diff --git a/ppstructure/vqa/eval_re.py b/ppstructure/vqa/eval_re.py
index 12bb9cabdb8b4d6482a121ca3b73089b3d0244ff..68c27bad8a8e236fc16ffad21acefe7a55fde561 100644
--- a/ppstructure/vqa/eval_re.py
+++ b/ppstructure/vqa/eval_re.py
@@ -24,7 +24,7 @@ import paddle
 from paddlenlp.transformers import LayoutXLMTokenizer, LayoutXLMModel, LayoutXLMForRelationExtraction
 
 from xfun import XFUNDataset
-from utils import parse_args, get_bio_label_maps, print_arguments
+from vqa_utils import parse_args, get_bio_label_maps, print_arguments
 from data_collator import DataCollator
 from metric import re_score
 
diff --git a/ppstructure/vqa/eval_ser.py b/ppstructure/vqa/eval_ser.py
index 52eeb8a1da82d6e7dcdc726c1a77fd9ab18f0608..95f428c721e103748d5cd722ff0e1d2bf0f09e52 100644
--- a/ppstructure/vqa/eval_ser.py
+++ b/ppstructure/vqa/eval_ser.py
@@ -33,7 +33,7 @@ from paddlenlp.transformers import LayoutLMModel, LayoutLMTokenizer, LayoutLMFor
 
 from xfun import XFUNDataset
 from losses import SERLoss
-from utils import parse_args, get_bio_label_maps, print_arguments
+from vqa_utils import parse_args, get_bio_label_maps, print_arguments
 
 from ppocr.utils.logging import get_logger
 
diff --git a/ppstructure/vqa/infer_re.py b/ppstructure/vqa/infer_re.py
index 7937700a78c28eff19a78dc5ddc8ccef82cc0148..b6774e77befe6ba8954d5f552bcade86cb44e644 100644
--- a/ppstructure/vqa/infer_re.py
+++ b/ppstructure/vqa/infer_re.py
@@ -15,7 +15,7 @@ import paddle
 from paddlenlp.transformers import LayoutXLMTokenizer, LayoutXLMModel, LayoutXLMForRelationExtraction
 
 from xfun import XFUNDataset
-from utils import parse_args, get_bio_label_maps, draw_re_results
+from vqa_utils import parse_args, get_bio_label_maps, draw_re_results
 from data_collator import DataCollator
 
 from ppocr.utils.logging import get_logger
diff --git a/ppstructure/vqa/infer_ser.py b/ppstructure/vqa/infer_ser.py
index 7994b5449af97988e9ad458f4047b59b81f72560..f5fb581fa7e7613216d2e4feb8e39ed8c2541dc9 100644
--- a/ppstructure/vqa/infer_ser.py
+++ b/ppstructure/vqa/infer_ser.py
@@ -14,6 +14,10 @@
 
 import os
 import sys
+
+__dir__ = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(__dir__)
+
 import json
 import cv2
 import numpy as np
@@ -22,7 +26,7 @@ from copy import deepcopy
 import paddle
 
 # relative reference
-from utils import parse_args, get_image_file_list, draw_ser_results, get_bio_label_maps
+from vqa_utils import parse_args, get_image_file_list, draw_ser_results, get_bio_label_maps
 from paddlenlp.transformers import LayoutXLMModel, LayoutXLMTokenizer, LayoutXLMForTokenClassification
 from paddlenlp.transformers import LayoutLMModel, LayoutLMTokenizer, LayoutLMForTokenClassification
 
diff --git a/ppstructure/vqa/infer_ser_e2e.py b/ppstructure/vqa/infer_ser_e2e.py
index 6bb0247501bc98ea709d3ad47c284268b3d9503b..33fe4dbb5e809388b135ee467d7e7c230f0eabcc 100644
--- a/ppstructure/vqa/infer_ser_e2e.py
+++ b/ppstructure/vqa/infer_ser_e2e.py
@@ -14,6 +14,10 @@
 
 import os
 import sys
+
+__dir__ = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(__dir__)
+
 import json
 import cv2
 import numpy as np
@@ -25,9 +29,9 @@ from paddlenlp.transformers import LayoutXLMModel, LayoutXLMTokenizer, LayoutXLM
 from paddlenlp.transformers import LayoutLMModel, LayoutLMTokenizer, LayoutLMForTokenClassification
 
