bug fix. Train and test runs

Author: MzxlM

logHelper.py

@@ -13,7 +13,8 @@ def setup_logger(logger_name, output_file=None):
     formatter = logging.Formatter('[%(asctime)s.%(msecs)03d] %(message)s', datefmt='%H:%M:%S')
 
     # Always log to the default "logging.txt" file
-    default_fh = logging.FileHandler(global_logging_txt)
+    # file_handler = logging.FileHandler(filename='your_log_file.log', encoding='utf-8')
+    default_fh = logging.FileHandler(filename=global_logging_txt, encoding='utf-8')
     default_fh.setLevel(logging.INFO)
     default_fh.setFormatter(formatter)
     logger.addHandler(default_fh)

train.py

@@ -125,6 +125,15 @@ def __len__(self):
         return size
 
 
+def format_progress_message(iteration, total):
+
+    progress_percentage = (iteration / total) * 100
+    progress_bar_length = int(50 * iteration // total)
+    progress_bar = '#' * progress_bar_length + '-' * (50 - progress_bar_length)
+
+    return f"{progress_percentage:3.0f}%|{progress_bar}| {iteration}/{total}"
+
+
 # if __name__ == "__main__":
 def train(batch_size=4, max_epochs=200, base_lr=0.01, seed=1234, n_gpu=1, img_size=224, model_name='swinunet'):
     if os.environ.get('current_fold') is None:
@@ -282,7 +291,8 @@ def train(batch_size=4, max_epochs=200, base_lr=0.01, seed=1234, n_gpu=1, img_si
                 param_group['lr'] = lr_
 
             iter_num = iter_num + 1
-
+        progress_message = format_progress_message(epoch_num + 1, max_epoch)
+        logger.info(progress_message)
             # logger.info('iteration %d : loss : %f' % (iter_num, loss.item()))
 
         save_interval = 5

web.py

@@ -575,7 +575,7 @@ def run_test():
     # TODO convert it from command line to python code
     # threading.Thread(target=run_command_async, args=(complete_command,)).start()
 
-    test_process = Process(target=run_command_async, args="test")
+    test_process = Process(target=run_command_async, args=("test", 4, 2, 0.01))
     test_process.start()
     test_process.join()
     print("Test process started")
@@ -598,69 +598,6 @@ def run_test():
                 data_json = json.load(f)
             label_num = len(data_json['labels'])
 
-            # if os.environ['MODEL_NAME'] == 'nnunet3d':
-            #     if file_ext in ['.gz', '.nrrd', '.mha', '.nii']:
-            #         img = sitk.ReadImage(img_path)
-            #         img = sitk.GetArrayFromImage(img)
-            #
-            #         pred = sitk.ReadImage(prediction_path)
-            #         pred = sitk.GetArrayFromImage(pred)
-            #         ground_truth = sitk.ReadImage(ground_truth_path)
-            #         ground_truth = sitk.GetArrayFromImage(ground_truth)
-            #     else:
-            #         return jsonify({'status': 'Please use png, bmp, tif, nii.gz, nrrd or mha format.'})
-            #
-            #     half_layer = int(np.argmax(ground_truth.sum(axis=(1, 2))))
-            #     img = img[half_layer]
-            #     img = np.repeat(img[:, :, np.newaxis], 3, axis=2)
-            #
-            #     mask_result = np.zeros_like(img)
-            #     img_with_GT = img.copy()
-            #
-            #     each_metric = []
-            #     for i in range(1, label_num):
-            #         each_metric.append(calculate_metric_percase(pred == i, ground_truth == i))
-            #
-            #         mask = np.where(pred[half_layer] == i, 1, 0).astype(np.uint8)
-            #         colored_mask = np.zeros_like(img)
-            #         colored_mask[mask == 1] = colors[i - 1]
-            #         img = cv2.addWeighted(img, 1, colored_mask, 0.5, 0)
-            #
-            #         gt_mask = np.where(ground_truth[half_layer] == i, 1, 0).astype(np.uint8)
-            #         colored_mask_gt = np.zeros_like(img_with_GT)
-            #         colored_mask_gt[gt_mask == 1] = colors[i - 1]
-            #         img_with_GT = cv2.addWeighted(img_with_GT, 1, colored_mask_gt, 0.5, 0)
-            #
-            #         mask_result[mask == 1] = colors[i - 1]
-            #
-            #     metric_list.append(each_metric)
-            #
-            #     img_with_mask_save_path = os.path.join(os.environ['nnUNet_results'], os.environ['MODEL_NAME'],
-            #                                            os.environ['current_dataset'], nnUNetPlans,
-            #                                            'visualization_result', test_img_name + '.png')
-            #     os.makedirs(
-            #         os.path.join(os.environ['nnUNet_results'], os.environ['MODEL_NAME'], os.environ['current_dataset'],
-            #                      nnUNetPlans, 'visualization_result'), exist_ok=True)
-            #     cv2.imwrite(img_with_mask_save_path, img)
-            #
-            #     img_with_GT_save_path = os.path.join(os.environ['nnUNet_results'], os.environ['MODEL_NAME'],
-            #                                          os.environ['current_dataset'], nnUNetPlans, 'GT_result',
-            #                                          test_img_name + '.png')
-            #     os.makedirs(
-            #         os.path.join(os.environ['nnUNet_results'], os.environ['MODEL_NAME'], os.environ['current_dataset'],
-            #                      nnUNetPlans, 'GT_result'), exist_ok=True)
-            #     cv2.imwrite(img_with_GT_save_path, img_with_GT)
-            #
-            #     mask_save_path = os.path.join(os.environ['nnUNet_results'], os.environ['MODEL_NAME'],
-            #                                   os.environ['current_dataset'], nnUNetPlans, 'mask_result',
-            #                                   test_img_name + '.png')
-            #     os.makedirs(
-            #         os.path.join(os.environ['nnUNet_results'], os.environ['MODEL_NAME'], os.environ['current_dataset'],
-            #                      nnUNetPlans, 'mask_result'), exist_ok=True)
-            #     cv2.imwrite(mask_save_path, mask_result)
-
-            # else:
-
             if file_ext in ['.png', '.bmp', '.tif']:
 
                 img = cv2.imread(img_path, cv2.IMREAD_UNCHANGED)