Fix vocab_size

tokenization_qwen.py

@@ -237,7 +237,10 @@ class QWenTokenizer(PreTrainedTokenizer):
 
     @property
     def vocab_size(self):
-        return self.tokenizer.n_vocab
+        """
+        Size of the base vocabulary (without the added tokens).
+        """
+        return len(self.mergeable_ranks)
 
     def _convert_id_to_token(self, index: int) -> Union[bytes, str]:
         """Converts an id to a token, special tokens included"""