This model is a reimagining of the fictional character Manuela Vicious, by Tower13Studios.
Flux.2 Klein 9B
Checkpoint: Flux.2 Klein 9B
LoRA strength: 0.8–1.6
Sampler: euler, euler_ancestral, res_multistep, or dpmpp_2s_ancestral
Scheduler: simple or beta
Steps: 4–12; 30–50 for base model
CFG Scale: 1, or try 1.2–3; 2–5 for base model
Trigger word (for superior results): ATRX_MANUELA_FLUX2KLEIN9B
Z-Image —Turbo
Checkpoint: Z-Image Turbo, or try Z-Image [Base]
LoRA strength: 1–1.6
Sampler: dpmpp_2s_ancestral
Scheduler: simple or beta
Steps: 4–12
CFG Scale: 1, or try 1.2–3; CFG Norm: 1–1.12
Shift: 7 (first sampler) and 0.6 (second sampler)
Trigger word (for superior results): ATRX_MANUELA_ZIMAGE
Negative prompt (as needed): "Asian woman, black hair"
Qwen-Image 2512
Checkpoint: Qwen-Image 2512
LoRA strength: 0.8–1
Sampler: res_multistep
Scheduler: simple
Steps: 20+, or 4–8 with lightning LoRA
CFG Scale: 2–5, or 1 with lightning LoRA; CFG Norm: 0.94–0.98
Shift: 3.1
Trigger word (strictly required!): ATRX_MANUELA_QWEN2512
Qwen-Image
Checkpoint: Qwen-Image
LoRA strength: 0.75–1
Sampler: euler
Scheduler: simple
Steps: 15+, but 20+ is better, or 4–8 with lightning LoRA
CFG Scale: 2–5, or 1 with lightning LoRA; CFG Norm: 0.8–0.9
Shift: 3.1
IMG-2-IMG denoise: 0.2–0.6
Flux.1
Checkpoint: Flux.Dev 1.0
LoRA strength: 0.8–1.4
Sampler: deis or euler
Scheduler: beta or simple (or ddim_uniform for different look, but add steps)
Steps: 15+, but 20+ is better, or 8–12 with turbo LoRA
CFG Scale: 2–5
IMG-2-IMG denoise: 0.65–0.85
Pony
Checkpoint: Real Dream SDXL Pony 12
LoRA strength: 0.6–1
Sampler: DPM++ SDE or euler
Scheduler: turbo or simple
Steps: 20–30+, or 8–12 with turbo LoRA
CFG Scale: 2–5
ADetailer recommended
Description
Pair with a realism LoRA. Can produce artistic nudes pretty well on its own. Works well for img-2-img with denoise 0.2–0.6.
Trained using AI Toolkit following this official tutorial, but the learning rate seemed to be too low so I continued rerunning with additional steps until it was something over 20,000, I think? At some point in the training, I also recall changing the linear rank from 16 to 32 and some other adjustments.
The dataset included 44 diverse images—different settings, hair styles, emotions, etc.—with almost all being upper body or closer and about one-fourth nudes. The captions were drafted mostly by Mistrel Le Chat—which didn't seem to mind the nudity—and Qwen at about 100 words, including the trigger word ATRX_MANUELA_QWEN, although the LoRA seems to respond without it. The captions described all of the elements not associated with the character or of temporal nature (e.g., hair style and clothing).