diff --git a/ACL_PyTorch/built-in/nlp/Bert_Base_Uncased_for_Pytorch/bert_base_get_info.py b/ACL_PyTorch/built-in/nlp/Bert_Base_Uncased_for_Pytorch/bert_base_get_info.py
index ce1b597ce2293a39d5e12d66838f7571e6f648a6..3d9146da31a0d9c87be5c6bbb7a4d3478ddb75f9 100644
--- a/ACL_PyTorch/built-in/nlp/Bert_Base_Uncased_for_Pytorch/bert_base_get_info.py
+++ b/ACL_PyTorch/built-in/nlp/Bert_Base_Uncased_for_Pytorch/bert_base_get_info.py
@@ -13,12 +13,20 @@
 # limitations under the License.
 # ============================================================================
 import os
+import argparse
+
 
-test_num = 10649
 base_path = './bert_bin/'
+# three inputs in each eval
+test_num = len(os.listdir(base_path)) // 3
+parser = argparse.ArgumentParser(description='manual to this script')
+parser.add_argument('--batchsize', type=int, default=8)
+args = parser.parse_args()
+batchsize = args.batchsize
+real_test = test_num // batchsize * batchsize
 
 with open('./bert_base_uncased.info', 'w') as f:
-    for i in range(test_num):
+    for i in range(real_test):
         ids_name = base_path + 'input_ids_{}.bin'.format(i)
         segment_name = base_path + 'segment_ids_{}.bin'.format(i)
         mask_name = base_path + 'input_mask_{}.bin'.format(i)
diff --git a/ACL_PyTorch/built-in/nlp/Bert_Base_Uncased_for_Pytorch/bert_postprocess_data.py b/ACL_PyTorch/built-in/nlp/Bert_Base_Uncased_for_Pytorch/bert_postprocess_data.py
index 11e0ada13c7998fa4fa3ed11dffa9be3d8c54b3b..ebf52e30e9e945dcd24efda6565854009114287c 100644
--- a/ACL_PyTorch/built-in/nlp/Bert_Base_Uncased_for_Pytorch/bert_postprocess_data.py
+++ b/ACL_PyTorch/built-in/nlp/Bert_Base_Uncased_for_Pytorch/bert_postprocess_data.py
@@ -395,8 +395,8 @@ def main():
                         required=True,
                         help="NPU benchmark infer result path")
     args = parser.parse_args()
-    test_num = 10649
     npu_path = args.npu_result
+    test_num = len(os.listdir(npu_path)) // 2
     tokenizer = BertTokenizer(args.vocab_file, do_lower_case=args.do_lower_case, max_len=512) # for bert large
     eval_examples = read_squad_examples(
         input_file=args.predict_file, is_training=False)
@@ -432,7 +432,6 @@ def main():
         all_results.append(RawResult(unique_id=unique_id,
                                      start_logits=start_logits,
                                      end_logits=end_logits))
-        print(" [INFO] i == ", i)
     time_to_infer = time.time() - infer_start
     output_prediction_file = os.path.join("./", "predictions.json")
     output_nbest_file = os.path.join("./", "nbest_predictions.json")
@@ -442,7 +441,8 @@ def main():
         f.write(json.dumps(answers, indent=4) + "\n")
     with open(output_nbest_file, "w") as f:
         f.write(json.dumps(nbest_answers, indent=4) + "\n")
-    
+    print("Completed")
+
 if __name__ == '__main__':
    main() 
     
\ No newline at end of file