Beardsleyflux LoRA BySilverAgePoets

🚀 Aubrey Beardsley Style via a Tokyo Expo Selection

This project is a Low(ish) Rank Adapter (LoRA) for FLUX, fine - tuned to capture the unique style of Aubrey Beardsley. It offers a way to generate text - to - image outputs in the iconic Beardsley style, leveraging a set of exclusive photos from a Tokyo exhibition.

✨ Features

• Unique Training Data: Fine - tuned on about 40 exclusive and original photos of Beardsley originals currently on display in Tokyo.
• High - Quality Parameters: With a relatively high rank (64), step - count (4000), and training resolution (1280 and 768).
• Specific Fine - Tuning: Fine - tuned at a 2e - 4 learning rate on one A100 over the Abliterated (V.2) (de - censored) modification of Flux Dev.

📦 Installation

To use this LoRA, you need to have the diffusers library installed. You can install it via the following command (assuming you have the necessary Python environment set up):

pip install diffusers

💻 Usage Examples

Basic Usage

from diffusers import AutoPipelineForText2Image
import torch

pipeline = AutoPipelineForText2Image.from_pretrained('black-forest-labs/FLUX.1-dev', torch_dtype=torch.float16).to('cuda')
pipeline.load_lora_weights('AlekseyCalvin/BeardsleyFLUX_LoRA_BySilverAgePoets')
image = pipeline('your prompt').images[0]

Trigger Words

You should use trigger words like Beardsley style illustration clip, by Aubrey Beardsley, Edwardian erotic Symbolist Decadent india ink masterful cartoon art by Aubrey Beardsley, Refined masterful line - work to generate outputs in the desired Beardsley style.

📄 License

This project is licensed under the Apache 2.0 license.

Information Table