SD3.5-Turbo-Portrait-LoRA

Prompt

Turbo Portrait, A close-up shot of a medium-sized young mans face and upper torso. The mans hair is a vibrant shade of brown, and his eyes are a piercing blue. His eyebrows are a light brown color, and he has a slight smile on his face. His lips are a darker shade of pink, with a slight pink tint. He is wearing a black short-sleeved t-shirt, and the background is a light gray.

Prompt

Turbo Portrait, A close-up portrait of a young Asian man with dark brown eyes and a black short-sleeved t-shirt. The mans face is angled slightly to the left, with a slight smile on his lips. His eyebrows are a dark brown, and he has dark brown eyebrows. He is standing in front of a plain white wall, and his hair is pulled back in a ponytail. His left ear is slightly bent at the elbow, while his right ear is facing the right. His upper lip is a lighter shade of pink, and the mans hair is a darker shade of brown.

Prompt

Turbo Portrait, A close-up shot of a womans face, positioned in front of a gray backdrop. She is wearing a maroon sleeveless tank top, adorned with a pair of silver hoop earrings. Her hair is pulled back in a ponytail, framing her face. Her eyes are a piercing blue, and her lips is a vibrant shade of pink. Her eyebrows are a light brown, adding a pop of color to the scene.

The model is still in the training phase. This is not the final version and may contain artifacts and perform poorly in some cases.

Sample Image Grid [ 3 x 2 ]

Model description

prithivMLmods/SD3.5-Large-Turbo-HyperRealistic-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	21 & 2.2K
Epoch	20	Save Every N Epochs	1

Labeling: florence2-en(natural language & English)

Total Images Used for Training : 15

Setting Up

from diffusers import DiffusionPipeline
import torch

device = "cuda" if torch.cuda.is_available() else "cpu"
model_repo_id = "stabilityai/stable-diffusion-3.5-large-turbo"

torch_dtype = torch.bfloat16 if torch.cuda.is_available() else torch.float32

pipe = DiffusionPipeline.from_pretrained(model_repo_id, torch_dtype=torch_dtype)
pipe = pipe.to(device)

pipe.load_lora_weights("strangerzonehf/SD3.5-Turbo-Portrait-LoRA", weight_name="SD3.5-Turbo-Portrait.safetensors")
trigger_word = "Turbo Portrait"  # Specify trigger word for LoRA
pipe.fuse_lora(lora_scale=1.0))

Trigger words

You should use Turbo Portrait to trigger the image generation.

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.