namespace of mxnet More...

Namespaces
	common

	cpp

	csr

	engine
	namespace of engine internal types.

	op
	namespace of arguments

	rowsparse

	rtc

Classes
struct	Context
	Context information about the execution environment. More...

struct	DataBatch
	DataBatch of NDArray, returned by Iterator. More...

struct	DataInst
	a single data instance More...

struct	DataIteratorReg
	Registry entry for DataIterator factory functions. More...

class	Engine
	Dependency engine that schedules operations. More...

class	Executor
	Executor of a computation graph. Executor can be created by Binding a symbol. More...

class	IIterator
	iterator type More...

class	Imperative
	runtime functions for NDArray More...

class	KVStore
	distributed key-value store More...

class	NDArray
	ndarray interface More...

struct	NDArrayFunctionReg
	Registry entry for NDArrayFunction. More...

struct	OpContext
	All the possible information needed by Operator.Forward and Backward This is the superset of RunContext. We use this data structure to bookkeep everything needed by Forward and Backward. More...

class	Operator
	Operator interface. Operator defines basic operation unit of optimized computation graph in mxnet. This interface relies on pre-allocated memory in TBlob, the caller need to set the memory region in TBlob correctly before calling Forward and Backward. More...

class	OperatorProperty
	OperatorProperty is a object that stores all information about Operator. It also contains method to generate context(device) specific operators. More...

struct	OperatorPropertyReg
	Registry entry for OperatorProperty factory functions. More...

class	OpStatePtr
	Operator state. This is a pointer type, its content is mutable even if OpStatePtr is const. More...

struct	Resource
	Resources used by mxnet operations. A resource is something special other than NDArray, but will still participate. More...

class	ResourceManager
	Global resource manager. More...

struct	ResourceRequest
	The resources that can be requested by Operator. More...

struct	RunContext
	execution time context. The information needed in runtime for actual execution. More...

class	Storage
	Storage manager across multiple devices. More...

class	TBlob
	tensor blob class that can be used to hold tensor of any dimension, any device and any data type, This is a weak type that can be used to transfer data through interface TBlob itself do not involve any arithmentic operations, but it can be converted to tensor of fixed dimension for further operations More...

Typedefs
typedef mshadow::cpu	cpu
	mxnet cpu More...

typedef mshadow::gpu	gpu
	mxnet gpu More...

typedef mshadow::index_t	index_t
	index type usually use unsigned More...

typedef mshadow::default_real_t	real_t
	data type that will be used to store ndarray More...

using	TShape = nnvm::TShape
	Shape data structure used to record shape information. More...

using	Op = nnvm::Op
	operator structure from NNVM More...

using	StorageTypeVector = std::vector< int >
	The result holder of storage type of each NodeEntry in the graph. More...

using	DispatchModeVector = std::vector< DispatchMode >
	The result holder of dispatch mode of each Node in the graph. More...

using	CachedOpPtr = std::shared_ptr< Imperative::CachedOp >

typedef std::function< IIterator< DataBatch > *()>	DataIteratorFactory
	typedef the factory function of data iterator More...

typedef std::function< void(NDArray *used_vars, real_t scalars, NDArray mutate_vars, int num_params, char param_keys, char **param_vals)>	NDArrayAPIFunction
	definition of NDArray function More...

using	FCreateOpState = std::function< OpStatePtr(const NodeAttrs &attrs, Context ctx, const std::vector< TShape > &in_shape, const std::vector< int > &in_type)>
	Create a Layer style, forward/backward operator. This is easy to write code that contains state. OpStatePtr is a pointer type, it's content is mutable even if OpStatePtr is constant. More...

using	FExecType = std::function< ExecType(const NodeAttrs &attrs)>
	Execution mode of this operator. More...

using	FStatefulCompute = std::function< void(const OpStatePtr &state, const OpContext &ctx, const std::vector< TBlob > &inputs, const std::vector< OpReqType > &req, const std::vector< TBlob > &outputs)>
	Resiger a compute function for stateful operator. OpStatePtr is a pointer type, it's content is mutable even if OpStatePtr is constant. More...

