This is Kingabdul he's an online social media fitness model with the most insane body like a godly body!
I am using my own checkpoint model Photonic Fusion SDXL, this LoRA was also trained on this checkpoint so they pair perfectly together! For this particular LORA I have not tested it on other checkpoints or online generators so I have no idea how it performs outside of my own, sorry. Tested on A1111.
Trigger Word: (K1ngabdul), other triggers are... african american, black male, dark skin, muscular, fitness model, vascularity, vascular, toned body, detailed eyes, full lips, short dreads
I'm using Photonic Fusion SDXL V1 or V3 available on TensorArt (V1 on TensorArt is hidden in the NSFW category)
I'm using DPM++ 3M SDE with Karras
Steps 30-40, CFG 5.5-6.5
768 x 1024 or 768 x 1100 Upscaled 1.6 - 2x
Hires Fix, 13-15 Steps, 0.35-0.40 Denoising, 1x-ITF-SkinDiffDetail-Lite-v1 is the best on A1111 and ForgeUI or specifically on TensorArt I suggest using 4x_NMKD-Superscale-SP_178000_G
LORA strength between 1 or 0.9
Adetailer - face_yolov8n_pt: in the positive prompt paste <lora:(K1ngabdul)SDXLv1:0.8> (K1ngabdul), african american, black male, dark skin, detailed face, detailed lips, full lips
Adetailer - mediapipe_face_mesh_eyes_only: in the positive prompt paste <lora:PerfectEyesXL:1> muted hazel eyes, detailed eyes, matching iris <lora:(K1ngabdul)SDXLv1:0.6> (K1ngabdul)
The two Adetailer additions above are not necessary but often will help improve his face especially in full body shots.