Upload 2 files

Remove the 'neutral' label in MARC-ja when 'remove_neutral' is specified in the config

JGLUE.py

@@ -178,11 +178,12 @@ def dataset_info_jcommonsenseqa() -> ds.DatasetInfo:
     )
 
 
-def dataset_info_marc_ja() -> ds.DatasetInfo:
+def dataset_info_marc_ja(remove_netural: bool) -> ds.DatasetInfo:
+    labels = ["positive", "negative"] if remove_netural else ["positive", "negative", "neutral"]
     features = ds.Features(
         {
             "sentence": ds.Value("string"),
-            "label": ds.ClassLabel(num_classes=3, names=["positive", "negative", "neutral"]),
+            "label": ds.ClassLabel(num_classes=len(labels), names=labels),
             "review_id": ds.Value("string"),
         }
     )
@@ -235,7 +236,7 @@ class JGLUE(ds.GeneratorBasedBuilder):
         elif self.config.name == "JCommonsenseQA":
             return dataset_info_jcommonsenseqa()
         elif self.config.name == "MARC-ja":
-            return dataset_info_marc_ja()
+            return dataset_info_marc_ja(self.config.remove_netural)
         else:
             raise ValueError(f"Invalid config name: {self.config.name}")
 

preprocess_marc_ja.py

@@ -23,7 +23,7 @@ class MarcJaConfig(ds.BuilderConfig):
         is_han_to_zen: bool = False,
         max_instance_num: Optional[int] = None,
         max_char_length: int = 500,
-        is_pos_neg: bool = True,
+        remove_netural: bool = True,
         train_ratio: float = 0.94,
         val_ratio: float = 0.03,
         test_ratio: float = 0.03,
@@ -55,20 +55,20 @@ class MarcJaConfig(ds.BuilderConfig):
         self.is_han_to_zen = is_han_to_zen
         self.max_instance_num = max_instance_num
         self.max_char_length = max_char_length
-        self.is_pos_neg = is_pos_neg
+        self.remove_netural = remove_netural
         self.output_testset = output_testset
 
         self.filter_review_id_list_valid = filter_review_id_list_valid
         self.label_conv_review_id_list_valid = label_conv_review_id_list_valid
 
 
-def get_label(rating: int, is_pos_neg: bool = False) -> Optional[str]:
+def get_label(rating: int, remove_netural: bool = False) -> Optional[str]:
     if rating >= 4:
         return "positive"
     elif rating <= 2:
         return "negative"
     else:
-        if is_pos_neg:
+        if remove_netural:
             return None
         else:
             return "neutral"
@@ -225,7 +225,7 @@ def preprocess_marc_ja(
 
     # convert the rating to label
     tqdm.pandas(dynamic_ncols=True, desc="Convert the rating to the label")
-    df = df.assign(label=df["rating"].progress_apply(lambda rating: get_label(rating, config.is_pos_neg)))
+    df = df.assign(label=df["rating"].progress_apply(lambda rating: get_label(rating, config.remove_netural)))
 
     # remove rows where the label is None
     df = df[~df["label"].isnull()]