Softmax

Description

Performs softmax on the input.

Input

One input:

x: tensor of type float16, float, or double.

Output

One output:

y: tensor of type float16, float, or double.

Attribute

axis: (optional) int, direction for softmax. Defaults to -1. Must be in the range of [ -len(x.shape), len(x.shape)-1].

ONNX Opset Support

Opset v8/v9/v10/v11/v12/v13/v14/v15/v16/v17/v18