# Stable Diffusion

Stable Diffusion is a text-to-image latent diffusion model.
Check out this [blog post](https://huggingface.co/blog/stable_diffusion) for more information.

## How to generate images?

To generate images with Stable Diffusion on Gaudi, you need to instantiate two instances:
- A pipeline with `GaudiStableDiffusionPipeline`. This pipeline supports *text-to-image generation*.
- A scheduler with `GaudiDDIMScheduler`. This scheduler has been optimized for Gaudi.

When initializing the pipeline, you have to specify `use_habana=True` to deploy it on HPUs.
Furthermore, to get the fastest possible generations you should enable **HPU graphs** with `use_hpu_graphs=True`.
Finally, you will need to specify a [Gaudi configuration](../package_reference/gaudi_config) which can be downloaded from the Hugging Face Hub.

```python
from optimum.habana.diffusers import GaudiDDIMScheduler, GaudiStableDiffusionPipeline

model_name = "CompVis/stable-diffusion-v1-4"

scheduler = GaudiDDIMScheduler.from_pretrained(model_name, subfolder="scheduler")

pipeline = GaudiStableDiffusionPipeline.from_pretrained(
    model_name,
    scheduler=scheduler,
    use_habana=True,
    use_hpu_graphs=True,
    gaudi_config="Habana/stable-diffusion",
)
```

You can then call the pipeline to generate images from one or several prompts:
```python
outputs = pipeline(
    prompt=["High quality photo of an astronaut riding a horse in space", "Face of a yellow cat, high resolution, sitting on a park bench"],
    num_images_per_prompt=10,
    batch_size=4,
    output_type="pil",
)
```

Generated images can be returned as either PIL images or NumPy arrays, depending on the `output_type` option.

Check out the [example](/examples/stable-diffusion) provided in the official Github repository.

## Stable Diffusion 2

DISCLAIMER: Stable Diffusion 2 models family has been discontinued and withdrawn by Stability AI. The following instruction uses mirrored models maintained by sd2-community, not maintained by Stability AI.

[Stable Diffusion 2](https://huggingface.co/docs/diffusers/api/pipelines/stable_diffusion/stable_diffusion_2) can be used with the exact same classes.
Here is an example:

```python
from optimum.habana.diffusers import GaudiDDIMScheduler, GaudiStableDiffusionPipeline

model_name = "sd2-community/stable-diffusion-2-1"

scheduler = GaudiDDIMScheduler.from_pretrained(model_name, subfolder="scheduler")

pipeline = GaudiStableDiffusionPipeline.from_pretrained(
    model_name,
    scheduler=scheduler,
    use_habana=True,
    use_hpu_graphs=True,
    gaudi_config="Habana/stable-diffusion-2",
)

outputs = pipeline(
    ["An image of a squirrel in Picasso style"],
    num_images_per_prompt=10,
    batch_size=2,
    height=768,
    width=768,
)
```

There are two different checkpoints for Stable Diffusion 2:

- use [sd2-community/stable-diffusion-2-1](https://huggingface.co/sd2-community/stable-diffusion-2-1) for generating 768x768 images
- use [sd2-community/stable-diffusion-2-1-base](https://huggingface.co/sd2-community/stable-diffusion-2-1-base) for generating 512x512 images

## Super-resolution

The Stable Diffusion upscaler diffusion model was created by the researchers and engineers from CompVis, Stability AI, and LAION. It is used to enhance the resolution of input images by a factor of 4.

See [here](https://huggingface.co/docs/diffusers/api/pipelines/stable_diffusion/upscale) for more information.

### How to upscale low resolution images?

To generate RGB and depth images with Stable Diffusion Upscale on Gaudi, you need to instantiate two instances:
- A pipeline with `GaudiStableDiffusionUpscalePipeline`.
- A scheduler with `GaudiDDIMScheduler`. This scheduler has been optimized for Gaudi.

When initializing the pipeline, you have to specify `use_habana=True` to deploy it on HPUs.
Furthermore, to get the fastest possible generations you should enable **HPU graphs** with `use_hpu_graphs=True`.
Finally, you will need to specify a [Gaudi configuration](../package_reference/gaudi_config) which can be downloaded from the Hugging Face Hub.

```python
import requests
from io import BytesIO
from optimum.habana.diffusers import (
    GaudiDDIMScheduler,
    GaudiStableDiffusionUpscalePipeline,
)
from optimum.habana.utils import set_seed
from PIL import Image

set_seed(42)

model_name_upscale = "stabilityai/stable-diffusion-x4-upscaler"
scheduler = GaudiDDIMScheduler.from_pretrained(model_name_upscale, subfolder="scheduler")
url = "https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd2-upscale/low_res_cat.png"
response = requests.get(url)
low_res_img = Image.open(BytesIO(response.content)).convert("RGB")
low_res_img = low_res_img.resize((128, 128))
low_res_img.save("low_res_cat.png")
prompt = "a white cat"

pipeline = GaudiStableDiffusionUpscalePipeline.from_pretrained(
    model_name_upscale,
    scheduler=scheduler,
    use_habana=True,
    use_hpu_graphs=True,
    gaudi_config="Habana/stable-diffusion",
)
upscaled_image = pipeline(prompt=prompt, image=low_res_img).images[0]
upscaled_image.save("upsampled_cat.png")

```

## Tips

To accelerate your Stable Diffusion pipeline, you can run it in full *bfloat16* precision.
This will also save memory.
You just need to pass `torch_dtype=torch.bfloat16` to `from_pretrained` when instantiating your pipeline.
Here is how to do it:

```python
import torch

pipeline = GaudiStableDiffusionPipeline.from_pretrained(
    "CompVis/stable-diffusion-v1-4",
    scheduler=scheduler,
    use_habana=True,
    use_hpu_graphs=True,
    gaudi_config="Habana/stable-diffusion",
    torch_dtype=torch.bfloat16
)
```

## Textual Inversion Fine-Tuning

[Textual Inversion](https://arxiv.org/abs/2208.01618) is a method to personalize text2image models like Stable Diffusion on your own images using just 3-5 examples.

You can find [here](https://github.com/huggingface/optimum-habana/blob/main/examples/stable-diffusion/textual_inversion.py) an example script that implements this training method.