Commit
•
0b077d8
1
Parent(s):
0e13e08
Update readme
Browse files
README.md
CHANGED
@@ -13,16 +13,16 @@ tags:
|
|
13 |
# Stable Diffusion XL 1.0 TensorRT
|
14 |
|
15 |
### Introduction
|
16 |
-
This repository contains Stable Diffusion XL 1.0 ONNX models compatible with TensorRT.
|
17 |
|
18 |
-
|
|
|
19 |
|
20 |
![examples](./examples.jpg)
|
21 |
|
22 |
|
23 |
### Performance Comparison
|
24 |
|
25 |
-
Timings for 30 steps at 1024x1024
|
26 |
|
27 |
| Accelerator | Baseline (non-optimized) | NVIDIA TensorRT (optimized) | Percentage improvement |
|
28 |
|-------------|--------------------------|-----------------------------|------------------------|
|
@@ -30,7 +30,7 @@ Timings for 30 steps at 1024x1024
|
|
30 |
| A100 | 3704 ms | 2742 ms | ~26% |
|
31 |
| H100 | 2496 ms | 1471 ms | ~41% |
|
32 |
|
33 |
-
Image throughput for 30 steps
|
34 |
|
35 |
| Accelerator | Baseline (non-optimized) | NVIDIA TensorRT (optimized) | Percentage improvement |
|
36 |
|-------------|--------------------------|-----------------------------|------------------------|
|
@@ -44,5 +44,4 @@ Image throughput for 30 steps
|
|
44 |
- **Developed by:** Stability AI
|
45 |
- **Model type:** Diffusion-based text-to-image generative model
|
46 |
- **License:** [CreativeML Open RAIL++-M License](https://huggingface.co/stabilityai/stable-diffusion-xl-refiner-1.0/blob/main/LICENSE.md)
|
47 |
-
- **Model Description:** This is a conversion of the [
|
48 |
-
|
|
|
13 |
# Stable Diffusion XL 1.0 TensorRT
|
14 |
|
15 |
### Introduction
|
|
|
16 |
|
17 |
+
This repository hosts the TensorRT versions of **Stable Diffusion XL 1.0** created in collaboration with [NVIDIA](https://huggingface.co/nvidia). The optimized versions give substantial improvements in speed and efficiency.
|
18 |
+
|
19 |
|
20 |
![examples](./examples.jpg)
|
21 |
|
22 |
|
23 |
### Performance Comparison
|
24 |
|
25 |
+
#### Timings for 30 steps at 1024x1024
|
26 |
|
27 |
| Accelerator | Baseline (non-optimized) | NVIDIA TensorRT (optimized) | Percentage improvement |
|
28 |
|-------------|--------------------------|-----------------------------|------------------------|
|
|
|
30 |
| A100 | 3704 ms | 2742 ms | ~26% |
|
31 |
| H100 | 2496 ms | 1471 ms | ~41% |
|
32 |
|
33 |
+
#### Image throughput for 30 steps
|
34 |
|
35 |
| Accelerator | Baseline (non-optimized) | NVIDIA TensorRT (optimized) | Percentage improvement |
|
36 |
|-------------|--------------------------|-----------------------------|------------------------|
|
|
|
44 |
- **Developed by:** Stability AI
|
45 |
- **Model type:** Diffusion-based text-to-image generative model
|
46 |
- **License:** [CreativeML Open RAIL++-M License](https://huggingface.co/stabilityai/stable-diffusion-xl-refiner-1.0/blob/main/LICENSE.md)
|
47 |
+
- **Model Description:** This is a conversion of the [SDXL base 1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0) and [SDXL refiner 1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-refiner-1.0) models for [NVIDIA TensorRT](https://developer.nvidia.com/tensorrt) optimized inference
|
|