Qwen-Image-Sketch-Smudge

Prompt

Sketch Smudge, A medium-sized sketch of a womans face is depicted on a stark white background. The womans head is facing the left side of the frame, her hair cascading over her shoulders. Her eyes are squinted and her lips are pursed. She is wearing a short-sleeved brown shirt with a collar around her neck. Her hair is a dark brown color, and her bangs are a darker shade of brown. Her eyebrows are a lighter shade of black, and she has a slight smile on her face. There are three blue circles surrounding her head, adding a pop of color to the scene.

Prompt

Sketch Smudge, A black and white portrait of a mans head and shoulders. The mans hair is short and wavy. His eyes are open and he is looking to the right. His hair is a light brown color. He is wearing a black turtleneck with a white collar around his neck. The backdrop is a dark gray color and there is a shadow on the wall behind him.

Prompt

Sketch Smudge, a pencil sketch of a womans face is displayed against a light pink background. The womans head is tilted slightly to the left, her eyes are wide open, and her lips are slightly parted. Her hair is pulled back, cascading over her shoulders, framing her face. She is wearing a sleeveless blouse adorned with a pattern of black and white flowers, adding a pop of color to her outfit. Her earrings are adorned with silver earrings. Behind her, a backdrop of brown leaves can be seen, adding depth to the scene.

---

Model description for Qwen-Image-Sketch-Smudge

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	22 & 2850
Epoch	20	Save Every N Epochs	1

Labeling: florence2-en(natural language & English)

Total Images Used for Training : 30 [HQ Images]

Data Sources

Source	Link
Playground	playground.com
ArtStation	artstation.com
4K Wallpapers	4kwallpapers.com

Best Dimensions & Inference

Dimensions	Aspect Ratio	Recommendation
1472 x 1140	4:3 (approx.)	Best
1024 x 1024	1:1	Default

Inference Range

• Recommended Inference Steps: 35-50

Setting Up

import torch
from diffusers import DiffusionPipeline

base_model = "Qwen/Qwen-Image"
pipe = DiffusionPipeline.from_pretrained(base_model, torch_dtype=torch.bfloat16)

lora_repo = "prithivMLmods/Qwen-Image-Sketch-Smudge"
trigger_word = "Sketch Smudge"  
pipe.load_lora_weights(lora_repo)

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

Trigger words

You should use Sketch Smudge to trigger the image generation.

Download model

Download them in the Files & versions tab.