added tokenizers serialization tests

pytorch_pretrained_bert/tokenization_openai.py

@@ -263,7 +263,10 @@ class OpenAIGPTTokenizer(object):
         return out_string
 
     def save_vocabulary(self, vocab_path):
-        """Save the tokenizer vocabulary to a path."""
+        """Save the tokenizer vocabulary and merge files to a directory."""
+        if not os.path.isdir(vocab_path):
+            logger.error("Vocabulary path ({}) should be a directory".format(vocab_path))
+            return
         vocab_file = os.path.join(vocab_path, VOCAB_NAME)
         merge_file = os.path.join(vocab_path, MERGES_NAME)
         json.dump(self.encoder, vocab_file)
@@ -277,3 +280,4 @@ class OpenAIGPTTokenizer(object):
                     index = token_index
                 writer.write(bpe_tokens + u'\n')
                 index += 1
+        return vocab_file, merge_file