Flux-Kontext-Style / README.md
ginipick's picture
Update README.md
6bd2ff0 verified
metadata
title: Flux Kontext Style
emoji: ๐ŸŒ–
colorFrom: green
colorTo: yellow
sdk: gradio
sdk_version: 5.36.2
app_file: app.py
pinned: false
license: apache-2.0
short_description: Transform images into 22 stunning art styles

Link: https://huggingface.co/spaces/ginigen/Flux-Kontext-Style

๐ŸŽจ Flux Kontext Style LoRA - AI-Powered Artistic Style Transfer

Transform your images into 22 stunning art styles with state-of-the-art AI

Features โ€ข Quick Start โ€ข Styles โ€ข Installation โ€ข Usage โ€ข Community

๐Ÿš€ What is Flux Kontext Style LoRA? Flux Kontext Style LoRA is a cutting-edge web application that leverages the power of FLUX.1-Kontext-dev model with 22 professionally crafted style LoRAs to transform ordinary images into extraordinary artistic masterpieces. Whether you're a digital artist, content creator, or AI enthusiast, this tool opens up endless creative possibilities. โœจ Key Features ๐ŸŽฏ One-Click Style Transfer

Visual Style Selection: Browse and click thumbnails to instantly apply styles Real-Time Preview: See your style selection update immediately High-Resolution Output: All images rendered at 1024x1024 resolution

๐ŸŽจ 22 Professional Art Styles Transform your images into:

Anime & Cartoon: Studio Ghibli, American Cartoon, JoJo's Bizarre Adventure, Snoopy, Rick & Morty 3D & Geometric: 3D Chibi, Low Poly, LEGO, Clay Toy Traditional Art: Chinese Ink Painting, Oil Painting, Van Gogh, Picasso, Pop Art Craft & Material: Fabric, Origami, Paper Cutting, Macaron Digital & Modern: Pixel Art, Line Art, Vector Graphics

โšก Performance & Quality

GPU Accelerated: Optimized for CUDA-enabled GPUs Fast Generation: 30-60 seconds per image Memory Efficient: Smart GPU memory management Reproducible Results: Seed control for consistent outputs

๐Ÿ“‹ Quick Start Prerequisites

Python 3.8 or higher CUDA-capable GPU (12GB+ VRAM recommended) 15GB free disk space

Installation bash# Clone the repository git clone https://github.com/yourusername/flux-kontext-style-lora.git cd flux-kontext-style-lora

Install dependencies

pip install -r requirements.txt

Set HuggingFace token (optional, for private model access)

export HF_TOKEN="your_huggingface_token"

Run the application

python app.py ๐ŸŽฏ How to Use

Upload Image: Drag and drop or click to upload your image Select Style: Click any thumbnail from the style gallery Customize (Optional): Add additional instructions like "make it more vibrant" Generate: Click "Transform Image" and wait for the magic Download: Save your transformed artwork

Advanced Controls

Inference Steps: 10-50 (higher = better quality, slower) Guidance Scale: 1.0-5.0 (controls style strength) Seed: For reproducible results

๐ŸŒŸ Use Cases

Social Media Content: Create unique, eye-catching posts NFT Art Creation: Generate distinctive digital collectibles Design Prototyping: Quickly explore different visual styles Educational Projects: Demonstrate AI capabilities in art Personal Projects: Transform photos into artistic memories

๐Ÿ“„ License This project follows the FLUX.1-Kontext-dev model license. Please review the license terms before commercial use.

๐Ÿ™ Acknowledgments Black Forest Labs for the FLUX.1-Kontext-dev model Owen777 for the high-quality style LoRAs The open-source AI community

Community: https://discord.gg/openfreeai


๐ŸŽจ Flux Kontext Style LoRA - AI ์ด๋ฏธ์ง€ ์Šคํƒ€์ผ ๋ณ€ํ™˜

๋งํฌ: https://huggingface.co/spaces/ginigen/Flux-Kontext-Style

ํ”„๋กœ์ ํŠธ ์†Œ๊ฐœ Flux Kontext Style LoRA๋Š” FLUX.1-Kontext-dev ๋ชจ๋ธ๊ณผ 22๊ฐ€์ง€ ๊ณ ํ’ˆ์งˆ ์Šคํƒ€์ผ LoRA๋ฅผ ํ™œ์šฉํ•˜์—ฌ ์ด๋ฏธ์ง€๋ฅผ ๋‹ค์–‘ํ•œ ์˜ˆ์ˆ ์  ์Šคํƒ€์ผ๋กœ ๋ณ€ํ™˜ํ•˜๋Š” AI ๊ธฐ๋ฐ˜ ์›น ์• ํ”Œ๋ฆฌ์ผ€์ด์…˜์ž…๋‹ˆ๋‹ค.

