Update wit-dataset.py
Browse filesfixing filepath variable
- wit-dataset.py +3 -4
wit-dataset.py
CHANGED
|
@@ -1,8 +1,7 @@
|
|
| 1 |
import datasets
|
| 2 |
|
| 3 |
logger = datasets.logging.get_logger(__name__)
|
| 4 |
-
_DESCRIPTION = """
|
| 5 |
-
Wikipedia-based Image Text (WIT) Dataset is a large multimodal multilingual dataset.
|
| 6 |
WIT is composed of a curated set of 37.6 million entity rich image-text examples with 11.5 million unique images across 108 Wikipedia languages.
|
| 7 |
Its size enables WIT to be used as a pretraining dataset for multimodal machine learning models.
|
| 8 |
"""
|
|
@@ -193,8 +192,8 @@ class WIT(datasets.GeneratorBasedBuilder):
|
|
| 193 |
data_fields = list(self._info().features.keys())
|
| 194 |
path_idx = data_fields.index("image_url")
|
| 195 |
# ToDO: Remove after debugging..
|
| 196 |
-
print(
|
| 197 |
-
with open(
|
| 198 |
lines = f.readlines()
|
| 199 |
headline = line[0]
|
| 200 |
|
|
|
|
| 1 |
import datasets
|
| 2 |
|
| 3 |
logger = datasets.logging.get_logger(__name__)
|
| 4 |
+
_DESCRIPTION = """\\nWikipedia-based Image Text (WIT) Dataset is a large multimodal multilingual dataset.
|
|
|
|
| 5 |
WIT is composed of a curated set of 37.6 million entity rich image-text examples with 11.5 million unique images across 108 Wikipedia languages.
|
| 6 |
Its size enables WIT to be used as a pretraining dataset for multimodal machine learning models.
|
| 7 |
"""
|
|
|
|
| 192 |
data_fields = list(self._info().features.keys())
|
| 193 |
path_idx = data_fields.index("image_url")
|
| 194 |
# ToDO: Remove after debugging..
|
| 195 |
+
print(filepath)
|
| 196 |
+
with open(filepath, encoding="utf-8") as f:
|
| 197 |
lines = f.readlines()
|
| 198 |
headline = line[0]
|
| 199 |
|