Commit
·
f9fca7f
1
Parent(s):
ef2db1d
Update winogavil.py
Browse files- winogavil.py +6 -6
winogavil.py
CHANGED
@@ -52,6 +52,7 @@ class Winogavil(datasets.GeneratorBasedBuilder):
|
|
52 |
BUILDER_CONFIGS = [
|
53 |
datasets.BuilderConfig(name="TEST", version=VERSION, description="winogavil dataset"),
|
54 |
]
|
|
|
55 |
|
56 |
def _info(self):
|
57 |
img = datasets.Image()
|
@@ -92,12 +93,14 @@ class Winogavil(datasets.GeneratorBasedBuilder):
|
|
92 |
# It can accept any type or nested list/dict and will give back the same structure with the url replaced with path to local files.
|
93 |
# By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
|
94 |
data_dir = dl_manager.download_and_extract({
|
95 |
-
"examples_csv": hf_hub_url("datasets/nlphuji/winogavil", filename="winogavil_dataset.csv")
|
|
|
|
|
96 |
|
97 |
return [datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs=data_dir)]
|
98 |
|
99 |
# method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
|
100 |
-
def _generate_examples(self, examples_csv):
|
101 |
# The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
|
102 |
|
103 |
df = pd.read_csv(examples_csv)
|
@@ -109,11 +112,8 @@ class Winogavil(datasets.GeneratorBasedBuilder):
|
|
109 |
for r_idx, r in df.iterrows():
|
110 |
r_dict = r.to_dict()
|
111 |
r_dict['candidates'] = json.loads(r_dict['candidates'])
|
112 |
-
candidates_images = [
|
113 |
r_dict['candidates_images'] = candidates_images
|
114 |
r_dict['associations'] = json.loads(r_dict['associations'])
|
115 |
key = r['ID']
|
116 |
yield key, r_dict
|
117 |
-
|
118 |
-
def get_img_url(image_name):
|
119 |
-
return 'https://winogavil.s3.eu-west-1.amazonaws.com/{}'.format(image_name + ".jpg")
|
|
|
52 |
BUILDER_CONFIGS = [
|
53 |
datasets.BuilderConfig(name="TEST", version=VERSION, description="winogavil dataset"),
|
54 |
]
|
55 |
+
IMAGE_EXTENSION = ".jpg"
|
56 |
|
57 |
def _info(self):
|
58 |
img = datasets.Image()
|
|
|
93 |
# It can accept any type or nested list/dict and will give back the same structure with the url replaced with path to local files.
|
94 |
# By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
|
95 |
data_dir = dl_manager.download_and_extract({
|
96 |
+
"examples_csv": hf_hub_url("datasets/nlphuji/winogavil", filename="winogavil_dataset.csv"),
|
97 |
+
"images_dir": hf_hub_url("datasets/facebook/winoground", filename="winogavil_images.zip")
|
98 |
+
})
|
99 |
|
100 |
return [datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs=data_dir)]
|
101 |
|
102 |
# method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
|
103 |
+
def _generate_examples(self, examples_csv, images_dir):
|
104 |
# The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
|
105 |
|
106 |
df = pd.read_csv(examples_csv)
|
|
|
112 |
for r_idx, r in df.iterrows():
|
113 |
r_dict = r.to_dict()
|
114 |
r_dict['candidates'] = json.loads(r_dict['candidates'])
|
115 |
+
candidates_images = [os.path.join(images_dir, f"{x}.{self.IMAGE_EXTENSION}") for x in r['candidates']]
|
116 |
r_dict['candidates_images'] = candidates_images
|
117 |
r_dict['associations'] = json.loads(r_dict['associations'])
|
118 |
key = r['ID']
|
119 |
yield key, r_dict
|
|
|
|
|
|