File size: 2,088 Bytes
87f247b 6de1dec 862cb55 87f247b 12210fd 87f247b b5398cf 00c4c5d d1ca045 87f247b 2d4de60 504b925 2d4de60 87f247b 2d4de60 504b925 2d4de60 b5398cf c8f7c0c d1e46c0 2d4de60 b5398cf 7427aae affb45e b5398cf affb45e |
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 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 |
---
license: cc-by-nc-4.0
library_name: diffusers
base_model: runwayml/stable-diffusion-v1-5
tags:
- lora
- text-to-image
- diffusers
- stable-diffusion
inference: False
---
# ⚡ Flash Diffusion: FlashSD ⚡
Flash Diffusion is a diffusion distillation method proposed in [ADD ARXIV]() *by Clément Chadebec, Onur Tasar, Eyal Benaroche, and Benjamin Aubin.*
This model is a **26.4M** LoRA distilled version of SD1.5 model that is able to generate images in **2-4 steps**. The main purpose of this model is to reproduce the main results of the paper.
See our [live demo](https://huggingface.co/spaces/jasperai/FlashPixart).
<p align="center">
<img style="width:700px;" src="images/flash_sd.jpg">
</p>
# How to use?
The model can be used using the `StableDiffusionPipeline` from `diffusers` library directly. It can allow reducing the number of required sampling steps to **2-4 steps**.
```python
from diffusers import StableDiffusionPipeline, LCMScheduler
adapter_id = "jasperai/flash-sd"
pipe = StableDiffusionPipeline.from_pretrained(
"runwayml/stable-diffusion-v1-5",
use_safetensors=True,
)
pipe.scheduler = LCMScheduler.from_pretrained(
"runwayml/stable-diffusion-v1-5",
subfolder="scheduler",
timestep_spacing="trailing",
)
pipe.to("cuda")
# Fuse and load LoRA weights
pipe.load_lora_weights(adapter_id)
pipe.fuse_lora()
prompt = "A raccoon reading a book in a lush forest."
image = pipe(prompt, num_inference_steps=4, guidance_scale=0).images[0]
```
<p align="center">
<img style="width:400px;" src="images/raccoon.png">
</p>
# Training Details
The model was trained for 20k iterations on 2 H100 GPUs (representing approx. a total **26 GPU hours** of training). Please refer to the [paper]() for further parameters details.
**Metrics on COCO 2017 validation set (Table 1)**
- FID-5k: 22.6 (2 NFE) / 22.5 (4 NFE)
- CLIP Score (ViT-g/14): 0.306 (2 NFE) / 0.311 (4 NFE)
**Metrics on COCO 2014 validation (Table 2)**
- FID-30k: 12.41 (4 NFE)
- FID-30k: 12.27 (2 NFE)
## License
This model is released under the the Creative Commons BY-NC license. |