Flux Super Portrait Lo R A

Introduction

Flux-Super-Portrait-LoRA is a text-to-image diffusion model designed to generate high-quality, detailed portraits. It utilizes LoRA (Low-Rank Adaptation) techniques to enhance the text-to-image generation capabilities, making it suitable for creating diverse and vivid portrait images.

Architecture

The model is built on the base model "black-forest-labs/FLUX.1-dev" and leverages diffusion techniques to transform text prompts into images. It uses LoRA for fine-tuning, allowing for efficient adaptation and improved performance in generating portrait images.

Training

The model was trained using the AdamW optimizer with a constant learning rate scheduler. It incorporates noise handling parameters such as a noise offset of 0.03 and multiresolution noise discount of 0.1. The training involved 19 images at a resolution of 4K, repeated over 14 epochs with steps set to 2650 per epoch.

Guide: Running Locally

1. Setup Environment:

   • Install PyTorch and Hugging Face's transformers library.
   • Ensure you have access to a GPU for optimal performance.

2. Load the Model:

   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-Super-Portrait-LoRA"
   trigger_word = "Super Portrait"  
   pipe.load_lora_weights(lora_repo)
   
   device = torch.device("cuda")
   pipe.to(device)
   

3. Generate Images:

   • Use the trigger word "Super Portrait" to produce images based on your text descriptions.

4. Cloud GPUs:

   • Consider using cloud services like AWS, Google Cloud, or Azure for access to high-performance GPUs.

License

The model is licensed under the CreativeML Open RAIL-M license, which allows for open and collaborative use, with certain restrictions and responsibilities outlined in the license agreement.