Prompt

Halftone Recraft, A close-up of a black and white photograph of a cars tail light. The car is facing towards the right side of the frame. The tail light is glowing red. The bumper of the car is visible. The background of the photograph is white with black

Prompt

Halftone Recraft, A black and white photograph of a white sports car on a road. The car is facing the left side of the frame. There is a yellow bus behind the car. The bus has a red stripe on the side of it. There are black letters on the top of the bus.

Prompt

Halftone Recraft, Captured in a close-up, eye-level perspective, a collage of a car and a car is featured prominently in the center of the frame. The car is a silver sedan, with a black tint on the windows. The sedan is facing the viewer, while the car is facing towards the viewer. The background is a dark brown.

Prompt

Halftone Recraft, An eye-level view of a white BMW car. The car is facing the left side of the frame, and the front bumper is visible. The front grill is also visible, with a blue light on the right side. Behind the car is a red and white sign that reads "MOVEMENT" in bold black letters.

Model description for Halftone-Recraft-Flux

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 & 3100
Epoch	23	Save Every N Epochs	1

Labeling: florence2-en(natural language & English)

Total Images Used for Training : 49

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/Halftone-Recraft-Flux"
trigger_word = "Halftone Recraft"  
pipe.load_lora_weights(lora_repo)

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

Trigger words

You should use Halftone Recraft to trigger the image generation.

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.