odp_ml_quant_param_t Struct Reference

Quantization parameters. More...

#include <ml_types.h>

Data Fields
odp_ml_data_type_t	type
	Type of quantization scale value. More...
float	scale_fp32
	Quantization scale.
int32_t	zerop_i32
	Quantization zero point.

Detailed Description

Quantization parameters.

These parameters are used to convert between floating point and integer data. Scale and zerop values can be used directly with the odp_ml_fp32_from_*() and odp_ml_fp32_to_*() functions.

Definition at line 521 of file api/spec/ml_types.h.

Field Documentation

type

odp_ml_data_type_t odp_ml_quant_param_t::type

Type of quantization scale value.

Valid quantization scale and zero point values are provided, if set to something other than ODP_ML_DATA_TYPE_NONE. Allowed types are ODP_ML_DATA_TYPE_NONE and ODP_ML_DATA_TYPE_FP32.

Definition at line 529 of file api/spec/ml_types.h.

---

The documentation for this struct was generated from the following file:

• api/spec/ml_types.h