--- tags: - text-to-image - stable-diffusion - lora - diffusers - template:sd-lora widget: - text: >- Screenshot (This isn't a prompt we're just showing you the output samples from Civit trainer) output: url: images/Screenshot 2024-02-23 at 19.55.18.png - text: Screenshot output: url: images/Screenshot 2024-02-23 at 19.55.13.png - text: Screenshot output: url: images/Screenshot 2024-02-23 at 19.48.08.png base_model: stabilityai/stable-diffusion-xl-base-1.0 instance_prompt: 0b4k3sks license: creativeml-openrail-m --- # Poltergeist Mix Style for SDXL ## Model description # Details Trained off Duskart7 outputs, Poltergeist 1.5 (several versions from 3.0 onwards) as well as Osenayan Illustration. These are all AI art outputs, rather than datasets created from existing media # Scope/ Intention The intention is to attempt to RECREATE Poltergeist Mix from SD 1.5, by way of training the lora into the SDXL model. This is because while we would naturally attempt at generating millions of pictures at a time over the course of days even weeks - we don't have the funds to fully even with sponsorship, train a full SDXL model. We're still working on porting some styles to Pony/Animagine but felt that for now to keep compatibility for the larger comic crowd we'd train on base. Our Website https://end-media.org/ # WE ARE PROUDLY SPONSORED FINANCIALLY BY https://www.piratediffusion.com/ https://civitai.com WE ARE PROUDLY SUPPORTED BY https://yodayo.com/ https://diffusionhub.io Please note Suppport sites are not financial backers, but hold a special place as they've given support via either service testing, or other merits towards E&D. ## Trigger words You should use `0b4k3sks` to trigger the image generation. ## Download model Weights for this model are available in Safetensors format. [Download](/EarthnDusk/PoltergeistMixXL_Lora/tree/main) them in the Files & versions tab.