In this part, we preprocess image data.
Your job is to create a tensor images_pt from image_list that has shape (1000,3,224,224) and datatype float64.
The data range is from -1 to 1.
Hint: If a is a PIL object, then you can use a.resize to resize it.
Print images_pt.shape.
Print images_pt.dtype.
Print images_pt[5].
### WRITE YOUR SOLUTION HERE ###
images_pt_list = []
for image in image_list:
image = image.resize((224,224)) # Resize the image
image_np = np.array(image) # Convert to numpy array
image_pt = torch.from_numpy(image_np) # Convert to pytorch tensor
image_pt = image_pt.permute(2,0,1) # Permute the dimension
image_pt = image_pt / 255 # Normalize value between 0 and 1
image_pt = image_pt * 2 - 1 # Normalize value between -1 and 1
images_pt_list.append(image_pt)
images_pt = torch.stack(images_pt_list)
print(images_pt.shape)
print(images_pt.dtype)
print(images_pt[5])
""" END OF THIS PART """