run black

Signed-off-by: Sarah Yurick <sarahyurick@gmail.com>
NVIDIA · Oct 23, 2024 · 72c444b · 72c444b
1 parent 81d8fec
commit 72c444b
Show file tree

Hide file tree

Showing 2 changed files with 2 additions and 3 deletions.
diff --git a/nemo_curator/utils/text_utils.py b/nemo_curator/utils/text_utils.py
@@ -45,6 +45,7 @@ def mecab_splitter(text):
         return mecab_splitter
 
     else:
+
         def default_splitter(text):
             return text.split()
 

diff --git a/tests/test_filters.py b/tests/test_filters.py
@@ -431,9 +431,7 @@ def test_wordcount(self):
         ), f"Expected {expected_data} but got {filtered_data}"
 
     def test_wordcount_zh(self):
-        dataset = list_to_dataset(
-            ["", "你好。", "我喜欢学习中文。"]
-        )
+        dataset = list_to_dataset(["", "你好。", "我喜欢学习中文。"])
         filters = ScoreFilter(WordCountFilter(min_words=2, max_words=5, lang="zh"))
         filtered_data = filters(dataset)