MLIR
15.0.0git
|
Represents per-axis (also known as per-channel quantization). More...
#include "mlir/Dialect/Quant/QuantTypes.h"
Public Member Functions | |
ArrayRef< double > | getScales () const |
Gets the quantization scales. More... | |
ArrayRef< int64_t > | getZeroPoints () const |
Gets the storage values corresponding to the real value 0 in the affine equation. More... | |
int32_t | getQuantizedDimension () const |
Specifies the dimension of the Tensor's shape that the scales and zero_points correspond to. More... | |
bool | isFixedPoint () const |
Fixed point values are real numbers divided by a scale. More... | |
Static Public Member Functions | |
static UniformQuantizedPerAxisType | get (unsigned flags, Type storageType, Type expressedType, ArrayRef< double > scales, ArrayRef< int64_t > zeroPoints, int32_t quantizedDimension, int64_t storageTypeMin, int64_t storageTypeMax) |
Gets an instance of the type with all parameters specified but not checked. More... | |
static UniformQuantizedPerAxisType | getChecked (function_ref< InFlightDiagnostic()> emitError, unsigned flags, Type storageType, Type expressedType, ArrayRef< double > scales, ArrayRef< int64_t > zeroPoints, int32_t quantizedDimension, int64_t storageTypeMin, int64_t storageTypeMax) |
Gets an instance of the type with all specified parameters checked. More... | |
static LogicalResult | verify (function_ref< InFlightDiagnostic()> emitError, unsigned flags, Type storageType, Type expressedType, ArrayRef< double > scales, ArrayRef< int64_t > zeroPoints, int32_t quantizedDimension, int64_t storageTypeMin, int64_t storageTypeMax) |
Verifies construction invariants and issues errors/warnings. More... | |
![]() | |
static TypeID | getTypeID () |
Return a unique identifier for the concrete type. More... | |
template<typename T > | |
static bool | classof (T val) |
Provide an implementation of 'classof' that compares the type id of the provided value with that of the concrete type. More... | |
static detail::InterfaceMap | getInterfaceMap () |
Returns an interface map for the interfaces registered to this storage user. More... | |
static HasTraitFn | getHasTraitFn () |
Returns the function that returns true if the given Trait ID matches the IDs of any of the traits defined by the storage user. More... | |
template<typename... IfaceModels> | |
static void | attachInterface (MLIRContext &context) |
Attach the given models as implementations of the corresponding interfaces for the concrete storage user class. More... | |
template<typename... Args> | |
static ConcreteT | get (MLIRContext *ctx, Args... args) |
Get or create a new ConcreteT instance within the ctx. More... | |
template<typename... Args> | |
static ConcreteT | getChecked (const Location &loc, Args... args) |
Get or create a new ConcreteT instance within the ctx, defined at the given, potentially unknown, location. More... | |
template<typename... Args> | |
static ConcreteT | getChecked (function_ref< InFlightDiagnostic()> emitErrorFn, MLIRContext *ctx, Args... args) |
Get or create a new ConcreteT instance within the ctx. More... | |
static ConcreteT | getFromOpaquePointer (const void *ptr) |
Get an instance of the concrete type from a void pointer. More... | |
Additional Inherited Members | |
![]() | |
using | Base = StorageUserBase< ConcreteT, BaseT, StorageT, UniquerT, Traits... > |
Utility declarations for the concrete attribute class. More... | |
using | ImplType = StorageT |
using | HasTraitFn = bool(*)(TypeID) |
![]() | |
template<typename... Args> | |
LogicalResult | mutate (Args &&...args) |
Mutate the current storage instance. More... | |
ImplType * | getImpl () const |
Utility for easy access to the storage instance. More... | |
![]() | |
template<typename... Args> | |
static LogicalResult | verify (Args... args) |
Default implementation that just returns success. More... | |
Represents per-axis (also known as per-channel quantization).
Syntax synopsis: Per-axis, all parameters expressed: !quant<uniform[StorageType:ExpressedType:QuantizedDim]{QuantParams}> Per-axis, optional parameters omitted: !quant<uniform[StorageType]{Scale}>
StorageType: 'i'|'u' NumBits ExpressedType: 'f16', 'f32', 'bf16', 'f64' QuantizedDim: An integer value QuantParams: (Scale ':' ZeroPoint)+ Scale: A legal double value ZeroPoint: An integer value
Definition at line 314 of file QuantTypes.h.
|
static |
Gets an instance of the type with all parameters specified but not checked.
Definition at line 303 of file QuantTypes.cpp.
References mlir::Type::getContext().
Referenced by mlirUniformQuantizedPerAxisTypeGet().
|
static |
Gets an instance of the type with all specified parameters checked.
Returns a nullptr convertible type on failure.
Definition at line 313 of file QuantTypes.cpp.
References mlir::emitError(), and mlir::Type::getContext().
Referenced by mlir::quant::fakeQuantAttrsToType().
int32_t UniformQuantizedPerAxisType::getQuantizedDimension | ( | ) | const |
Specifies the dimension of the Tensor's shape that the scales and zero_points correspond to.
For example, a tensor t, with dims=[4, 3, 2, 1] with quantization params: scales=[1.0, 2.0, 3.0], zeroPoints=[1, 2, 3], quantizedDimension=1 will be quantized across the second dimension of t. t[:, 0, :, :] will have scale[0]=1.0, zero_point[0]=1 t[:, 1, :, :] will have scale[1]=2.0, zero_point[0]=2 t[:, 2, :, :] will have scale[2]=3.0, zero_point[0]=3
Definition at line 366 of file QuantTypes.cpp.
Referenced by printUniformQuantizedPerAxisType().
ArrayRef< double > UniformQuantizedPerAxisType::getScales | ( | ) | const |
Gets the quantization scales.
The scales designate the difference between the real values corresponding to consecutive quantized values differing by 1. The ith scale corresponds to the ith slice in the quantized_dimension.
Definition at line 358 of file QuantTypes.cpp.
Referenced by printUniformQuantizedPerAxisType().
ArrayRef< int64_t > UniformQuantizedPerAxisType::getZeroPoints | ( | ) | const |
Gets the storage values corresponding to the real value 0 in the affine equation.
The ith zero point corresponds to the ith slice in the quantized_dimension.
Definition at line 362 of file QuantTypes.cpp.
Referenced by printUniformQuantizedPerAxisType().
|
inline |
Fixed point values are real numbers divided by a scale.
Currently, only signed storage types are treated as fixed point. A fixed point value can be obtained from an affine value by subtracting the zeroPoint. In the future, this may be explicit versus implied by type and zeroPoint.
Definition at line 371 of file QuantTypes.h.
|
static |
Verifies construction invariants and issues errors/warnings.
Definition at line 323 of file QuantTypes.cpp.
References mlir::emitError(), mlir::failed(), mlir::failure(), mlir::Type::isa(), mlir::success(), and mlir::quant::QuantizedType::verify().