From e7a17bb15360e17a8d6c646e521793565e7c0ff7 Mon Sep 17 00:00:00 2001
From: pruthwik mishra <pruthwik.mishra@research.iiit.ac.in>
Date: Tue, 16 Aug 2022 11:03:59 +0530
Subject: [PATCH] Update tokenizer_for_indian_languages_on_files.py

---
 tokenizer_for_indian_languages_on_files.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tokenizer_for_indian_languages_on_files.py b/tokenizer_for_indian_languages_on_files.py
index f439cf7..52bb3d5 100644
--- a/tokenizer_for_indian_languages_on_files.py
+++ b/tokenizer_for_indian_languages_on_files.py
@@ -137,7 +137,7 @@ def main():
         os.makedirs(args.out)
     if args.lang in ['hi', 'or', 'mn', 'as', 'bn', 'pa']:
         lang = 0
-    elif args.lang == 'ur':
+    elif args.lang in ['ur', 'ks']:
         lang = 1
     elif args.lang in ['en', 'gu', 'mr', 'ml', 'kn', 'te', 'ta']:
         lang = 2
-- 
GitLab