 # relative reference
-from utils import parse_args, get_image_file_list, draw_ser_results, get_bio_label_maps
+from vqa_utils import parse_args, get_image_file_list, draw_ser_results, get_bio_label_maps
 
-from utils import pad_sentences, split_page, preprocess, postprocess, merge_preds_list_with_ocr_info
+from vqa_utils import pad_sentences, split_page, preprocess, postprocess, merge_preds_list_with_ocr_info
 
 MODELS = {
     'LayoutXLM':
diff --git a/ppstructure/vqa/infer_ser_re_e2e.py b/ppstructure/vqa/infer_ser_re_e2e.py
index 32d8850a16eebee7e43fd58cbaa604fc2bc00b7c..e24c9f69e0836d64fbe67609623e4b6409f7658c 100644
--- a/ppstructure/vqa/infer_ser_re_e2e.py
+++ b/ppstructure/vqa/infer_ser_re_e2e.py
@@ -24,7 +24,7 @@ import paddle
 from paddlenlp.transformers import LayoutXLMModel, LayoutXLMTokenizer, LayoutXLMForRelationExtraction
 
 # relative reference
-from utils import parse_args, get_image_file_list, draw_re_results
+from vqa_utils import parse_args, get_image_file_list, draw_re_results
 from infer_ser_e2e import SerPredictor
 
 
diff --git a/ppstructure/vqa/train_re.py b/ppstructure/vqa/train_re.py
index 47d694678013295a1c664a7bdb6a7fe13a0b36a5..eeff2bfbbe466b29b8b46e83058e2199fd5cafed 100644
--- a/ppstructure/vqa/train_re.py
+++ b/ppstructure/vqa/train_re.py
@@ -27,7 +27,7 @@ import paddle
 from paddlenlp.transformers import LayoutXLMTokenizer, LayoutXLMModel, LayoutXLMForRelationExtraction
 
 from xfun import XFUNDataset
-from utils import parse_args, get_bio_label_maps, print_arguments, set_seed
+from vqa_utils import parse_args, get_bio_label_maps, print_arguments, set_seed
 from data_collator import DataCollator
 from eval_re import evaluate
 
diff --git a/ppstructure/vqa/train_ser.py b/ppstructure/vqa/train_ser.py
index 2670ef9eeaf75cc1c9bb7d8f41d6f76d4300e597..226172050e9a5ea3b7c6534444ef24278de07043 100644
--- a/ppstructure/vqa/train_ser.py
+++ b/ppstructure/vqa/train_ser.py
@@ -32,7 +32,7 @@ from paddlenlp.transformers import LayoutXLMModel, LayoutXLMTokenizer, LayoutXLM
 from paddlenlp.transformers import LayoutLMModel, LayoutLMTokenizer, LayoutLMForTokenClassification
 
 from xfun import XFUNDataset
-from utils import parse_args, get_bio_label_maps, print_arguments, set_seed
+from vqa_utils import parse_args, get_bio_label_maps, print_arguments, set_seed
 from eval_ser import evaluate
 from losses import SERLoss
 from ppocr.utils.logging import get_logger
diff --git a/ppstructure/vqa/utils.py b/ppstructure/vqa/vqa_utils.py
similarity index 99%
rename from ppstructure/vqa/utils.py
rename to ppstructure/vqa/vqa_utils.py
index b9f2edc860b1ce48c22bf602cef48466c357834f..8f5ff2ef46ed2ec4027979d6411459059f15b7fa 100644
--- a/ppstructure/vqa/utils.py
+++ b/ppstructure/vqa/vqa_utils.py
@@ -366,9 +366,9 @@ def parse_args():
     parser.add_argument("--max_seq_length", default=512, type=int,)
     parser.add_argument("--evaluate_during_training", action="store_true",)
     parser.add_argument("--num_workers", default=8, type=int,)
-    parser.add_argument("--per_gpu_train_batch_size", default=8,
+    parser.add_argument("--per_gpu_train_batch_size", default=1,
                         type=int, help="Batch size per GPU/CPU for training.",)
-    parser.add_argument("--per_gpu_eval_batch_size", default=8,
+    parser.add_argument("--per_gpu_eval_batch_size", default=1,
                         type=int, help="Batch size per GPU/CPU for eval.",)
     parser.add_argument("--learning_rate", default=5e-5,
                         type=float, help="The initial learning rate for Adam.",)