Prompt

Sketch Light, A black and white drawing of a mans face. The man has a black face mask covering his mouth. His eyes are black and he has blonde hair. His eyebrows are black. His mouth is covered with a black mask. The mask has a red line coming from the left side of the mask. He is wearing a black sweater over his neck. The background is a grayish gray.

Prompt

Sketch Light, An eye-level drawing of a womans face with long, wavy hair. The womans eyes are a piercing blue, and her lips are a lighter shade of blue. Her eyebrows are a darker shade of brown, and she has a slight smile on her lips. The background is a dark gray, and there is a star in the lower left corner of the image.

Prompt

Sketch Light, A digital painting of a womans head and neck. The womans hair is styled in a bun. The hair is a vibrant orange and black. The background is a dark gray color.

Prompt

Sketch Light, A detailed black-and-white sketch of a man sitting on a stool, leaning forward with his elbows on his knees. He is wearing a leather jacket, and his hair is styled in a messy pompadour. His eyes are shadowed, but his intense expression is still visible. The background is dark with faint streaks of light creating a moody atmosphere.

Prompt

Sketch Light, A minimalist black and white drawing of a young man wearing a hoodie with the hood pulled up. His face is partially hidden in shadow, but his sharp black eyes and serious expression stand out. The drawstrings of the hoodie are slightly frayed. The background is a textured dark gray with faint white diagonal lines adding depth.

Prompt

-Sketch Light, A black-and-white sketch of a boy standing in the doorway of an old wooden house. He is holding a lantern that casts a warm glow on his face, illuminating his curious expression. His outfit is simple, with suspenders over a button-up shirt. The background is a textured gray with faint woodgrain patterns adding depth.

Model description for Flux-Sketch-Light-LoRA

Image Processing Parameters

Parameter	Value	Parameter	Value
LR Scheduler	constant	Noise Offset	0.03
Optimizer	AdamW	Multires Noise Discount	0.1
Network Dim	64	Multires Noise Iterations	10
Network Alpha	32	Repeat & Steps	18 & 2190
Epoch	15	Save Every N Epochs	1

Labeling: florence2-en(natural language & English)

Total Images Used for Training : 14

Best Dimensions & Inference

Dimensions	Aspect Ratio	Recommendation
1280 x 832	3:2	Best
1024 x 1024	1:1	Default

Inference Range

• Recommended Inference Steps: 30–35

Setting Up

import torch
from pipelines import DiffusionPipeline

base_model = "black-forest-labs/FLUX.1-dev"
pipe = DiffusionPipeline.from_pretrained(base_model, torch_dtype=torch.bfloat16)

lora_repo = "strangerzonehf/Flux-Sketch-Light-LoRA"
trigger_word = "Sketch Light"  
pipe.load_lora_weights(lora_repo)

device = torch.device("cuda")
pipe.to(device)

Trigger words

You should use Sketch Light to trigger the image generation.

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.