README.md

---
license: mit
pipeline_tag: text-to-image
tags:
  - onnx
  - text-to-image
inference: false
---

## Model Descriptions:

This repo contains ONNX model files for [madebyollin's Tiny AutoEncoder for Stable Diffusion](https://huggingface.co/madebyollin/taesd).

## Using in 🧨 diffusers

To install the requirements for this demo, do pip install optimum["onnxruntime"]. 

```python
from optimum.onnxruntime import ORTStableDiffusionPipeline

model_id = "CompVis/stable-diffusion-v1-4"
pipeline = ORTStableDiffusionPipeline.from_pretrained(model_id, revision="onnx")

# Inject TAESD

taesd_id = "deinferno/taesd-onnx"
pipeline.vae_decoder = OnnxRuntimeModel.from_pretrained(taesd_id, subfolder = "vae_decoder")
pipeline.vae_encoder = OnnxRuntimeModel.from_pretrained(taesd_id, subfolder = "vae_encoder")

prompt = "sailing ship in storm by Leonardo da Vinci"
image = pipeline(prompt).images[0]
output.save("result.png")
```