For best outputs and usage:

Sampler: DPM++ 2M Karras.

Steps: 35, cfg 7, CLIP skip 2.

VAE: none required.

Pos prompt: wrap your text between "score_9, score_8_up, score_7_up, realistic, photo", being "realistic" and "photo" two of the model's triggers, and "(b0dy3:1.3), ", being "body3" other of the model's triggers, and the LoRa "add-detail-xl" included in my example runs.

Note about the body types: you can use "body1" to create petite bodies, "body2" to create athletic, "body3" to create thick bodies, and "body4" to create chubby bodies.

Note about the pos prompt: you may need to add big weights specially when naming the positions, or if you're gonna try to create furry, for example, "(source_furry)" or "((doggystyle))".

Neg prompt: "score_3, score_2, score_1, source_anime, source_cartoon, (blur), (blurry), mosaic censoring, bar censoring, censored, text, signature, logo, artist_name, username, crop, frame, (simple background), worst quality, low quality, normal quality, messy drawing, amateur drawing, lowres, bad anatomy, bad hands, source filmmaker, 3d, old woman, granny, showing teeth, muscles, censor, bar censor, mosaic censorship, fat".

Checkpoint Merge.