🍰 Tiny AutoEncoder for Stable Diffusion 3

TAESD3 is very tiny autoencoder which uses the same "latent API" as Stable Diffusion 3's VAE. TAESD3 is useful for real-time previewing of the SD3 generation process.

This repo contains .safetensors versions of the TAESD3 weights.

Using in 🧨 diffusers

import torch
from diffusers import StableDiffusion3Pipeline, AutoencoderTiny

pipe = StableDiffusion3Pipeline.from_pretrained(
    "stabilityai/stable-diffusion-3-medium-diffusers", torch_dtype=torch.float16
)
pipe.vae = AutoencoderTiny.from_pretrained("madebyollin/taesd3", torch_dtype=torch.float16)
pipe.vae.config.shift_factor = 0.0
pipe = pipe.to("cuda")

prompt = "slice of delicious New York-style berry cheesecake"
image = pipe(prompt, num_inference_steps=25).images[0]
image.save("cheesecake.png")