Versioned name: SoftMaxBackprop-1

Category: Activation

Short description: Reference

Detailed description: Reference


  • axis

    • Description: axis represents the axis of which the Softmax is calculated.

    • Range of values: [-r, r-1] where r = rank(input)

    • Type: s64

    • Default value: 1

    • Required: no


  • 1: output_delta - the gradient tensor with respect to the output. Required.

    • Type: T

  • 2: result_forward - original output tensor of SoftMax op. Required.

    • Type: T


  • 1: input_delta - the gradient tensor with respect to the input of SoftMax.

    • Type: T


  • T: f32, f16, bf16.

  • Note: Inputs and outputs have the same data type denoted by T. For example, if input is f32 tensor, then all other tensors have f32 data type.