我正在尝试创建一个3dcnn分类模型，该模型将10个（320，864）大小的灰度图像作为输入，并输出一个输出。运行训练时，弹出以下错误：
运行时间错误：给定groups=1，权重大小为[16，1，3，3，3]，预期输入[2，10，1，320，864]有1个通道，但实际上得到了10个通道
下面的代码是加载输入的代码。我想让输入形状[2，1，10，320，864]。我应该在加载输入的代码中修改什么？

def __getitem__(self, idx):
  video_idx = idx // 260
  frame_idx = idx % 260 + 41
  video_dir = self.video_dirs[video_idx]
  frames = []

  # Get all files in the directory
  all_files = os.listdir(video_dir)

  # Select only .jpg files
  jpg_files = [file for file in all_files if file.endswith('.jpg')]

  # Extract the number from the file name and sort
  numbered_files = sorted(jpg_files, key=lambda x: int(re.findall(r'\d+', x)[-1]))

  for i in range(frame_idx, frame_idx + 10):
      # Get the file with the corresponding number
      frame_file = numbered_files[i-1]  # -1 because indexing starts from 0
      frame_path = os.path.join(video_dir, frame_file)
      print(f"Loading image from {frame_path}")
      frame = cv2.imread(frame_path)
      if frame is None:
          raise ValueError(f"Could not load image at {frame_path}")
      frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)  # Convert to grayscale
      frame = frame.reshape(1, frame.shape[0], frame.shape[1])  # Add channel dimension
      frame = torch.tensor(frame, dtype=torch.float32)
      print(frame.shape)
      frames.append(frame)

  frames = torch.stack(frames)  # Stack frames along the channel dimension
  label = self.labels[idx]
  return frames, label

字符串