Update arman.py

arman.py

@@ -26,9 +26,9 @@ ArmanPersoNERCorpus includes 250,015 tokens and 7,682 Persian sentences in total
 """
 _HOMEPAGE = "https://github.com/HaniehP/PersianNER"
 _DATA_URL = "https://github.com/HaniehP/PersianNER/raw/master/ArmanPersoNERCorpus.zip"
-_TRAINING_FILE = r'/data/train.txt'
-_DEV_FILE = r'/data/dev.txt'
-_TEST_FILE = r'/data/test.txt'
+_TRAINING_FILE = r'arman/data/train.txt'
+_DEV_FILE = r'arman/data/dev.txt'
+_TEST_FILE = r'arman/data/test.txt'
 
 class Arman(datasets.GeneratorBasedBuilder):
     """ArmanPersoNER Corpus"""