flux handwriting

Introduction
The FLUX Handwriting model is a text-to-image model designed to generate handwriting images based on text prompts. It is a LoRA-based model developed by "fofr" and hosted on Hugging Face, leveraging the diffusers library for image generation tasks.

Architecture
The model builds upon the base model black-forest-labs/FLUX.1-dev and utilizes LoRA (Low-Rank Adaptation) technology to perform handwriting generation. It employs the diffusers library to handle text-to-image transformations effectively.

Training
FLUX Handwriting was trained on Replicate using the flux-dev-lora-trainer. This training approach allows the model to effectively capture and replicate various handwriting styles based on given textual prompts.

Guide: Running Locally

1. Install the diffusers library: Ensure you have the diffusers Python library installed.
2. Load the model: Use the provided Python code snippet to load the model and LoRA weights using the diffusers library.

   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('fofr/flux-handwriting', weight_name='lora.safetensors')
   image = pipeline('your prompt').images[0]
   

3. Trigger words: Use the trigger word HWRIT handwriting in prompts to generate handwriting images.
4. Hardware recommendation: For optimal performance, using a cloud GPU service like AWS EC2 with a CUDA-enabled GPU is recommended.

License
The model is licensed under the flux-1-dev-non-commercial-license. Further details can be found in the license documentation.