DynamicQuantize¶

General¶

DynamicQuantize operation converts a f32 tensor to a quantized (s8 or u8) tensor. It supports both per-tensor and per-channel asymmetric linear quantization. The target quantized data type is specified via the data type of dst logical tensor. Rounding mode is library-implementation defined.

For per-tensor quantization

\[dst = round(src/scales + zps)\]

For per-channel quantization, taking channel axis = 1 as an example:

\[{dst}_{\cdots,i,\cdots,\cdots} = round(src_{\cdots,i,\cdots,\cdots}/scales_i + zps_i),i\in [0,channelNum-1]\]

Operation attributes¶

Attribute Name	Description	Value Type	Supported Values	Required or Optional
qtype	Specifies which de-quantization type is used.	string	`per_tensor` (default), `per_channel`	Optional
axis	Specifies dimension on which per-channel de-quantization is applied.	s64	A s64 value in the range of [-r, r-1] where r = rank(src), `1` by default. Negative value means counting the dimension backwards from the end.	Optional

Execution arguments¶

The inputs and outputs must be provided according to below index order when constructing an operation.

Inputs¶

Index	Argument Name	Required or Optional
0	`src`	Required
1	`scales`	Required
2	`zps`	Optional

Note

scales is a f32 1D tensor to be applied to the quantization formula. For qtype = per-tensor, there should be only one element in the scales tensor. For qtype = per-channel, the element number should be equal to the element number of src tensor along the dimension axis.

Note

zps is a 1D tensor with offset values that map to zero. For qtype = per-tensor, there should be only one element in the zps tensor. For qtype = per-channel, the element number should be equal to the element number of input tensor along the dimension axis. If not specified, the library can assume the operator is symmetric quantization and perform kernel optimization accordingly.

Outputs¶

Index	Argument Name	Required or Optional
0	`dst`	Required

Supported data types¶

DynamicQuantize operation supports the following data type combinations.

Src	Scales	Zps	Dst
f32	f32	s8, u8, s32	s8
f32	f32	s8, u8, s32	u8

DynamicDequantize Elu