PyTorch .expm1()

In PyTorch, the .expm1() function takes in an input tensor and returns a tensor where each element is the exponential of the input element minus 1. It provides greater precision than performing $$exp(x) - 1$$ directly, especially for small values of x. The function computes each element as:

$$y_i= e^{x_i}-1$$

Syntax

torch.expm1(input, *, out=None) → Tensor

Or,

torch.special.expm1(input, *, out=None) → Tensor

Parameters:

• input: The input tensor.
• out (Optional): A tensor to store the output. If provided, the result is written to this tensor.

Return value:

It returns a new tensor of the same shape as the input, containing the computed values for each corresponding element.

Example

This example demonstrates the usage of the .expm1() function:

import torch

# Define a tensor
x = torch.tensor([0, math.log(2.)])

result = torch.special.expm1(x)

print(result)

Copy to clipboard
Copy to clipboard

The output of this code is:

tensor([ 0.,  1.])

Copy to clipboard
Copy to clipboard