shivangibithel
commited on
Commit
•
5094fd8
1
Parent(s):
2ee35ff
Update sotab.py
Browse files
sotab.py
CHANGED
@@ -34,7 +34,7 @@ _LICENSE = ""
|
|
34 |
# The HuggingFace Datasets library doesn't host the datasets but only points to the original files.
|
35 |
# This can be an arbitrary nested dict/list of URLs (see below in `_split_generators` method)
|
36 |
_DATA_URL = (
|
37 |
-
"https://github.com/ppasupat/WikiTableQuestions/releases/download/v1.0.2/WikiTableQuestions-1.0.2-compact.zip"
|
38 |
)
|
39 |
|
40 |
|
@@ -58,7 +58,7 @@ class WikiTableQuestions(datasets.GeneratorBasedBuilder):
|
|
58 |
datasets.BuilderConfig(
|
59 |
name="random-split-1",
|
60 |
version=VERSION,
|
61 |
-
description="
|
62 |
),
|
63 |
]
|
64 |
|
@@ -70,7 +70,7 @@ class WikiTableQuestions(datasets.GeneratorBasedBuilder):
|
|
70 |
features = datasets.Features(
|
71 |
{
|
72 |
"id": datasets.Value("string"),
|
73 |
-
"column_index": datasets.Value("
|
74 |
"label": datasets.features.Sequence(datasets.Value("string")),
|
75 |
"table": {
|
76 |
"header": datasets.features.Sequence(datasets.Value("string")),
|
@@ -96,27 +96,33 @@ class WikiTableQuestions(datasets.GeneratorBasedBuilder):
|
|
96 |
)
|
97 |
|
98 |
def _split_generators(self, dl_manager):
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
|
104 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
105 |
return [
|
106 |
datasets.SplitGenerator(
|
107 |
name=datasets.Split.TRAIN,
|
108 |
# These kwargs will be passed to _generate_examples
|
109 |
-
gen_kwargs={"main_filepath": os.path.join(root_dir, "
|
110 |
),
|
111 |
datasets.SplitGenerator(
|
112 |
name=datasets.Split.TEST,
|
113 |
# These kwargs will be passed to _generate_examples
|
114 |
-
gen_kwargs={"main_filepath": os.path.join(root_dir, "
|
115 |
),
|
116 |
datasets.SplitGenerator(
|
117 |
name=datasets.Split.VALIDATION,
|
118 |
# These kwargs will be passed to _generate_examples
|
119 |
-
gen_kwargs={"main_filepath": os.path.join(root_dir, "
|
120 |
),
|
121 |
]
|
122 |
|
|
|
34 |
# The HuggingFace Datasets library doesn't host the datasets but only points to the original files.
|
35 |
# This can be an arbitrary nested dict/list of URLs (see below in `_split_generators` method)
|
36 |
_DATA_URL = (
|
37 |
+
# "https://github.com/ppasupat/WikiTableQuestions/releases/download/v1.0.2/WikiTableQuestions-1.0.2-compact.zip"
|
38 |
)
|
39 |
|
40 |
|
|
|
58 |
datasets.BuilderConfig(
|
59 |
name="random-split-1",
|
60 |
version=VERSION,
|
61 |
+
description="",
|
62 |
),
|
63 |
]
|
64 |
|
|
|
70 |
features = datasets.Features(
|
71 |
{
|
72 |
"id": datasets.Value("string"),
|
73 |
+
"column_index": datasets.Value("int32"),
|
74 |
"label": datasets.features.Sequence(datasets.Value("string")),
|
75 |
"table": {
|
76 |
"header": datasets.features.Sequence(datasets.Value("string")),
|
|
|
96 |
)
|
97 |
|
98 |
def _split_generators(self, dl_manager):
|
99 |
+
train_url = "https://huggingface.co/datasets/shivangibithel/SOTAB/resolve/main/CTA_Training.zip"
|
100 |
+
dev_url = "https://huggingface.co/datasets/shivangibithel/SOTAB/resolve/main/CTA_Validation.zip"
|
101 |
+
test_url = "https://huggingface.co/datasets/shivangibithel/SOTAB/resolve/main/CTA_Test.zip"
|
102 |
+
|
103 |
+
CTA_Training = os.path.join(dl_manager.download_and_extract(train_url), "CTA_Training")
|
104 |
+
CTA_Validation = os.path.join(dl_manager.download_and_extract(dev_url), "CTA_Validation")
|
105 |
+
CTA_Test = os.path.join(dl_manager.download_and_extract(test_url), "CTA_Test")
|
106 |
+
|
107 |
+
train_file = "{}.json.gz".format(self.config.name)
|
108 |
+
test_file = "{}.json.gz".format(self.config.name)
|
109 |
+
dev_file = "{}.json.gz".format(self.config.name)
|
110 |
+
|
111 |
return [
|
112 |
datasets.SplitGenerator(
|
113 |
name=datasets.Split.TRAIN,
|
114 |
# These kwargs will be passed to _generate_examples
|
115 |
+
gen_kwargs={"main_filepath": os.path.join(root_dir, "Train", train_file), "root_dir": CTA_Training},
|
116 |
),
|
117 |
datasets.SplitGenerator(
|
118 |
name=datasets.Split.TEST,
|
119 |
# These kwargs will be passed to _generate_examples
|
120 |
+
gen_kwargs={"main_filepath": os.path.join(root_dir, "Test", test_file), "root_dir": CTA_Test},
|
121 |
),
|
122 |
datasets.SplitGenerator(
|
123 |
name=datasets.Split.VALIDATION,
|
124 |
# These kwargs will be passed to _generate_examples
|
125 |
+
gen_kwargs={"main_filepath": os.path.join(root_dir, "Validation", dev_file), "root_dir": CTA_Validation},
|
126 |
),
|
127 |
]
|
128 |
|