See <a target="_blank" rel="ugc" href="https://civarchive.com/articles/1771">here</a> for an extensive article about my entire training workflow and configuration if you want to train your own SDXL 1.0 LoRa's and achieve results similar as me!Recommended weight: 1.0 - or lower when you combine it with other LoRa'sRecommended model and VAE: <a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/sd_xl_base_1.0_0.9vae.safetensors">SDXL 1.0 with 0.9 vae</a>Recommended settings: 35 steps | DPM++ 2M Karras | CFG 7Recommended negative prompt: text, watermark, signature, abstract, burned, instagram, low-quality, low-res, action figure, child drawing, jpeg, low contrast, glossy, puppet, cliché, stock photo, magazine coverDescriptionA LoRa for SDXL 1.0 for a more consistent Supergirl outfit than the one base SDXL provides. It should work for any and all models and LoRa combinations, but you may have to adjust weights for some.Weights used for some of the sample images:<ul><li>Ghibli style sample image: Ghibli LoRa 1.0 | Supergirl LoRa 1.0</li><li>Nausicaä character sample image: Ghibli LoRa 0.1 | Supergirl LoRa 0.8 | Nausicaä LoRa 0.5</li><li>Dark and Darker style sample image: Dark and Darker LoRa 0.9 | Supergirl LoRa 0.9</li><li>Inkpunk style sample image: Inkpunk LoRa 0.8 | Supergirl LoRa 0.8</li><li>Zelensky celebrity sample image: Zelensky LoRa 0.3 | Supergirl LoRa 1.0</li></ul>Trigger words<ul><li>"wearing supergirl outfit with cape"</li></ul>Changelogv2.0<ul><li>much less overtraining, more flexibility</li></ul>v1.0<ul><li>Initial release</li></ul>Additional InfoCredit is not mandatory but I would still appreciate credit when posting images using these LoRa's on social media! Feel free to @ my Twitter and/or Instagram accounts too (@ai_characters)! I would love to know when it is used!I share all of my model's and training knowledge for free, but donations to my <a target="_blank" rel="ugc" href="https://ko-fi.com/aicharacters">Ko-Fi</a> would be greatly appreciated! Testing so much and training so many model's costs a lot of money (and time).Feel free to also try out my RVC2 AI voice models which you can find <a target="_blank" rel="ugc" href="https://huggingface.co/ai-characters/AI-Voice-Models-by-AI_Characters">here</a>!Stats (v2.0)Dataset:<ul><li>21x fanart and cosplay photos</li></ul>Epoch chosen for the final model:<ul><li>50</li></ul>Learning rate:<ul><li>2e-5</li></ul>For the configuration used to train this model see my guide which I linked in the beginning of this post.

Files with SHA256: 38f6de00af3ae8ca32a42ba9893ac072ede7ca4578c8dcf85b16c32762edcca6