Update blimp_classification.py

blimp_classification.py

@@ -67,7 +67,6 @@ class BlimpConfig(datasets.BuilderConfig):
         self,
         text_features,
         label_classes=None,
-        process_label=lambda x: x,
         **kwargs,
     ):
         """BuilderConfig for Blimp.
@@ -99,7 +98,6 @@ class BlimpConfig(datasets.BuilderConfig):
         self.data_url = DATA_URL
         self.data_dir = os.path.join("blimp", self.name)
         self.citation = textwrap.dedent(_Blimp_CITATION)
-        self.process_label = process_label
         self.description = ""
         self.url = ""
 
@@ -113,23 +111,23 @@ class Blimp(datasets.GeneratorBasedBuilder):
     BUILDER_CONFIGS = [
         BlimpConfig(
             name="syntax_semantics",
-            text_features={"text": "text"},
+            text_features={"sentence": "sentence"},
         ),
         BlimpConfig(
-            name=r"syntax+semantics",
-            text_features={"text": "text"},
+            name="syntax+semantics",
+            text_features={"sentence": "sentence"},
         ),
         BlimpConfig(
             name="morphology",
-            text_features={"text": "text"},
+            text_features={"sentence": "text"},
         ),
         BlimpConfig(
             name="syntax",
-            text_features={"text": "text"},
+            text_features={"sentence": "sentence"},
         ),
         BlimpConfig(
             name="semantics",
-            text_features={"text": "text"},
+            text_features={"sentence": "sentence"},
         ),
     ]
 
@@ -168,7 +166,6 @@ class Blimp(datasets.GeneratorBasedBuilder):
 
     def _generate_examples(self, data_file, split):
 
-        process_label = self.config.process_label
         label_classes = self.config.label_classes
 
         with open(data_file) as f:
@@ -190,4 +187,6 @@ class Blimp(datasets.GeneratorBasedBuilder):
                     example["label"] = process_label(label)
                 else:
                     example["label"] = process_label(-1)
+                example['sentence'] = example['text']
+                del example['text']
                 yield example["idx"], example