Module quantization

Tensor quantization module.

Structs

CalibrationRange

The observed input calibration range.

QParams

The quantization tensor data parameters.

QuantScheme

Describes a quantization scheme/configuration.

QuantizationParametersPrimitive

The quantization parameters primitive.

QuantizedBytes

Quantized data bytes representation.

SymmetricQuantization

Symmetric quantization scheme.

Enums

Calibration

Calibration method used to compute the quantization range mapping.

QuantAccPrecision

Quantization accumulator precision. This is the precision to used when accumulating values while executing algorithms such as matmul.

QuantInputType

Data type used to represent quantized values.

QuantLevel

Level or granularity of quantization.

QuantMode

Strategy used to quantize values.

QuantPropagation

Specify if the output of an operation is quantized using the scheme of the input or returned unquantized.

QuantizationStrategy

Quantization strategy.

Traits

QTensorPrimitive

Quantized tensor primitive.

Quantization

Quantization scheme to convert elements of a higher precision data type E to a lower precision data type Q and vice-versa.

Functions

pack_i8s_to_u32s

Pack signed 8-bit integer values into a sequence of unsigned 32-bit integers.

unpack_u32s_to_i8s

Unpack 32-bit unsigned integer values into a sequence of signed 8-bit integers.

Type Aliases

QuantizationParameters

The tensor quantization parameters.