216 lines
8.5 KiB
Markdown
216 lines
8.5 KiB
Markdown
---
|
|
license: openrail++
|
|
tags:
|
|
- text-to-image
|
|
- stable-diffusion
|
|
---
|
|
# SD-XL 1.0-base Model Card
|
|
|
|
|
|
## Model
|
|
|
|
|
|
|
|
[SDXL](https://arxiv.org/abs/2307.01952) consists of an [ensemble of experts](https://arxiv.org/abs/2211.01324) pipeline for latent diffusion:
|
|
In a first step, the base model is used to generate (noisy) latents,
|
|
which are then further processed with a refinement model (available here: https://huggingface.co/stabilityai/stable-diffusion-xl-refiner-1.0/) specialized for the final denoising steps.
|
|
Note that the base model can be used as a standalone module.
|
|
|
|
Alternatively, we can use a two-stage pipeline as follows:
|
|
First, the base model is used to generate latents of the desired output size.
|
|
In the second step, we use a specialized high-resolution model and apply a technique called SDEdit (https://arxiv.org/abs/2108.01073, also known as "img2img")
|
|
to the latents generated in the first step, using the same prompt. This technique is slightly slower than the first one, as it requires more function evaluations.
|
|
|
|
Source code is available at https://github.com/Stability-AI/generative-models .
|
|
|
|
### Model Description
|
|
|
|
- **Developed by:** Stability AI
|
|
- **Model type:** Diffusion-based text-to-image generative model
|
|
- **License:** [CreativeML Open RAIL++-M License](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/LICENSE.md)
|
|
- **Model Description:** This is a model that can be used to generate and modify images based on text prompts. It is a [Latent Diffusion Model](https://arxiv.org/abs/2112.10752) that uses two fixed, pretrained text encoders ([OpenCLIP-ViT/G](https://github.com/mlfoundations/open_clip) and [CLIP-ViT/L](https://github.com/openai/CLIP/tree/main)).
|
|
- **Resources for more information:** Check out our [GitHub Repository](https://github.com/Stability-AI/generative-models) and the [SDXL report on arXiv](https://arxiv.org/abs/2307.01952).
|
|
|
|
### Model Sources
|
|
|
|
For research purposes, we recommend our `generative-models` Github repository (https://github.com/Stability-AI/generative-models), which implements the most popular diffusion frameworks (both training and inference) and for which new functionalities like distillation will be added over time.
|
|
[Clipdrop](https://clipdrop.co/stable-diffusion) provides free SDXL inference.
|
|
|
|
- **Repository:** https://github.com/Stability-AI/generative-models
|
|
- **Demo:** https://clipdrop.co/stable-diffusion
|
|
|
|
|
|
## Evaluation
|
|
|
|
The chart above evaluates user preference for SDXL (with and without refinement) over SDXL 0.9 and Stable Diffusion 1.5 and 2.1.
|
|
The SDXL base model performs significantly better than the previous variants, and the model combined with the refinement module achieves the best overall performance.
|
|
|
|
|
|
### 🧨 Diffusers
|
|
|
|
Make sure to upgrade diffusers to >= 0.19.0:
|
|
```
|
|
pip install diffusers --upgrade
|
|
```
|
|
|
|
In addition make sure to install `transformers`, `safetensors`, `accelerate` as well as the invisible watermark:
|
|
```
|
|
pip install invisible_watermark transformers accelerate safetensors
|
|
```
|
|
|
|
To just use the base model, you can run:
|
|
|
|
```py
|
|
from diffusers import DiffusionPipeline
|
|
import torch
|
|
|
|
pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16, use_safetensors=True, variant="fp16")
|
|
pipe.to("cuda")
|
|
|
|
# if using torch < 2.0
|
|
# pipe.enable_xformers_memory_efficient_attention()
|
|
|
|
prompt = "An astronaut riding a green horse"
|
|
|
|
images = pipe(prompt=prompt).images[0]
|
|
```
|
|
|
|
To use the whole base + refiner pipeline as an ensemble of experts you can run:
|
|
|
|
```py
|
|
from diffusers import DiffusionPipeline
|
|
import torch
|
|
|
|
# load both base & refiner
|
|
base = DiffusionPipeline.from_pretrained(
|
|
"stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16, variant="fp16", use_safetensors=True
|
|
)
|
|
base.to("cuda")
|
|
refiner = DiffusionPipeline.from_pretrained(
|
|
"stabilityai/stable-diffusion-xl-refiner-1.0",
|
|
text_encoder_2=base.text_encoder_2,
|
|
vae=base.vae,
|
|
torch_dtype=torch.float16,
|
|
use_safetensors=True,
|
|
variant="fp16",
|
|
)
|
|
refiner.to("cuda")
|
|
|
|
# Define how many steps and what % of steps to be run on each experts (80/20) here
|
|
n_steps = 40
|
|
high_noise_frac = 0.8
|
|
|
|
prompt = "A majestic lion jumping from a big stone at night"
|
|
|
|
# run both experts
|
|
image = base(
|
|
prompt=prompt,
|
|
num_inference_steps=n_steps,
|
|
denoising_end=high_noise_frac,
|
|
output_type="latent",
|
|
).images
|
|
image = refiner(
|
|
prompt=prompt,
|
|
num_inference_steps=n_steps,
|
|
denoising_start=high_noise_frac,
|
|
image=image,
|
|
).images[0]
|
|
```
|
|
|
|
When using `torch >= 2.0`, you can improve the inference speed by 20-30% with torch.compile. Simple wrap the unet with torch compile before running the pipeline:
|
|
```py
|
|
pipe.unet = torch.compile(pipe.unet, mode="reduce-overhead", fullgraph=True)
|
|
```
|
|
|
|
If you are limited by GPU VRAM, you can enable *cpu offloading* by calling `pipe.enable_model_cpu_offload`
|
|
instead of `.to("cuda")`:
|
|
|
|
```diff
|
|
- pipe.to("cuda")
|
|
+ pipe.enable_model_cpu_offload()
|
|
```
|
|
|
|
For more information on how to use Stable Diffusion XL with `diffusers`, please have a look at [the Stable Diffusion XL Docs](https://huggingface.co/docs/diffusers/api/pipelines/stable_diffusion/stable_diffusion_xl).
