Download
1 variant available
Trained on NoobAI-XL-v1.1
Trained on 30 images based off various image generations I made with my previous artist mix, bigusmix. The end result is hopefully gens that are both sharp, colourful and detailed, assuming you're using a good model like Xeno443/StableMondAI-SDG · Hugging Face.
No activation word. Just throw it in your prompt and it should work.
Recommended settings:
Sampling method: Eular A
Sampling Steps: 35
CFG scale: 4
Image size: 1152x896
What I do is I first make a txt2img gen without using hiresfix, then I upscale the image with img2img at cfg 7 and denoise 0.45 and with the lora weight set to 0.8. This usually yields very high quality results.