File size: 1,249 Bytes
5a11670 8f79696 42505ec 8f79696 73b7d1c b7456fc 73b7d1c bf2a139 21efc66 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 |
---
license: mit
---
# 🍰 Tiny AutoEncoder for Stable Diffusion
[TAESD](https://github.com/madebyollin/taesd) is very tiny autoencoder which uses the same "latent API" as Stable Diffusion's VAE.
TAESD is useful for [real-time previewing](https://twitter.com/madebyollin/status/1679356448655163394) of the SD generation process.
Comparison on my laptop:
This repo contains `.safetensors` versions of the TAESD weights.
For SDXL, use [TAESDXL](https://huggingface.co/madebyollin/taesdxl/) instead (the SD and SDXL VAEs are [incompatible](https://huggingface.co/madebyollin/sdxl-vae-fp16-fix/discussions/6#64b8a9c13707b7d603c6ac16)).
## Using in 🧨 diffusers
```python
import torch
from diffusers import DiffusionPipeline, AutoencoderTiny
pipe = DiffusionPipeline.from_pretrained(
"stabilityai/stable-diffusion-2-1-base", torch_dtype=torch.float16
)
pipe.vae = AutoencoderTiny.from_pretrained("madebyollin/taesd", torch_dtype=torch.float16)
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")
``` |