Commit
•
92613ad
1
Parent(s):
bc40e9e
Update README.md
Browse files
README.md
CHANGED
|
@@ -23,4 +23,9 @@ RuDOLPH 350M is a fast and light text-image-text transformer (350M GPT-3) design
|
|
| 23 |
|
| 24 |
The primary proposed method is to modify the sparse transformer's attention mask to better control multi-modalities and up to the next level with "hyper-modality". It allows us to calculate the transitions of modalities in both directions, unlike another similar work DALL-E Transformer, which used only one direction, "text to image". The proposed "image to right text" direction is achieved by extension sparse attention mask to the right for auto-repressively text generation with image condition without attention to left text.
|
| 25 |
|
| 26 |
-
<img src="https://raw.githubusercontent.com/shonenkov/ru-dolph/master/pics/attention_masks.png?token=AHV2MCP7BH3CQBAK74UVA7TB4CXQE" height="40" border="2"/>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23 |
|
| 24 |
The primary proposed method is to modify the sparse transformer's attention mask to better control multi-modalities and up to the next level with "hyper-modality". It allows us to calculate the transitions of modalities in both directions, unlike another similar work DALL-E Transformer, which used only one direction, "text to image". The proposed "image to right text" direction is achieved by extension sparse attention mask to the right for auto-repressively text generation with image condition without attention to left text.
|
| 25 |
|
| 26 |
+
<img src="https://raw.githubusercontent.com/shonenkov/ru-dolph/master/pics/attention_masks.png?token=AHV2MCP7BH3CQBAK74UVA7TB4CXQE" height="40" border="2"/>
|
| 27 |
+
|
| 28 |
+
# Authors
|
| 29 |
+
|
| 30 |
+
+ Alex Shonenkov: [Github](https://github.com/shonenkov), [Kaggle GM](https://www.kaggle.com/shonenkov)
|
| 31 |
+
+ Michael Konstantinov: [Mishin Learning](https://t.me/mishin_learning), [Transformer Community](https://transformer.community/)
|