Spaces:
Running
Running
Fix typo
Browse files- introduction.md +1 -1
introduction.md
CHANGED
|
@@ -47,7 +47,7 @@ working on the training and on the loss gave us the final increase that you can
|
|
| 47 |
We eventually had to deal with the fact that we do not have the same data that OpenAI had during the training of CLIP.
|
| 48 |
Thus, we tried to add as much data as possible while keeping the data-quality as high as possible.
|
| 49 |
|
| 50 |
-
We considered
|
| 51 |
|
| 52 |
+ [WIT](https://github.com/google-research-datasets/wit) is an image-caption dataset collected from Wikipedia (see,
|
| 53 |
[Srinivasan et al., 2021](https://arxiv.org/pdf/2103.01913.pdf)). We focused on the *Reference Description* captions described in the paper as they are
|
|
|
|
| 47 |
We eventually had to deal with the fact that we do not have the same data that OpenAI had during the training of CLIP.
|
| 48 |
Thus, we tried to add as much data as possible while keeping the data-quality as high as possible.
|
| 49 |
|
| 50 |
+
We considered four main sources of data:
|
| 51 |
|
| 52 |
+ [WIT](https://github.com/google-research-datasets/wit) is an image-caption dataset collected from Wikipedia (see,
|
| 53 |
[Srinivasan et al., 2021](https://arxiv.org/pdf/2103.01913.pdf)). We focused on the *Reference Description* captions described in the paper as they are
|