Main prompt tags: gruffigummixl, anthro, brown bear, solo, green cap, beige tunic <------ should always be included
Style control:
for the retro TV show style use following positive prompts: cartoon, toony, retro
to avoid the retro cartoon style put cartoon, toony, retro into negative prompt
for mixed style only put cartoon into positive prompt
Recommended samplers:
- Euler Ancestral Scheduler: Normal CFG: 4.0-6.0 Steps: 15-20 Lora strength: 1.0
Note: I recommended starting with lower CFG/steps first
Prompt tips (green - positive, red - negative):
- beige tunic, green cap, belt - full base clothing
- >:( - his default grumpy expression, 0.9 weight recommended
- solid eyes - generates with pure black eyes (no white pupils)
- 4 fingers / four fingers should help to get the canonical version of Gummi hands
- no tail to help the canonical lack of tails for Gummi bears
- tail, animal tail, bear tail - same as above but with further negatives
- nostril, nostrils, bear nose - helps remove nostril that sometimes appears
- vhs filter will give the style even more retro feel closer to the original cartoon
The LoRa should support all kinds of Illustrious checkpoints. Tested working well with the following:
Nova Furry XL Illustrious V50
WAI-NSFW-illustrious-SDXL v12.0
WAI-NSFW-illustrious-SDXL v14.0
plantMilk - Flax
To improve details:
- generate using higher base resolution (e.g. 1216x1216) or use latent upscaling / highres fix - especially when trying to do full body poses
- for fixing wonky face/eyes I use FaceDetailer node (comfyui) + Adetailer Face finder model
Last image in the gallery can be imported into ComfyUI with a basic workflow:
No custom nodes required. Workflow includes a HiresFix pass with x1.15 latent upscaling by default. For cowboy shot/full body shots I recommend increasing the upscaling ratio ('Upscale Latent By' node) - helps with details/quality.