Hardswish

class torch.ao.nn.quantized.Hardswish(scale, zero_point)

This is the quantized version of Hardswish.

Parameters:

• scale – quantization scale of the output tensor

• zero_point – quantization zero point of the output tensor