fix: HateBR labels

portuguese_benchmark.py

@@ -1,5 +1,5 @@
 import datasets
-from typing import Dict, List, Optional, Union
+from typing import Dict, List, Optional, Union, Callable
 import json
 import textwrap
 
@@ -113,7 +113,7 @@ _HATEBR_KWARGS = dict(
     ),
     task_type="classification",
     file_type="csv",
-    label_classes=[0, 1, 2, 3],
+    label_classes=["non-offensive", "slightly", "moderately", "highly"],
     data_urls={
         "train": "https://raw.githubusercontent.com/franciellevargas/HateBR/2d18c5b9410c2dfdd6d5394caa54d608857dae7c/dataset/HateBR.csv"
     },
@@ -131,6 +131,13 @@ _HATEBR_KWARGS = dict(
     text_and_label_columns=["instagram_comments", "offensiveness_levels"],
     indexes_url="https://huggingface.co/datasets/ruanchaves/hatebr/raw/main/indexes.json"
 )
+hatebr_map = {
+    "0": "non-offensive",
+    "1": "slightly",
+    "2": "moderately",
+    "3": "highly",
+}
+_HATEBR_KWARGS['process_label'] = lambda x: hatebr_map[x]
 
 class PTBenchmarkConfig(datasets.BuilderConfig):
     """BuilderConfig for PTBenchmark."""
@@ -144,7 +151,8 @@ class PTBenchmarkConfig(datasets.BuilderConfig):
         label_classes: Optional[List[Union[str, int]]] = None,
         file_type: Optional[str] = None, #filetype (csv, tsc, jsonl)
         text_and_label_columns: Optional[List[str]] = None, #columns for train, dev and test for csv datasets
-        indexes_url=None, #indexes for train, dev and test for single file datasets
+        indexes_url: Optional[str] = None, #indexes for train, dev and test for single file datasets
+        process_label: Optional[Callable[[str], str]] = lambda x: x,
         **kwargs,
     ):
         """BuilderConfig for GLUE.
@@ -174,6 +182,7 @@ class PTBenchmarkConfig(datasets.BuilderConfig):
         self.file_type = file_type
         self.text_and_label_columns = text_and_label_columns
         self.indexes_url = indexes_url
+        self.process_label = process_label
 
 def _get_classification_features(config):
     return datasets.Features(
@@ -220,11 +229,12 @@ def _get_sts_features(config):
     )
 
 def _csv_generator(file_path: str,
-                   columns: List[str],
                    indexes_path: Optional[str] = None,
-                   split: Optional[str] = None):
+                   split: Optional[str] = None,
+                   config: PTBenchmarkConfig = None):
     """Yields examples."""
     df = pd.read_csv(file_path)
+    columns = config.text_and_label_columns
     df = df[columns]
 
     with open(indexes_path, "r") as f:
@@ -236,11 +246,11 @@ def _csv_generator(file_path: str,
         example = {
             "idx": id_,
             "sentence": str(row[columns[0]]),
-            "label": int(row[columns[-1]])
+            "label": config.process_label(str(row[columns[-1]]))
         }
         yield id_, example
 
-def _conll_ner_generator(file_path):
+def _conll_ner_generator(file_path: str, config: PTBenchmarkConfig = None):
     with open(file_path, encoding="utf-8") as f:
 
         guid = 0
@@ -261,7 +271,7 @@ def _conll_ner_generator(file_path):
             else:
                 splits = line.split(" ")
                 tokens.append(splits[0])
-                ner_tags.append(splits[1].rstrip())
+                ner_tags.append(config.process_label(splits[1].rstrip()))
 
         # last example
         yield guid, {
@@ -270,7 +280,7 @@ def _conll_ner_generator(file_path):
             "ner_tags": ner_tags,
         }
 
-def _assin2_generator(file_path, type):
+def _assin2_generator(file_path, config: PTBenchmarkConfig = None):
     """Yields examples."""
     id_ = 0
 
@@ -279,6 +289,8 @@ def _assin2_generator(file_path, type):
         tree = ET.parse(f)
         root = tree.getroot()
 
+        task_type = config.task_type
+
         for pair in root:
 
             example = {
@@ -286,10 +298,10 @@ def _assin2_generator(file_path, type):
                 "sentence1": pair.find(".//t").text,
                 "sentence2": pair.find(".//h").text
             }
-            if type == "rte":
+            if task_type == "rte":
                 example["label"] = pair.attrib.get("entailment").upper()
-            elif type == "sts":
-                example["label"] = float(pair.attrib.get("similarity"))
+            elif task_type == "sts":
+                example["label"] = float(config.process_label(pair.attrib.get("similarity")))
 
             yield id_, example
             id_ += 1
@@ -378,16 +390,16 @@ class PTBenchmark(datasets.GeneratorBasedBuilder):
             if self.config.file_type == "csv":
                 yield from _csv_generator(
                     file_path, 
-                    self.config.text_and_label_columns,
                     indexes_path=indexes_path,
-                    split=split
+                    split=split,
+                    config=self.config
                 )
         elif self.config.task_type == "ner":
-            yield from _conll_ner_generator(file_path)
+            yield from _conll_ner_generator(file_path, config=self.config)
         elif self.config.task_type == "rte":
             if "assin2" in self.config.name:
-                yield from _assin2_generator(file_path, "rte")
+                yield from _assin2_generator(file_path, config=self.config)
         elif self.config.task_type == "sts":
             if "assin2" in self.config.name:
-                yield from _assin2_generator(file_path, "sts")
+                yield from _assin2_generator(file_path, config=self.config)