indexing reader

ml-superb-subset.py

@@ -91,7 +91,7 @@ class MLSuperb(datasets.GeneratorBasedBuilder):
         metadata = {}
         with open(transcript_path, encoding="utf-8") as f:
             reader = csv.reader(f, delimiter="\t", quoting=csv.QUOTE_NONE)
-            if len(reader[0]) == 1:
+            if len(next(reader)) == 1:
                 reader = csv.reader(f, delimiter=" ", quoting=csv.QUOTE_NONE)
             for row in reader:
                 id_ = row[0]