์ฃผ์š” ๊ธฐ๋Šฅ โœจ 22๊ฐ€์ง€ ๋‹ค์–‘ํ•œ ์•„ํŠธ ์Šคํƒ€์ผ

์• ๋‹ˆ๋ฉ”์ด์…˜/๋งŒํ™”: ์ง€๋ธŒ๋ฆฌ, ์•„๋ฉ”๋ฆฌ์นธ ์นดํˆฐ, ์กฐ์กฐ, ์Šค๋ˆ„ํ”ผ, ๋ฆญ์•ค๋ชจํ‹ฐ 3D/๊ธฐํ•˜ํ•™์ : 3D ์น˜๋น„, ํด๋ฆฌ, ๋ ˆ๊ณ , ํด๋ ˆ์ด ํ† ์ด ์ „ํ†ต ์˜ˆ์ˆ : ์ˆ˜๋ฌตํ™”, ์œ ํ™”, ๋ฐ˜ ๊ณ ํ, ํ”ผ์นด์†Œ, ํŒ์•„ํŠธ ๊ณต์˜ˆ/์žฌ๋ฃŒ: ํŒจ๋ธŒ๋ฆญ, ์ข…์ด์ ‘๊ธฐ, ํŽ˜์ดํผ ์ปคํŒ…, ๋งˆ์นด๋กฑ ๋””์ง€ํ„ธ/๋ชจ๋˜: ํ”ฝ์…€, ๋ผ์ธ์•„ํŠธ, ๋ฒกํ„ฐ

๐Ÿš€ ํ•ต์‹ฌ ํŠน์ง•

์›ํด๋ฆญ ์ธ๋„ค์ผ ์Šคํƒ€์ผ ์„ ํƒ 1024x1024 ๊ณ ํ•ด์ƒ๋„ ์ถœ๋ ฅ ์ถ”๊ฐ€ ํ”„๋กฌํ”„ํŠธ๋กœ ์„ธ๋ถ€ ์กฐ์ • ๊ฐ€๋Šฅ GPU ๊ฐ€์† ์‹ค์‹œ๊ฐ„ ์ฒ˜๋ฆฌ ์ง๊ด€์ ์ธ UI/UX ๋””์ž์ธ

๊ธฐ์ˆ  ์Šคํƒ

๋ชจ๋ธ: FLUX.1-Kontext-dev LoRA: Owen777/Kontext-Style-Loras ํ”„๋ ˆ์ž„์›Œํฌ: Gradio, Diffusers ๋Ÿฐํƒ€์ž„: PyTorch, CUDA

์‚ฌ์šฉ ๋ฐฉ๋ฒ•

์ด๋ฏธ์ง€ ์—…๋กœ๋“œ ๋˜๋Š” ๊ธฐ๋ณธ ์ด๋ฏธ์ง€ ์‚ฌ์šฉ ์ธ๋„ค์ผ ํด๋ฆญ์œผ๋กœ ์›ํ•˜๋Š” ์Šคํƒ€์ผ ์„ ํƒ (์„ ํƒ) ์ถ”๊ฐ€ ์ง€์‹œ์‚ฌํ•ญ ์ž…๋ ฅ "Transform Image" ๋ฒ„ํŠผ ํด๋ฆญ 30-60์ดˆ ๋‚ด ์Šคํƒ€์ผ ๋ณ€ํ™˜ ์™„๋ฃŒ

์„ค์น˜ ๋ฐ ์‹คํ–‰ bash# ํ™˜๊ฒฝ ์„ค์ • pip install -r requirements.txt

HuggingFace ํ† ํฐ ์„ค์ • (์„ ํƒ)

export HF_TOKEN="your_token_here"

์‹คํ–‰

python app.py ์‹œ์Šคํ…œ ์š”๊ตฌ์‚ฌํ•ญ

Python 3.8+ CUDA ์ง€์› GPU (์ตœ์†Œ 12GB VRAM) 15GB+ ๋””์Šคํฌ ๊ณต๊ฐ„ (๋ชจ๋ธ ๋‹ค์šด๋กœ๋“œ)

ํ™œ์šฉ ๋ถ„์•ผ

์†Œ์…œ ๋ฏธ๋””์–ด ์ฝ˜ํ…์ธ  ์ œ์ž‘ NFT ์•„ํŠธ ์ƒ์„ฑ ๋””์ž์ธ ํ”„๋กœํ† ํƒ€์ดํ•‘ ๊ต์œก ๋ฐ ์—ฐ๊ตฌ ๊ฐœ์ธ ์ฐฝ์ž‘ ํ™œ๋™

๋ผ์ด์„ ์Šค ๋ณธ ํ”„๋กœ์ ํŠธ๋Š” FLUX.1-Kontext-dev ๋ชจ๋ธ ๋ผ์ด์„ ์Šค๋ฅผ ๋”ฐ๋ฆ…๋‹ˆ๋‹ค.