Trained for 3 hours on a 4070ti using animagine sdxl model, and with around 45 pictures of Eris on episodes 4-6. Useful keywords for generation are: Eris Greyrat, Long red hair, Red eyes, black eyeliner, black and white dress, ahogeThis is version 1 so it may require some prompting to generate proper images, here are my recommended ones.Prompt:(best quality, masterpiece:1.4), &lt;lora:ERIS_GREYRAT:0.8&gt;, eris greyrat, long red hair, black and white dress, ahogeNegative Prompt:(bad quality, worst quality:1.5), monochromatic, improper limbs, cropped, blurry

Eris Greyrat Mushoku Tensai Sdxl 1.0 Lora

Slightly more training and hopefully corruption prevention from common prompts used to train the lora

Files with SHA256: 2ed8b7bbdcb6b2f864b2f8dadc2b6d62826afc8a22d6aacec0453ae085a1a87b