diff --git "a/MIR_Task_pytorch/\357\254\202ash guided non-\357\254\202ash image denoising/Code/CUdata_npy_denoise25.py" "b/MIR_Task_pytorch/\357\254\202ash guided non-\357\254\202ash image denoising/Code/CUdata_npy_denoise25.py"
new file mode 100644
index 0000000..dd980a3
--- /dev/null
+++ "b/MIR_Task_pytorch/\357\254\202ash guided non-\357\254\202ash image denoising/Code/CUdata_npy_denoise25.py"	
@@ -0,0 +1,43 @@
+import torch.utils.data as data
+from glob import glob
+import matplotlib.pyplot as plt
+from torchvision import transforms
+import cv2
+from PIL import Image
+import random
+import os
+import numpy as np
+import torch
+
+class cudataset(data.Dataset):
+    def __init__(self):
+        super(cudataset, self).__init__()
+        self.depth = np.load('trainset_denoise/data_x_25.npy')#(batch,height,width,c)
+        #print(self.depth.shape)
+        self.depth = np.transpose(self.depth, (0, 3, 1, 2))
+        self.depth_t = torch.from_numpy(self.depth)
+
+        self.gt = np.load('trainset_denoise/label_25.npy')  # (batch,height,width,c)
+        self.gt = np.transpose(self.gt, (0, 3, 1, 2))
+        self.gt_t = torch.from_numpy(self.gt)
+
+        self.rgb = np.load('trainset_denoise/data_y_25.npy')  # (batch,height,width,c)
+        self.rgb = np.transpose(self.rgb, (0, 3, 1, 2))
+        self.rgb_t = torch.from_numpy(self.rgb)
+
+    def __getitem__(self, item):
+        img_depth = self.depth_t[item]
+        img_gt = self.gt_t[item]
+        img_rgb = self.rgb_t[item]
+
+        return (img_depth, img_gt,img_rgb)
+
+    def __len__(self):
+        return len(self.depth)
+
+if __name__ =='__main__':
+    dataset=cudataset()
+    dataloader=data.DataLoader(dataset,batch_size=1)
+    for b1,(img_L,img_H,img_RGB) in enumerate(dataloader):
+        print(b1)
+        print(img_L.shape,img_H.shape,img_RGB.shape)
\ No newline at end of file
diff --git "a/MIR_Task_pytorch/\357\254\202ash guided non-\357\254\202ash image denoising/Code/testCUnoise.py" "b/MIR_Task_pytorch/\357\254\202ash guided non-\357\254\202ash image denoising/Code/testCUnoise.py"
new file mode 100644
index 0000000..dd56fb7
--- /dev/null
+++ "b/MIR_Task_pytorch/\357\254\202ash guided non-\357\254\202ash image denoising/Code/testCUnoise.py"	
@@ -0,0 +1,49 @@
+import torch.utils.data as data
+from glob import glob
+import matplotlib.pyplot as plt
+from torchvision import transforms
+import cv2
+from PIL import Image
+import random
+import os
+import numpy as np
+import torch
+
+
+class cudatatest(data.Dataset):
+    def __init__(self):
+        super(cudatatest, self).__init__()
+        # self.depth = np.load('testset/data_d1.npy',allow_pickle=True)#(batch,height,width,c)
+        self.depth = np.load('noise_testset/data_n.npy', allow_pickle=True)  # (batch,height,width,c)
+        self.depth = np.transpose(self.depth, (0, 3, 1, 2))
+        self.depth_t = torch.from_numpy(self.depth)
+
+        # self.gt = np.load('testset/data_g1.npy',allow_pickle=True)  # (batch,height,width,c)
+        self.gt = np.load('noise_testset/gt.npy', allow_pickle=True)  # (batch,height,width,c)
+        self.gt = np.transpose(self.gt, (0, 3, 1, 2))
+        self.gt_t = torch.from_numpy(self.gt)
+
+        # self.rgb = np.load('testset/data_c1.npy',allow_pickle=True)  # (batch,height,width,c)
+        self.rgb = np.load('noise_testset/data_f.npy', allow_pickle=True)  # (batch,height,width,c)
+        self.rgb = np.transpose(self.rgb, (0, 3, 1, 2))
+        self.rgb_t = torch.from_numpy(self.rgb)
+
+    def __getitem__(self, item):
+        img_depth = self.depth_t[item]
+        img_gt = self.gt_t[item]
+        img_rgb = self.rgb_t[item]
+
+        return (img_depth, img_gt, img_rgb)
+
+    def __len__(self):
+        return len(self.depth)
+
+
+
+
+if __name__ == '__main__':
+    dataset = cudatatest()
+    dataloader = data.DataLoader(dataset, batch_size=1)
+    for b1, (img_L, img_H, img_RGB) in enumerate(dataloader):
+        print(b1)
+        print(img_L.shape, img_H.shape, img_RGB.shape)
\ No newline at end of file