QLinearSoftmaxFloatmul< TT, INP_H, INP_W, INP_W_PAD > Class Template Reference

Vector implementation using fastexp2 method, float multiplication for exp estimation QLinearSoftmaxFloatmul<10,10,16> takes 6410 cycles requires INP_W_PAD%16=0.

#include <qlinearsoftmax.h>

Public Member Functions
	QLinearSoftmaxFloatmul (float x_scale, float y_scale, TT x_zero, TT y_zero)
void	filter (input_window< TT > *in, output_stream< TT > *out)

Static Public Member Functions
static void	registerKernelClass ()

---

The documentation for this class was generated from the following files:

• design/aie_src/qlinearsoftmax.h
• design/aie_src/qlinearsoftmax.cc