LUNA-NSFW Model Documentation

Introduction

LUNA-NSFW is a text-to-image model developed using FLUX and LoRA technologies. It is designed to be used with the Diffusers library for generating images from textual prompts. The model is particularly tailored to work with the keyword "LUNA" to trigger image generation.

Architecture

The model is based on the FLUX.1-dev architecture from Black-Forest-Labs and utilizes LoRA (Low-Rank Adaptation) for efficient training and deployment. It supports the integration of multiple technologies including diffusers and replicate.

Training

The model was trained on the Replicate platform using the FLUX-dev-LoRA-trainer. This setup allows for effective adaptation and fine-tuning of the model to handle specific image generation tasks.

Guide: Running Locally

To run LUNA-NSFW locally, follow these steps:

1. Install the necessary libraries:

   • Ensure you have Python installed and set up on your machine.
   • Install the Diffusers library using pip:

     pip install diffusers
     

2. Load the Model:

   • Use the following Python script to load the model and generate images:

     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('evstrahy/luna-NSFW', weight_name='lora.safetensors')
     image = pipeline('your prompt').images[0]
     

   • Replace 'your prompt' with the desired text to generate corresponding images.

3. Hardware Recommendations:

   • It is recommended to use a cloud GPU service like AWS, Google Cloud, or Azure for optimal performance, especially for large-scale or resource-intensive image generation tasks.

License

The LUNA-NSFW model is distributed under the FLUX-1-dev-non-commercial-license. For more details on the license terms, refer to the license document.