modidy time statistics

fluid/PaddleNLP/deep_attention_matching_net/train_and_evaluate.py

@@ -391,7 +391,7 @@ def train(args):
             global_step, last_cost = train_with_feed(global_step)
         train_time += time.time() - begin_time
         print("Pass {0}, pass_time_cost {1}"
-          .format(epoch, "%2.2f sec" % time.time() -begin_time ))
+          .format(epoch, "%2.2f sec" % (time.time() - begin_time)))
     # For internal continuous evaluation
     if "CE_MODE_X" in os.environ:
         print("kpis	train_cost	%f" % last_cost)