Update translation_utils.py

llm_toolkit/translation_utils.py

@@ -23,7 +23,7 @@ comet = evaluate.load("comet")
 
 
 def extract_answer(text, debug=False):
-    if text:
+    if text and isinstance(text, str):
         # Remove the begin and end tokens
         text = re.sub(
             r".*?(assistant|\[/INST\]).+?\b", "", text, flags=re.DOTALL | re.MULTILINE
@@ -240,6 +240,9 @@ def detect_repetition_scores(row, col, debug=False):
 
 
 def count_chinese_characters(text):
+    if isinstance(text, str) is False:
+        return 0
+    
     # Define a regular expression pattern for Chinese characters
     chinese_char_pattern = r"[\u4e00-\u9fff]"