File size: 2,330 Bytes
ec6ef19
105640e
5f10bb6
a9b3bf8
15735b5
a9b3bf8
5c762ce
327e673
 
105640e
 
 
f99066d
83803eb
 
803e17b
3e7e6b8
 
105640e
 
803e17b
 
3e7e6b8
 
105640e
 
3e7e6b8
105640e
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
This demo showcases a lightweight Stable Diffusion model (SDM) for general-purpose text-to-image synthesis. Our model [**BK-SDM-Small**](https://huggingface.co/nota-ai/bk-sdm-small) achieves **36% reduced** parameters and latency. This model is bulit with (i) removing several residual and attention blocks from the U-Net of [SDM-v1.4](https://huggingface.co/CompVis/stable-diffusion-v1-4) and (ii) distillation pretraining on only 0.22M LAION pairs (fewer than 0.1% of the full training set). Despite very limited training resources, our model can imitate the original SDM by benefiting from transferred knowledge.
- **For more information & acknowledgments**, please see [Paper](https://arxiv.org/abs/2305.15798), [GitHub](https://github.com/Nota-NetsPresso/BK-SDM), BK-SDM-{[Base](https://huggingface.co/nota-ai/bk-sdm-base), [Small](https://huggingface.co/nota-ai/bk-sdm-small), [Tiny](https://huggingface.co/nota-ai/bk-sdm-tiny)} Model Card.
  
<center>
    <img alt="U-Net architectures and KD-based pretraining" img src="https://huggingface.co/spaces/nota-ai/compressed-stable-diffusion/resolve/91f349ab3b900cbfec20163edd6a312d1e8c8193/docs/fig_model.png" width="65%">
</center>

<br/>

- This research was accepted to [**ICCV 2023 Demo Track**](https://iccv2023.thecvf.com/) & [**ICML 2023 Workshop on Efficient Systems for Foundation Models** (ES-FoMo)](https://es-fomo.com/).
- Please be aware that your prompts are logged, _without_ any personally identifiable information.
- For different images with the same prompt, please change _Random Seed_ in Advanced Settings (because of using the firstly sampled latent code per seed).

<br/>

**Demo Environment**: [Aug/01/2023] NVIDIA T4-small (4 vCPU · 15 GB RAM · 16GB VRAM) — 5~10 sec inference of the original SDM (for a 512×512 image with 25 denoising steps)
<br/>

<details>
<summary>Previous Env Setup:</summary>
[July/31/2023] Free CPU-basic (2 vCPU · 16 GB RAM) — 7~10 min slow inference of the original SDM.
<br/>
[July/27/2023] NVIDIA T4-small (4 vCPU · 15 GB RAM · 16GB VRAM) — 5~10 sec inference of the original SDM.
<br/>
[June/30/2023] Free CPU-basic (2 vCPU · 16 GB RAM) — 7~10 min slow inference of the original SDM.
<br/>
[May/31/2023] NVIDIA T4-small (4 vCPU · 15 GB RAM · 16GB VRAM) — 5~10 sec inference of the original SDM.
</details>