Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -11,5 +11,11 @@ size_categories:
|
|
11 |
- 10K<n<100K
|
12 |
---
|
13 |
|
14 |
-
This dataset contains
|
15 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
- 10K<n<100K
|
12 |
---
|
13 |
|
14 |
+
This dataset contains:
|
15 |
+
- 31865 images (71.36%) in train dataset
|
16 |
+
- 9815 images (21.98%) in test dataset
|
17 |
+
- 2971 images (6.65%) in validation dataset
|
18 |
+
- Total: 44651 images
|
19 |
+
- sourced from the extensive Synthetic Word Dataset, a large-scale word-image dataset.
|
20 |
+
|
21 |
+
The original and complete dataset (9 million images, 10.68GB) can be found and downloaded at [this academic torrent](https://academictorrents.com/details/3d0b4f09080703d2a9c6be50715b46389fdb3af1).
|