Update evi.py

evi.py

@@ -151,7 +151,7 @@ class Evi(datasets.GeneratorBasedBuilder):
             audio_path = audio_paths[lang]
             with open(text_path, encoding="utf-8") as fin:
                 reader = csv.DictReader(
-                    fin, delimiter="\t", skipinitialspace=True
+                    fin, delimiter=",", skipinitialspace=True
                 )
                 for dictrow in reader:
                     dialogue_id = dictrow["dialogue_id"]