classic anim diffusion

Introduction

The Classic Animation Diffusion model is a fine-tuned version of Stable Diffusion trained on screenshots from a popular animation studio. To achieve a classic animation style in your images, use the prompt tokens classic disney style. This model is capable of rendering various subjects like characters, animals, and landscapes with a distinctive animation style.

Architecture

This model operates similarly to other Stable Diffusion models and can be integrated into applications using the Diffusers library. It supports exporting to several formats, including ONNX, MPS, and FLAX/JAX, enhancing flexibility for deployment across different platforms.

Training

The model was trained using the Diffusers-based DreamBooth training methodology by ShivamShrirao. The training involved prior-preservation loss and the train-text-encoder flag over 9,000 steps, optimizing its ability to generate high-quality, stylized images.

Guide: Running Locally

To run the Classic Animation Diffusion model locally, follow these steps:

1. Install Dependencies: Ensure you have torch and diffusers installed in your Python environment.
2. Load the Model: Use the following code snippet to load and use the model.

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

3. Hardware: For optimal performance, using a cloud GPU is recommended. Platforms like AWS, Google Cloud, or Azure offer robust GPU instances suitable for this purpose.

License

The Classic Animation Diffusion model is distributed under the CreativeML OpenRAIL-M license, allowing open access with conditions:

1. The model cannot be used to produce or share illegal or harmful content.
2. Users have rights to the outputs they generate but must comply with the license terms.
3. Redistribution and commercial use are permitted, provided the same use restrictions are included, and the license is shared with all users. For detailed terms, please refer to the full license here.