Function Differences with torch.float_power

The following mapping relationships can be found in this file.

PyTorch APIs	MindSpore APIs
torch.float_power	mindspore.ops.float_power
torch.Tensor.float_power	mindspore.Tensor.float_power

torch.float_power

torch.float_power(input, exponent, *, out=None) -> Tensor

For more information, see torch.float_power.

mindspore.ops.float_power

mindspore.ops.float_power(input, exponent)

For more information, see mindspore.ops.float_power.

Differences

PyTorch: Raise the input tensor to double precision to calculate exponential powers. If neither input is complex, a torch.float64 tensor is returned, and if one or more inputs is complex, a torch.complex128 tensor is returned.

MindSpore: If the inputs are all real numbers, MindSpore API implements the same functionality as PyTorch, and only the parameter names are different. Currently, MindSpore does not support computation with complex numbers.

Categories	Subcategories	PyTorch	MindSpore	Differences
Parameter	Parameter 1	input	input	The function is the same
	Parameter 2	exponent	exponent	The function is the same
	Parameter 3	out	-	MindSpore does not have this Parameter

Code Example

When the input is a real number type, the functions of the two APIs are the same, and the usage is the same.

import numpy as np
input_np = np.array([2., 3., 4.], np.float32)
# PyTorch
import torch
input = torch.from_numpy(input_np)
out_torch = torch.float_power(input, 2.)
print(out_torch.detach().numpy())
# [ 4.  9. 16.]

# MindSpore
import mindspore
from mindspore import Tensor, ops
x = Tensor(input_np)
output = ops.float_power(x, 2.)
print(output.asnumpy())
# [ 4.  9. 16.]