Yihong Luo
commited on
Commit
•
e494dc2
1
Parent(s):
a662069
Update README.md
Browse files
README.md
CHANGED
|
@@ -4,6 +4,9 @@ This algorithm was proposed in YOSO.
|
|
| 4 |
|
| 5 |
This model is fine-tuning from PixArt, enabling one-step inference to perform text-to-image generation.
|
| 6 |
|
|
|
|
|
|
|
|
|
|
| 7 |
## usage
|
| 8 |
```python
|
| 9 |
from diffusers import PixArtAlphaPipeline, LCMScheduler, Transformer2DModel, DPMSolverMultistepScheduler
|
|
|
|
| 4 |
|
| 5 |
This model is fine-tuning from PixArt, enabling one-step inference to perform text-to-image generation.
|
| 6 |
|
| 7 |
+
We wanna highlight that the YOSO-PixArt was originally trained on 512 resolution. However we found that we can construct a YOSO that enables generating samples with 1024 resolution by merging with PixArt-1024 (Eq(17) in the paper) as follows:
|
| 8 |
+
|
| 9 |
+
|
| 10 |
## usage
|
| 11 |
```python
|
| 12 |
from diffusers import PixArtAlphaPipeline, LCMScheduler, Transformer2DModel, DPMSolverMultistepScheduler
|