Main prompt tags: jovithordennis <------ should always be included
Style control:
for the original artist style include cartoon in positives
Recommended samplers:
- Euler Ancestral Scheduler: Normal CFG: 3.0-5.0 Steps: 10-20 Lora strength: 1.0
Note: I recommend low CFG and lower step count. For higher CFG/steps latent upscaling and face/eye detailer fixes are recommened.
Prompt tips (green - positive, red - negative):
- greyscale - helps exclude greyscale images which he was trained on
- vest, black shirt, tank top, spiked bracelet, sweatband, black pants, black boots, gold earrings, black earrings, torn sleeves, torn clothing - common clothing items for Dennis
- chubby / fat - in case he is not generated fat enough
- 5 fingers - helps fix hands
- 1boy / male focus - shouldn't be needed, can be included in case female biased models flip his gender
The LoRa should support all kinds of Illustrious checkpoints. Tested working well with the following:
WAI-NSFW-illustrious-SDXL v12.0
HomoSimile ILXL v2.0
To improve details:
- generate using higher base resolution (e.g. 1384x1384) 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:
Requires download of FaceDetailer custom nodes + Eyeful_v2-Paired.pt + FacesV1.pt models (can be Googled). Workflow includes a HiresFix pass with x1.30 latent upscaling by default.
For portraits/upper body shots I recommend lowering the upscaling ratio ('Upscale Latent By' node) - otherwise you might get distorted details.