Sign In

Josan Gonzalez 🎨 Flux.1-D

219

Download

1 variant available

SafeTensor

164 MB

Verified:

Type
LoRA
Stats

284

219

536

Reviews
Published

Sep 14, 2024

Base Model

Flux.1 D

Training
Steps: 5,000
Usage Tips
Strength: 1
Hash
AutoV2
816306B360
Trigger Words
In style of Josan Gonzalez
default creator card background decoration
Uploads - 24

24

Downloads - 18710

18.7K

The FLUX.1 [dev] Model is licensed by Black Forest Labs. Inc. under the FLUX.1 [dev] Non-Commercial License. Copyright Black Forest Labs. Inc.

IN NO EVENT SHALL BLACK FOREST LABS, INC. BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.

ComfyUI_temp_jyees_00439_.png

Overview

This is a LoRA that attempts to emulate the style of Josan Gonzalez (also known as Deathburger), a Spanish artist and illustrator renowned for his science fiction works in the cyberpunk genre. His style is characterized by high levels of detail, vibrant colors, and the use of the ligne claire technique. Despite the grim themes, his works are distinguished by their lightness and playfulness, often incorporating elements of satire. His art is influenced by artists such as Moebius, Geof Darrow, Katsuhiro Otomo, and other masters of science fiction and comics. His works attract attention with their unique combination of 1980s nostalgic aesthetics and a contemporary vision of the future, making him one of the bright representatives of modern science fiction illustration.

Usage

Showcase images were generated in ComfyUI with the following parameters (other values may work correctly, but were not tested):

Model: flux1-dev (fp8e4m3fn)
Text Encoder: t5pxxl_fp16
Sampler: euler
Scheduler: 24 steps (normal)
Flux Guidance: 4
LoRA Strength: 1

Training

The fine-tuning process, carried out on an RTX 3090, involved 266 images, captioned using Florence-2 ("<MORE_DETAILED_CAPTION>" task), prefixed with the phrase "In style of Josan Gonzalez." LoRA was trained using the AI-Toolkit with the following hyperparameters:

Rank/alpha: 16/16
Batch Size: 1
Steps: 5000
Learning Rate: 1e-4
Optimizer: adamw8bit