mlir::quant Namespace Reference

Namespaces
	detail
	QuantizationFlags
	Enumeration of bit-mapped flags related to quantized types.
	QuantizationTypes

Classes
class	AnyQuantizedType
struct	ExpressedToQuantizedConverter
class	QuantizationDialect
	Defines the 'Quantization' dialect. More...
class	QuantizedType
class	UniformQuantizedPerAxisType
class	UniformQuantizedPerAxisValueConverter
class	UniformQuantizedType
class	UniformQuantizedValueConverter

Functions
UniformQuantizedType	fakeQuantAttrsToType (Location loc, unsigned numBits, double rmin, double rmax, bool narrowRange, Type expressedType, bool isSigned=false)
UniformQuantizedPerAxisType	fakeQuantAttrsToType (Location loc, unsigned numBits, int32_t quantizedDimension, ArrayRef< double > rmins, ArrayRef< double > rmax, bool narrowRange, Type expressedType, bool isSigned=false)
std::unique_ptr< OpPassBase< FuncOp > >	createConvertSimulatedQuantPass ()
std::unique_ptr< OpPassBase< FuncOp > >	createConvertConstPass ()
Attribute	quantizeAttr (Attribute realValue, QuantizedType quantizedElementType, Type &outConvertedType)
Attribute	quantizeAttrUniform (Attribute realValue, UniformQuantizedType quantizedElementType, const UniformQuantizedValueConverter &converter, Type &outConvertedType)

Function Documentation

createConvertConstPass()

std::unique_ptr< OpPassBase< FuncOp > > mlir::quant::createConvertConstPass

(

)

Creates a pass that converts constants followed by a qbarrier to a constant whose value is quantized. This is typically one of the last passes done when lowering to express actual quantized arithmetic in a low level representation. Because it modifies the constant, it is destructive and cannot be undone.

createConvertSimulatedQuantPass()

std::unique_ptr< OpPassBase< FuncOp > > mlir::quant::createConvertSimulatedQuantPass

(

)

Creates a pass that converts quantization simulation operations (i.e. FakeQuant and those like it) to casts into/out of supported QuantizedTypes.

fakeQuantAttrsToType() [1/2]

UniformQuantizedType mlir::quant::fakeQuantAttrsToType	(	Location	loc,
		unsigned	numBits,
		double	rmin,
		double	rmax,
		bool	narrowRange,
		Type	expressedType,
		bool	isSigned = false
	)

Converts per-layer FakeQuant attributes to the corresponding type. In the event that the parameters cannot be converted, returns a nullptr convertible Type and issues an appropriate error. Note that there are multiple variants of a per-layer FakeQuant op, so this function takes the attributes discretely vs taking a reference to the originating op.

fakeQuantAttrsToType() [2/2]

UniformQuantizedPerAxisType mlir::quant::fakeQuantAttrsToType	(	Location	loc,
		unsigned	numBits,
		int32_t	quantizedDimension,
		ArrayRef< double >	rmins,
		ArrayRef< double >	rmax,
		bool	narrowRange,
		Type	expressedType,
		bool	isSigned = false
	)

Converts per-channel FakeQuant attributes to the corresponding type. In the event that the parameters cannot be converted, returns a nullptr convertible Type and issues an appropriate error.

quantizeAttr()

Attribute mlir::quant::quantizeAttr	(	Attribute	realValue,
		QuantizedType	quantizedElementType,
		Type &	outConvertedType
	)

Converts an attribute from a type based on quantizedElementType.getExpressedType() to one based on quantizedElementType.getStorageType(), where quantizedElementType is as from QuantizedType::getQuantizedElementType(). Returns nullptr if the conversion is not supported. On success, stores the converted type in outConvertedType.

Examples:

1. realValue is a primitive value attribute: (realValue: FloatAttr, quantizedElementType: UniformQuantizedType[i8:f32]) -> (IntegerAttr, outConvertedType: i8)
2. realValue is an elements attribute: (realValue: DenseElementsAttr[tensor<2x2xf32>], quantizedElementType: UniformQuantizedType[i8:f32]) -> (DenseElementsAttr[tensor<2x2xi8>], outConvertedType: tensor<2x2xi8>)

Convert an attribute from a type based on quantizedElementType.getExpressedType() to one based on quantizedElementType.getStorageType(). Returns nullptr if the conversion is not supported. On success, stores the converted type in outConvertedType.

quantizeAttrUniform()

Attribute mlir::quant::quantizeAttrUniform	(	Attribute	realValue,
		UniformQuantizedType	quantizedElementType,
		const UniformQuantizedValueConverter &	converter,
		Type &	outConvertedType
	)

Converts an attribute from a type based on quantizedElementType.getExpressedType() to one based on quantizedElementType.getStorageType(), where quantizedElementType is as from QuantizedType::getQuantizedElementType() and casted to an UniformQuantizedType. Returns nullptr if the conversion is not supported. On success, stores the converted type in outConvertedType.

Examples:

1. realValue is a primitive value attribute: (realValue: FloatAttr, quantizedElementType: UniformQuantizedType[i8:f32]) -> (IntegerAttr, outConvertedType: i8)
2. realValue is an elements attribute: (realValue: DenseElementsAttr[tensor<2x2xf32>], quantizedElementType: UniformQuantizedType[i8:f32]) -> (DenseElementsAttr[tensor<2x2xi8>], outConvertedType: tensor<2x2xi8>)

Converts a real expressed Attribute to a corresponding Attribute containing quantized storage values assuming the given uniform quantizedElementType and converter.