|
|
|
|
### Optimum
|
|
[Optimum](https://github.com/huggingface/optimum) provides a Stable Diffusion pipeline compatible with both [OpenVINO](https://docs.openvino.ai/latest/index.html) and [ONNX Runtime](https://onnxruntime.ai/).
|
|
|
|
#### OpenVINO
|
|
|
|
To install Optimum with the dependencies required for OpenVINO :
|
|
|
|
```bash
|
|
pip install optimum[openvino]
|
|
```
|
|
|
|
To load an OpenVINO model and run inference with OpenVINO Runtime, you need to replace `StableDiffusionXLPipeline` with Optimum `OVStableDiffusionXLPipeline`. In case you want to load a PyTorch model and convert it to the OpenVINO format on-the-fly, you can set `export=True`.
|
|
|
|
```diff
|
|
- from diffusers import StableDiffusionXLPipeline
|
|
+ from optimum.intel import OVStableDiffusionXLPipeline
|
|
|
|
model_id = "stabilityai/stable-diffusion-xl-base-1.0"
|
|
- pipeline = StableDiffusionXLPipeline.from_pretrained(model_id)
|
|
+ pipeline = OVStableDiffusionXLPipeline.from_pretrained(model_id)
|
|
prompt = "A majestic lion jumping from a big stone at night"
|
|
image = pipeline(prompt).images[0]
|
|
```
|
|
|
|
You can find more examples (such as static reshaping and model compilation) in optimum [documentation](https://huggingface.co/docs/optimum/main/en/intel/inference#stable-diffusion-xl).
|
|
|
|
|
|
#### ONNX
|
|
|
|
To install Optimum with the dependencies required for ONNX Runtime inference :
|
|
|
|
```bash
|
|
pip install optimum[onnxruntime]
|
|
```
|
|
|
|
To load an ONNX model and run inference with ONNX Runtime, you need to replace `StableDiffusionXLPipeline` with Optimum `ORTStableDiffusionXLPipeline`. In case you want to load a PyTorch model and convert it to the ONNX format on-the-fly, you can set `export=True`.
|
|
|
|
```diff
|
|
- from diffusers import StableDiffusionXLPipeline
|
|
+ from optimum.onnxruntime import ORTStableDiffusionXLPipeline
|
|
|
|
model_id = "stabilityai/stable-diffusion-xl-base-1.0"
|
|
- pipeline = StableDiffusionXLPipeline.from_pretrained(model_id)
|
|
+ pipeline = ORTStableDiffusionXLPipeline.from_pretrained(model_id)
|
|
prompt = "A majestic lion jumping from a big stone at night"
|
|
image = pipeline(prompt).images[0]
|
|
```
|
|
|
|
You can find more examples in optimum [documentation](https://huggingface.co/docs/optimum/main/en/onnxruntime/usage_guides/models#stable-diffusion-xl).
|
|
|
|
|
|
## Uses
|
|
|
|
### Direct Use
|
|
|
|
The model is intended for research purposes only. Possible research areas and tasks include
|
|
|
|
- Generation of artworks and use in design and other artistic processes.
|
|
- Applications in educational or creative tools.
|
|
- Research on generative models.
|
|
- Safe deployment of models which have the potential to generate harmful content.
|
|
- Probing and understanding the limitations and biases of generative models.
|
|
|
|
Excluded uses are described below.
|
|
|
|
### Out-of-Scope Use
|
|
|
|
The model was not trained to be factual or true representations of people or events, and therefore using the model to generate such content is out-of-scope for the abilities of this model.
|
|
|
|
## Limitations and Bias
|
|
|
|
### Limitations
|
|
|
|
- The model does not achieve perfect photorealism
|
|
- The model cannot render legible text
|
|
- The model struggles with more difficult tasks which involve compositionality, such as rendering an image corresponding to “A red cube on top of a blue sphere”
|
|
- Faces and people in general may not be generated properly.
|
|
- The autoencoding part of the model is lossy.
|
|
|
|
### Bias
|
|
While the capabilities of image generation models are impressive, they can also reinforce or exacerbate social biases.
|