li></ul>

Files with SHA256: b32739b17978ca9f1ddd5cd7ccb0ffbb7bc0616ff818b98ed1c027ca06511a08