QuantizeLinearKernels

Collaboration diagram for QuantizeLinearKernels:

Classes
class	QuantizeLinearScalar< TT, INP_H, INP_W, OUT_W >
	Scalar implementation, requires INP_W <= OUT_W, QuantizeLinearScalar<1*1*28*28> takes 92121 cycles. More...
class	QuantizeLinear< TT, INP_H, INP_W, OUT_W >
	Vector implementation, requires INP_W%4==0, OUT_W%16==0, INP_W <= OUT_W, QuantizeLinear<1*1*28*28> takes 1977 cycles. More...
class	QuantizeLinearFmul< TT, INP_H, INP_W, OUT_W >
	Vector implementation, requires INP_W%4==0, OUT_W%16==0, INP_W <= OUT_W, QuantizeLinearFmul<1*1*28*28> takes 1534 cycles,. More...
class	QuantizeLinearFmulStream< TT, INP_H, INP_W, OUT_W >
	Vector stream implementation, requires INP_W%4==0, OUT_W%16==0, INP_W <= OUT_W, QuantizeLinearFmulStream<1*1*28*28> takes 1669 cycles,. More...

Detailed Description

https://github.com/onnx/onnx/blob/main/docs/Operators.md#QuantizeLinear y = saturate ((x / y_scale) + y_zero)