elden ring diffusion

Introduction

Elden Ring Diffusion is a fine-tuned Stable Diffusion model trained on game art from Elden Ring. The model allows users to generate images in the Elden Ring style by including the phrase "elden ring style" in prompts. It supports various export formats and can be used with the Stable Diffusion pipeline.

Architecture

The model is based on the Stable Diffusion architecture and utilizes the Diffusers library for implementation. It can be exported to ONNX, MPS, and FLAX/JAX for optimized performance across different platforms.

Training

Elden Ring Diffusion was trained using the diffusers-based Dreambooth training method with prior-preservation loss over 3,000 steps. It has been optimized with specific prompts and settings for generating portraits and landscapes in the Elden Ring style.

Guide: Running Locally

1. Install Required Libraries:

   pip install diffusers transformers scipy torch
   

2. Load and Run the Model:

   from diffusers import StableDiffusionPipeline
   import torch
   
   model_id = "nitrosocke/elden-ring-diffusion"
   pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
   pipe = pipe.to("cuda")
   
   prompt = "a magical princess with golden hair, elden ring style"
   image = pipe(prompt).images[0]
   image.save("./magical_princess.png")
   

3. Hardware Recommendations:

   • It is recommended to use a cloud GPU service such as AWS, Google Cloud, or Azure to handle the computational requirements efficiently.

License

The model is distributed under the CreativeML OpenRAIL-M license. Key points include:

1. Prohibition against using the model to produce illegal or harmful content.
2. No rights are claimed on generated outputs, but users are accountable for their use.
3. Redistribution of weights and commercial use are permitted, provided that the same license restrictions are maintained and shared with users.

For full license details, refer to the CreativeML OpenRAIL license.