diff --git a/official/nlp/lstm/src/imdb.py b/official/nlp/lstm/src/imdb.py
index 7e1976ed7d34b3b77724e74a85ca8a99fe14f3ee..6fcdd5847fa73b3305f6160c2c3ae3a1192c3d56 100644
--- a/official/nlp/lstm/src/imdb.py
+++ b/official/nlp/lstm/src/imdb.py
@@ -47,7 +47,7 @@ class ImdbParser():
         """
         parse imdb data to memory
         """
-        self.__wvmodel = gensim.models.KeyedVectors.load_word2vec_format(self.__glove_file, no_header=True)
+        self.__wvmodel = gensim.models.KeyedVectors.load_word2vec_format(self.__glove_file)
 
         for seg in self.__segs:
             self.__parse_imdb_datas(seg)