Datasets:
Tasks:
Text Classification
Sub-tasks:
acceptability-classification
Languages:
English
Size:
10K<n<100K
Tags:
cola
License:
Update blimp_classification.py
Browse files- blimp_classification.py +8 -9
blimp_classification.py
CHANGED
@@ -67,7 +67,6 @@ class BlimpConfig(datasets.BuilderConfig):
|
|
67 |
self,
|
68 |
text_features,
|
69 |
label_classes=None,
|
70 |
-
process_label=lambda x: x,
|
71 |
**kwargs,
|
72 |
):
|
73 |
"""BuilderConfig for Blimp.
|
@@ -99,7 +98,6 @@ class BlimpConfig(datasets.BuilderConfig):
|
|
99 |
self.data_url = DATA_URL
|
100 |
self.data_dir = os.path.join("blimp", self.name)
|
101 |
self.citation = textwrap.dedent(_Blimp_CITATION)
|
102 |
-
self.process_label = process_label
|
103 |
self.description = ""
|
104 |
self.url = ""
|
105 |
|
@@ -113,23 +111,23 @@ class Blimp(datasets.GeneratorBasedBuilder):
|
|
113 |
BUILDER_CONFIGS = [
|
114 |
BlimpConfig(
|
115 |
name="syntax_semantics",
|
116 |
-
text_features={"
|
117 |
),
|
118 |
BlimpConfig(
|
119 |
-
name=
|
120 |
-
text_features={"
|
121 |
),
|
122 |
BlimpConfig(
|
123 |
name="morphology",
|
124 |
-
text_features={"
|
125 |
),
|
126 |
BlimpConfig(
|
127 |
name="syntax",
|
128 |
-
text_features={"
|
129 |
),
|
130 |
BlimpConfig(
|
131 |
name="semantics",
|
132 |
-
text_features={"
|
133 |
),
|
134 |
]
|
135 |
|
@@ -168,7 +166,6 @@ class Blimp(datasets.GeneratorBasedBuilder):
|
|
168 |
|
169 |
def _generate_examples(self, data_file, split):
|
170 |
|
171 |
-
process_label = self.config.process_label
|
172 |
label_classes = self.config.label_classes
|
173 |
|
174 |
with open(data_file) as f:
|
@@ -190,4 +187,6 @@ class Blimp(datasets.GeneratorBasedBuilder):
|
|
190 |
example["label"] = process_label(label)
|
191 |
else:
|
192 |
example["label"] = process_label(-1)
|
|
|
|
|
193 |
yield example["idx"], example
|
|
|
67 |
self,
|
68 |
text_features,
|
69 |
label_classes=None,
|
|
|
70 |
**kwargs,
|
71 |
):
|
72 |
"""BuilderConfig for Blimp.
|
|
|
98 |
self.data_url = DATA_URL
|
99 |
self.data_dir = os.path.join("blimp", self.name)
|
100 |
self.citation = textwrap.dedent(_Blimp_CITATION)
|
|
|
101 |
self.description = ""
|
102 |
self.url = ""
|
103 |
|
|
|
111 |
BUILDER_CONFIGS = [
|
112 |
BlimpConfig(
|
113 |
name="syntax_semantics",
|
114 |
+
text_features={"sentence": "sentence"},
|
115 |
),
|
116 |
BlimpConfig(
|
117 |
+
name="syntax+semantics",
|
118 |
+
text_features={"sentence": "sentence"},
|
119 |
),
|
120 |
BlimpConfig(
|
121 |
name="morphology",
|
122 |
+
text_features={"sentence": "text"},
|
123 |
),
|
124 |
BlimpConfig(
|
125 |
name="syntax",
|
126 |
+
text_features={"sentence": "sentence"},
|
127 |
),
|
128 |
BlimpConfig(
|
129 |
name="semantics",
|
130 |
+
text_features={"sentence": "sentence"},
|
131 |
),
|
132 |
]
|
133 |
|
|
|
166 |
|
167 |
def _generate_examples(self, data_file, split):
|
168 |
|
|
|
169 |
label_classes = self.config.label_classes
|
170 |
|
171 |
with open(data_file) as f:
|
|
|
187 |
example["label"] = process_label(label)
|
188 |
else:
|
189 |
example["label"] = process_label(-1)
|
190 |
+
example['sentence'] = example['text']
|
191 |
+
del example['text']
|
192 |
yield example["idx"], example
|