Discovering how PyTorch takes 'dimension' in many cases, this case max

torch_max_dimension.py

import torch

sample_tensor = torch.Tensor([
    [1, 2, 3, 4, 5, 6],
    [7, 8, 9, 10, 11, 12],
    [13, 14, 15, 16, 17, 18],
])
sample_tensor = sample_tensor.view(3, 6//2, -1)
#tensor([[[ 1.,  2.],
#         [ 3.,  4.],
#         [ 5.,  6.]],
#
#        [[ 7.,  8.],
#         [ 9., 10.],
#         [11., 12.]],
#
#        [[13., 14.],
#         [15., 16.],
#         [17., 18.]]])

print(sample_tensor.max().values)
print(sample_tensor.max(0).values)
print(sample_tensor.max(1).values)
print(sample_tensor.max(2).values)