gelu¶

dragon.vm.torch.nn.functional.gelu(
  input,
  approximate='none'
)[source]¶

Apply the gaussian error linear unit to input. [Clevert et.al, 2015].

The GELU function is defined as:

\[\text{GELU}(x) = x\cdot\frac{1}{2}[1 + \text{erf}(x / \sqrt{2})] \]

Parameters:

• input (dragon.vm.torch.Tensor) – The input tensor.
• approximate (str, optional, default='none') – The approximate algorithm.

Returns:

dragon.vm.torch.Tensor – The output tensor.

See also

torch.nn.GELU(…)