pytorch矩阵乘法

torch.mm(mat1, mat2) performs a matrix multiplication of mat1 and mat2

?

1 2 3 4 5 6 7 8 9

a = torch.randint(0, 5, (2, 3))   # tensor([[3, 3, 2],                                   #         [2, 2, 2]])                                     b = torch.randint(0, 6, (3, 1))   # tensor([[1],                                   #         [4],                                   #         [5]])   torch.mm(a, b)   # tensor([[11],                  #        [17]])

torch.mul(input, other) multiplies each element of the 'input' with the scalar in 'other' and returns a new resulting tensor.

?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

a = torch.tensor([[1, 2, 3],                   [4, 5, 6]]) b = torch.tensor([1, 2, 3]) c = torch.tensor([[1],                   [2],                   [3]])   torch.mul(a, 10)   # tensor([[10, 20, 30],                    #         [40, 50, 60]])   torch.mul(a, b)    # tensor([[ 1,  4,  9],                    #         [ 4, 10, 18]])   torch.mul(b, c)    # tensor([[1, 2, 3],                    #         [2, 4, 6],                    #         [3, 6, 9]])