Update README.md
Browse files
README.md
CHANGED
|
@@ -13,4 +13,8 @@ Architecture details:
|
|
| 13 |
4. Multi-headed attention with learned position- and data-dependent temperature scaling
|
| 14 |
5. Vision encoder initialized from SigLIP-SO-400M, with multi-crop channel concatenation for token-efficient high resolution image processing
|
| 15 |
|
| 16 |
-
For more details, please refer to our ||coming soon release blog post||.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
4. Multi-headed attention with learned position- and data-dependent temperature scaling
|
| 14 |
5. Vision encoder initialized from SigLIP-SO-400M, with multi-crop channel concatenation for token-efficient high resolution image processing
|
| 15 |
|
| 16 |
+
For more details, please refer to our ||coming soon release blog post||.
|
| 17 |
+
|
| 18 |
+
## Usage
|
| 19 |
+
|
| 20 |
+
* TODO: Add usage examples
|