A LoRA for Stable Diffusion 1.5 checkpoints, trained to mimic the art style of glyvees. glyvees is an artist known for their stylized fantasy character illustrations. Trained on 2 repeats of 24 images.

Usage Tips

• The trigger word is "hwglyvees".

• Most standard NAI-based checkpoints should be compatible with the LoRA. AOM3 and AnyLoRA seem to work the best.

• Use Booru tags when prompting.

• LoRA strength should be 1.0 to 1.2. If the style doesn't bleed enough, strength can go up to 1.5, but this may cause artifacts. Female characters may become more androgynous/masculine.

• Any DPM++ sampler at 512x768 base resolution. Hires fix is not required.

• Optionally use certain tags to emphasize the style; try "no mouth, no nose, animal hood"

Remarks

• Stable Diffusion XL/Pony versions are on my bucket list; I don't have enough VRAM to train these LoRA types yet. Be patient!

• It's ideal to use img2img, inpainting and controlnet techniques to get good images, you won't get perfect images from generations alone.