我有一个二维Tensor和一个一维Tensor：

import torch
torch.manual_seed(0)

out = torch.randn((16,2))
target = torch.tensor([0, 1, 1, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 1, 1, 0])

对于out的每一行，我想选择由target索引的相应列。因此，我的输出将是(16,1)Tensor。我尝试了下面提到的解决方案：
https://stackoverflow.com/a/58937071
但我得到：

Traceback (most recent call last):
  File "/opt/conda/lib/python3.8/site-packages/IPython/core/interactiveshell.py", line 3369, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "<ipython-input-7-50d103c3b56c>", line 1, in <cell line: 1>
    out.gather(1, target)
RuntimeError: Index tensor must have the same number of dimensions as input tensor

你能帮忙吗？