Function Differences with torch.cross

class torch.cross(
    input,
    other,
    dim=-1,
    out=None
)

class mindspore.numpy.cross(
    a,
    b,
    axisa=-1,
    axisb=-1,
    axisc=-1,
    axis=None
)

import mindspore.numpy as np
import torch

# MindSpore
x = np.array([[1,2,3], [4,5,6]])
y = np.array([[4,5,6], [1,2,3]])
output = np.cross(x, y)
print(output)
# [[-3  6 -3]
# [ 3 -6  3]]
output = np.cross(x, y, axisc=0)
print(output)
# [[-3  3]
# [ 6 -6]
# [-3  3]]
x = np.array([[1,2], [4,5]])
y = np.array([[4,5], [1,2]])
print(np.cross(x, y))
# Tensor(shape=[2], dtype=Int32, value= [-3,  3])

# PyTorch
a = torch.tensor([[1,2,3], [4,5,6]], dtype=torch.int8)
b = torch.tensor([[4,5,6], [1,2,3]], dtype=torch.int8)
print(torch.cross(a, b, dim=1))
# tensor([[-3,  6, -3],
#         [ 3, -6,  3]], dtype=torch.int8)
print(torch.cross(a, b))
# tensor([[-3,  6, -3],
#         [ 3, -6,  3]], dtype=torch.int8)
a = torch.tensor([[1,2], [4,5]], dtype=torch.int8)
b = torch.tensor([[4,5], [1,2]], dtype=torch.int8)
print(torch.cross(a, b))
# RuntimeError: no dimension of size 3 in input