Datasets:
Tasks:
Question Answering
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
100K - 1M
ArXiv:
Tags:
webdataset
License:
hheiden-roots
commited on
Update README.md
Browse files
README.md
CHANGED
@@ -58,7 +58,19 @@ This flat version of the dataset was specifically compiled to aid Large Multimod
|
|
58 |
- `page_id`: str, unique ID for the particular page
|
59 |
- `question`: str, the question to answer
|
60 |
- `answer`: str, the actual answer
|
61 |
-
- `image`:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
62 |
|
63 |
## Data Statistics
|
64 |
|
|
|
58 |
- `page_id`: str, unique ID for the particular page
|
59 |
- `question`: str, the question to answer
|
60 |
- `answer`: str, the actual answer
|
61 |
+
- `image`: str, a base64 encoded binary string of the image.
|
62 |
+
|
63 |
+
The `image` is converted back to a PIL.Image with:
|
64 |
+
```python
|
65 |
+
import base64
|
66 |
+
import io
|
67 |
+
|
68 |
+
|
69 |
+
def decode_base64_to_image(base64_string):
|
70 |
+
img_data = base64.b64decode(base64_string)
|
71 |
+
img = Image.open(io.BytesIO(img_data))
|
72 |
+
return img
|
73 |
+
```
|
74 |
|
75 |
## Data Statistics
|
76 |
|