flux tarot v1

Introduction

The FLUX TAROT V1 is a text-to-image model designed to generate tarot-style images. It is built using the LoRA technique, trained on the 1920 Raider Waite tarot card public domain dataset. The model leverages the diffusers library for image generation.

Architecture

FLUX TAROT V1 utilizes the LoRA (Low-Rank Adaptation) technique to enhance the base model, black-forest-labs/FLUX.1-dev. The architecture is tailored to create images in a distinctive tarot style using specific trigger words.

Training

The model was trained using the fal-ai trainer and the ostris AI Toolkit. The training process employed the public domain dataset of the Raider Waite tarot cards from 1920, ensuring high-quality, style-consistent image generation.

Guide: Running Locally

1. Install Dependencies: Ensure you have Python installed and set up a virtual environment. Install the necessary libraries using pip:

   pip install torch diffusers
   

2. Load the Model: Use the following Python code to load and run the model with a suitable GPU:

   from diffusers import AutoPipelineForText2Image
   import torch
   
   pipeline = AutoPipelineForText2Image.from_pretrained('black-forest-labs/FLUX.1-dev', torch_dtype=torch.bfloat16).to('cuda')
   pipeline.load_lora_weights('multimodalart/flux-tarot-v1', weight_name='flux_tarot_v1_lora.safetensors')
   image = pipeline('in the style of TOK a trtcrd tarot style').images[0]
   

3. GPU Recommendation: For optimal performance, especially when handling high-resolution images, use a cloud GPU service such as AWS EC2 with a Tesla V100 or A100 instance.

License

The FLUX TAROT V1 model is distributed under the flux-1-dev-non-commercial-license. For more details, refer to the license document.