torch.max np.max

import torch
import numpy as np
np.random.seed(5)
torch.manual_seed(6)

a = torch.randn(3,3)
print(a)

输出:
tensor([[-0.4868, -0.6038, -0.5581],
        [ 0.6675, -0.1974,  1.9428],
        [-1.4017, -0.7626,  0.6312]])

b = torch.max(a,0)
print(b)

输出:
torch.return_types.max(
values=tensor([ 0.6675, -0.1974,  1.9428]),
indices=tensor([1, 1, 1]))


c = np.random.randn(3,3)
print(c)

输出:
[[-0.32986996 -1.19276461 -0.20487651]
 [-0.35882895  0.6034716  -1.66478853]
 [-0.70017904  1.15139101  1.85733101]]

d = np.max(c, 1)
print(d)

输出:
[-0.20487651  0.6034716   1.85733101]

①torch.max 同时返回最大值和索引，②np.max只返回最大值，③np.max等价于np.amax