Flux-Animex-v2-LoRA Open Source Model - Free Deployment for One-Click Generation of Anime-Style Images

Flux Animex V2 LoRA

Developed by strangerzonehf

LoRA fine-tuned version based on the FLUX.1-dev foundation model, specializing in anime-style image generation

Image Generation Open Source License:Openrail #Anime-style generation #LoRA fine-tuning #High-resolution images

Downloads 1,526

Release Time : 11/15/2024

Model Overview

This model employs LoRA technology to fine-tune the FLUX.1-dev diffusion model, specifically designed for generating high-quality anime-style images. It supports creating cartoon character images with specific styles through text descriptions

Model Features

Animex style generation

Specially optimized for generating anime-style images with distinctive characteristics

LoRA fine-tuning technology

Utilizes Low-Rank Adaptation (LoRA) technology for efficient fine-tuning of the base diffusion model

Detail control

Supports controlling facial features, clothing, backgrounds, and other details through detailed text descriptions

Model Capabilities

Text-to-anime-style image generation

Character feature detail control

Background style customization

Use Cases

Anime creation

Character design

Generate anime character images with specific features based on text descriptions

Can produce complete character images including details like hairstyle, eye color, and clothing

Scene construction

Create anime scenes with specific atmospheres and styles

Can generate scene images containing elements like background colors and lighting effects

🚀 Flux-Animex-v2-LoRA

A LoRA model for text-to-image generation, specializing in anime-style images.

🚀 Quick Start

Installation

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/Flux-Animex-v2-LoRA"
trigger_word = "Animex"  
pipe.load_lora_weights(lora_repo)

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

Usage

You should use Animex to trigger the image generation.

✨ Features

Text-to-Image: Generate anime-style images based on text prompts.
LoRA Integration: Works in conjunction with the base model black-forest-labs/FLUX.1-dev.

📦 Installation

The weights for this model are available in Safetensors format. You can download them in the Files & versions tab.

💻 Usage Examples

Basic Usage

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/Flux-Animex-v2-LoRA"
trigger_word = "Animex"  
pipe.load_lora_weights(lora_repo)

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

# Generate an image
prompt = f"{trigger_word}, a vibrant cartoon drawing of a mans face is depicted. The mans eyes are a piercing blue, and he has a serious expression on his face. His hair is spiky, and his bangs are a darker shade of purple. His eyebrows are a lighter shade of brown, while his mouth is slightly open. He is wearing a red hoodie, and a black jacket with a white stripe down the center of his chest. The hoodie is draped over his shoulders, adding a pop of color to the otherwise monochromatic image. The background is a deep, vibrant purple, and there are a few wispy red lines running across the image, adding depth to the composition."
image = pipe(prompt).images[0]
image.save("generated_image.png")

📚 Documentation

Model description

Image Processing Parameters

Parameter	Value
LR Scheduler	constant
Noise Offset	0.03
Optimizer	AdamW
Multires Noise Discount	0.1
Network Dim	64
Network Alpha	32
Epoch	20
Save Every N Epochs	1
Multires Noise Iterations	10
Repeat & Steps	35 & 4800

Labeling: florence2-en(natural language & English)

Total Images Used for Training: 57 [ 11@5 ] [ Hi -Res ]

Best Dimensions

768 x 1024 (Best)
1024 x 1024 (Default)

📄 License

This model is licensed under the creativeml-openrail-m license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご