---
license: openrail++
tags:
- text-to-image
- stable-diffusion
library_name: diffusers
---

# SDXL-Lightning

![Intro Image](images/intro.jpg)

SDXL-Lightning is a lightning fast text-to-image generative model. It can generate high-quality 1024px images under a few steps. For more information, please refer to our paper: [SDXL-Lightning: Progressive Adversarial Diffusion Distillation]() (coming soon). The models are released for research purposes only.

Our models are distilled from [stabilityai/stable-diffusion-xl-base-1.0](https://huggingface.co./stabilityai/stable-diffusion-xl-base-1.0). This repository contains checkpoints for 1-step, 2-step, 4-step, and 8-step distilled models.

We provide both full UNet and LoRA checkpoints. The full UNet models have the best quality while the LoRA models can be applied to other base models.


## Diffusers Usage

Please always use the correct checkpoint for the corresponding inference steps.

### 2-Step, 4-Step, 8-Step UNet

```python
import torch
from diffusers import StableDiffusionXLPipeline, EulerDiscreteScheduler
from huggingface_hub import hf_hub_download

base = "stabilityai/stable-diffusion-xl-base-1.0"
repo = "ByteDance/sdxl-lightning"
ckpt = "sdxl_lightning_4step_unet.pth" # Use the correct ckpt for your step setting!

# Load model.
pipe = StableDiffusionXLPipeline.from_pretrained(base, torch_dtype=torch.float16, variant="fp16").to("cuda")
pipe.unet.load_state_dict(torch.load(hf_hub_download(repo, ckpt), map_location="cuda"))

# Ensure sampler uses "trailing" timesteps.
pipe.scheduler = EulerDiscreteScheduler.from_config(pipe.scheduler.config, timestep_spacing="trailing")

# Ensure using the same inference steps as the loaded model and CFG set to 0.
pipe("A girl smiling", num_inference_steps=4, guidance_scale=0).images[0].save("output.png")
```

### 2-Step, 4-Step, 8-Step LoRA

```python
import torch
from diffusers import StableDiffusionXLPipeline, EulerDiscreteScheduler
from huggingface_hub import hf_hub_download

base = "stabilityai/stable-diffusion-xl-base-1.0"
repo = "ByteDance/sdxl-lightning"
ckpt = "sdxl_lightning_4step_lora.pth" # Use the correct ckpt for your step setting!

# Load model.
pipe = StableDiffusionXLPipeline.from_pretrained(base, torch_dtype=torch.float16, variant="fp16").to("cuda")
pipe.load_lora_weights(hf_hub_download(repo, ckpt))
pipe.fuse_lora()

# Ensure sampler uses "trailing" timesteps.
pipe.scheduler = EulerDiscreteScheduler.from_config(pipe.scheduler.config, timestep_spacing="trailing")

# Ensure using the same inference steps as the loaded model and CFG set to 0.
pipe("A girl smiling", num_inference_steps=4, guidance_scale=0).images[0].save("output.png")
```

### 1-Step UNet

```python
import torch
from diffusers import StableDiffusionXLPipeline, EulerDiscreteScheduler
from huggingface_hub import hf_hub_download

base = "stabilityai/stable-diffusion-xl-base-1.0"
repo = "ByteDance/sdxl-lightning"
ckpt = "sdxl_lightning_1step_unet.pth" # Use the correct ckpt for your step setting!

# Load model.
pipe = StableDiffusionXLPipeline.from_pretrained(base, torch_dtype=torch.float16, variant="fp16").to("cuda")
pipe.unet.load_state_dict(torch.load(hf_hub_download(repo, ckpt), map_location="cuda"))

# Ensure sampler uses "trailing" timesteps and "sample" prediction type.
pipe.scheduler = EulerDiscreteScheduler.from_config(pipe.scheduler.config, timestep_spacing="trailing", prediction_type="sample")

# Ensure using the same inference steps as the loaded model and CFG set to 0.
pipe("A girl smiling", num_inference_steps=1, guidance_scale=0).images[0].save("output.png")
```


## ComfyUI Usage

Please always use the correct checkpoint for the corresponding inference steps.
Please use Euler sampler with sgm_uniform scheduler.

### 2-Step, 4-Step, 8-Step UNet

1. Download the UNet checkpoint to `/ComfyUI/models/unet`.
2. Download our [ComfyUI UNet workflow](comfyui/sdxl_lightning_unet.json).

![SDXL-Lightning ComfyUI UNet Workflow](images/comfyui_unet.png)

### 2-Step, 4-Step, 8-Step LoRA

1. Download the LoRA checkpoint to `/ComfyUI/models/loras`
2. Download our [ComfyUI LoRA workflow](comfyui/sdxl_lightning_lora.json).

![SDXL-Lightning ComfyUI UNet Workflow](images/comfyui_lora.png)

### 1-Step UNet

ComfyUI does not support changing model formulation to x0-prediction, so it is not usable in ComfyUI yet. Hopefully ComfyUI gets updated soon.