model card updated
Browse files
README.md
CHANGED
|
@@ -65,3 +65,14 @@ images = pipe(
|
|
| 65 |
|
| 66 |
images[0]
|
| 67 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
|
| 66 |
images[0]
|
| 67 |
```
|
| 68 |
+
|
| 69 |
+
|
| 70 |
+
## Limitations
|
| 71 |
+
|
| 72 |
+
The model is intended for research purposes only to show a way to reduce the unnecessary resource usage in existing T2I research.
|
| 73 |
+
|
| 74 |
+
As this prior model is trained using very small LAION subset and CLIP supervision, it will observe the limitations from the CLIP model such as:
|
| 75 |
+
* Lack of spatial understanding.
|
| 76 |
+
* Cannot render legible text
|
| 77 |
+
* Complex compositionality is still a big challenge that can be improved if CLIP is improved.
|
| 78 |
+
* While the capabilities of image generation models are impressive, they can also reinforce or exacerbate social biases.
|