NAType values to None

rixvox.py

@@ -177,6 +177,8 @@ class Rixvox(datasets.GeneratorBasedBuilder):
             df_meta = pd.read_parquet(meta_paths[self.config.name])
 
         df_meta = df_meta.set_index("filename", drop=False)
+        # Column contains NAType, so we convert to object type column and NAType to None values.
+        df_meta["birth_year"] = df_meta["birth_year"].astype("object").where(df_meta["birth_year"].notnull(), None)
 
         for i, audio_archive in enumerate(archive_iters):
             for filename, file in audio_archive: