fix bug with age_group ClassLabel

author_profiling.py

@@ -92,12 +92,12 @@ class AuthorProfiling(datasets.GeneratorBasedBuilder):
                     "account_id": datasets.Value("string"),
                     "author_id": datasets.Value("string"),
                     "age": datasets.Value("int32"),
-                    "age_group": datasets.ClassLabel(names=["1-19", "20-29", "30-39", "40-49", "50+"]),
+                    "age_group": datasets.ClassLabel(names=["0-19", "20-29", "30-39", "40-49", "50+"]),
                     "gender": datasets.ClassLabel(names=["male", "female"]),
                     "no_imitation": datasets.ClassLabel(names=[0,1]),
-                    "age_imitation": datasets.Value("float16"), #datasets.ClassLabel(names=[0.0, 1.0, float("nan")]),
-                    "gender_imitation": datasets.Value("float16"), #datasets.ClassLabel(names=[0.0, 1.0, float("nan")]),
-                    "style_imitation": datasets.Value("float16"), #datasets.ClassLabel(names=[0.0, 1.0, float("nan")]),
+                    "age_imitation": datasets.Value("float32"), #datasets.ClassLabel(names=[0.0, 1.0, float("nan")]),
+                    "gender_imitation": datasets.Value("float32"), #datasets.ClassLabel(names=[0.0, 1.0, float("nan")]),
+                    "style_imitation": datasets.Value("float32"), #datasets.ClassLabel(names=[0.0, 1.0, float("nan")]),
                     # These are the features of your dataset like images, labels ...
                 }
             )