diff --git a/dataset/cityscapes/filter_dataset.py b/dataset/cityscapes/filter_dataset.py
index 715cc92..29073f3 100644
--- a/dataset/cityscapes/filter_dataset.py
+++ b/dataset/cityscapes/filter_dataset.py
@@ -1,12 +1,12 @@
 # coding: utf-8
 __author__ = 'RocketFlash: https://github.com/RocketFlash'
 
-from tqdm import tqdm
 from scipy.io import loadmat
 from shutil import copyfile, rmtree
 from pathlib import Path
 import os
 from config import data_filtering as cfg
+import argparse
 
 ''' 
     class_label =0: ignore regions (fake humans, e.g. people on posters, reflections etc.)
@@ -41,7 +41,7 @@ def filter_data(mat_file_path, allowed_classes=['pedestrian'],
     
     filtered_file_names = []
     
-    for img_idx in tqdm(range(len(mat)), leave=True, position=0):
+    for img_idx in tqdm(range(len(mat)), position=0):
         img_anno = mat[img_idx][0, 0]
 
         img_name_with_ext = img_anno[1][0]
@@ -62,6 +62,15 @@ def filter_data(mat_file_path, allowed_classes=['pedestrian'],
 
 
 if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--notebook', action='store_true', help='true if run it in jupyter notebook')
+    args = parser.parse_args()
+
+    if args.notebook:
+        from tqdm.notebook import tqdm
+    else:
+        from tqdm import tqdm
+
     print('START DATASET FILTERING')
     if cfg.dataset_type=='citypersons':
         DATASET_ROOT = cfg.dataset_root
diff --git a/dataset/cityscapes/generate_and_filter_dataset.sh b/dataset/cityscapes/generate_and_filter_dataset.sh
index 77f1162..c8391db 100755
--- a/dataset/cityscapes/generate_and_filter_dataset.sh
+++ b/dataset/cityscapes/generate_and_filter_dataset.sh
@@ -1,3 +1,9 @@
 #!/bin/bash
-python generate_dataset.py
-python filter_dataset.py
\ No newline at end of file
+
+if [ ! -z $1 ]; then   
+    python generate_dataset.py --notebook
+    python filter_dataset.py --notebook
+else
+    python generate_dataset.py
+    python filter_dataset.py
+fi
\ No newline at end of file
diff --git a/dataset/cityscapes/generate_dataset.py b/dataset/cityscapes/generate_dataset.py
index be99fca..948a7f4 100644
--- a/dataset/cityscapes/generate_dataset.py
+++ b/dataset/cityscapes/generate_dataset.py
@@ -5,9 +5,9 @@
 import cv2
 import numpy as np
 from pathlib import Path
-from tqdm import tqdm
 from shutil import rmtree
 from config import data_generation as cfg
+import argparse
 
 NAME_TO_ID = {
     'pedestrian': 24
@@ -22,19 +22,19 @@ def generate_object_dataset_cityscapes(annotations_path, images_path, save_dir,
     postfix_instance = 'gtFine_instanceIds'
     obj_id = NAME_TO_ID[object_name]
     
-    for split_dir in tqdm(split_dirs, position=0, leave=True, desc="Total splits"):
+    for split_dir in tqdm(split_dirs, position=0, desc="Total splits"):
         SPLIT_ANNOS = annotations_path / split_dir
         SPLIT_IMAGES = images_path / split_dir
 
         CITY_DIRS = SPLIT_ANNOS.glob('*/')
         CITY_NAMES = [city_path.name for city_path in CITY_DIRS]
 
-        for city_name in tqdm(CITY_NAMES, position=1, leave=True, desc=f'{split_dir:5} split cities: '):
+        for city_name in tqdm(CITY_NAMES, position=1, desc=f'{split_dir:5} split cities: '):
             city_images_path = SPLIT_IMAGES / city_name
             city_annos_path = SPLIT_ANNOS / city_name
             city_images = city_images_path.glob('*.png')
             
-            for city_image in tqdm(list(city_images), leave=True, position=2, desc=f'{city_name:15} city images: '):
+            for city_image in tqdm(list(city_images), position=2, desc=f'{city_name:15} city images: '):
                 mask_label_path = city_annos_path / city_image.name.replace(postfix_image, postfix_label)
                 mask_instance_path = city_annos_path / city_image.name.replace(postfix_image, postfix_instance)
 
@@ -76,7 +76,14 @@ def generate_object_dataset_cityscapes(annotations_path, images_path, save_dir,
                         cv2.imwrite(str(obj_region_image_path), obj_rgba)
 
 if __name__ == '__main__':
-
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--notebook', action='store_true', help='true if run it in jupyter notebook')
+    args = parser.parse_args()
+
+    if args.notebook:
+        from tqdm.notebook import tqdm
+    else:
+        from tqdm import tqdm
     print('START DATASET GENERATION')
     if cfg.dataset_type=='cityscapes':
         DATASET_ROOT = cfg.dataset_root