using	FStatefulComputeEx = std::function< void(const OpStatePtr &state, const OpContext &ctx, const std::vector< NDArray > &inputs, const std::vector< OpReqType > &req, const std::vector< NDArray > &outputs)>
	Resiger a compute function for stateful operator using NDArray interface. OpStatePtr is a pointer type, it's content is mutable even if OpStatePtr is constant. More...

using	FResourceRequest = std::function< std::vector< ResourceRequest >(const NodeAttrs &n)>
	The resource request from the operator. More...

using	FNDArrayFunction = std::function< void(const nnvm::NodeAttrs &attrs, const std::vector< NDArray > &inputs, std::vector< NDArray > *outputs)>
	Register an operator called as a NDArray function. More...

using	FCompute = std::function< void(const nnvm::NodeAttrs &attrs, const OpContext &ctx, const std::vector< TBlob > &inputs, const std::vector< OpReqType > &req, const std::vector< TBlob > &outputs)>
	Resiger a compute function for simple stateless forward only operator. More...

using	FComputeEx = std::function< void(const nnvm::NodeAttrs &attrs, const OpContext &ctx, const std::vector< NDArray > &inputs, const std::vector< OpReqType > &req, const std::vector< NDArray > &outputs)>
	Resiger an NDArray compute function for simple stateless forward only operator. More...

using	FInferStorageType = std::function< bool(const NodeAttrs &attrs, const int dev_mask, DispatchMode dispatch_mode, std::vector< int > in_attrs, std::vector< int > *out_attrs)>
	Resiger a storage and dispatch mode inference function based on storage types of the inputs and outputs, and the dev_mask for the operator. More...

typedef std::function< OperatorProperty *()>	OperatorPropertyFactory
	typedef the factory function of operator property More...

Enumerations
enum	FnProperty { FnProperty::kNormal, FnProperty::kCopyFromGPU, FnProperty::kCopyToGPU, FnProperty::kCPUPrioritized, FnProperty::kAsync, FnProperty::kDeleteVar }
	Function property, used to hint what action is pushed to engine. More...

enum	NDArrayStorageType { kUndefinedStorage = -1, kDefaultStorage, kRowSparseStorage, kCSRStorage }

enum	NDArrayFormatErr { kNormalErr, kCSRShapeErr, kCSRIndPtrErr, kCSRIdxErr, kRSPShapeErr, kRSPIdxErr }

enum	NDArrayFunctionTypeMask { kNDArrayArgBeforeScalar = 1, kScalarArgBeforeNDArray = 1 << 1, kAcceptEmptyMutateTarget = 1 << 2 }
	mask information on how functions can be exposed More...

enum	OpReqType { kNullOp, kWriteTo, kWriteInplace, kAddTo }
	operation request type to Forward and Backward More...

enum	ExecType { ExecType::kSync, ExecType::kAsync, ExecType::kCrossDeviceCopy }
	the execution type of the operator More...

enum	DispatchMode { DispatchMode::kUndefined = -1, DispatchMode::kFCompute, DispatchMode::kFComputeEx, DispatchMode::kFComputeFallback, DispatchMode::kVariable }
	the dispatch mode of the operator More...

Functions
size_t	num_aux_data (NDArrayStorageType stype)

void	CopyFromTo (const NDArray &from, const NDArray *to, int priority=0)
	issue an copy operation from one NDArray to another the two ndarray can sit on different devices this operation will be scheduled by the engine More...

void	CopyFromTo (const NDArray &from, const NDArray &to, int priority=0)
	issue an copy operation from one NDArray to another the two ndarray can sit on different devices this operation will be scheduled by the engine More...

void	ElementwiseSum (const std::vector< NDArray > &source, NDArray *out, int priority=0)
	Perform elementwise sum over each data from source, store result into out. More...

NDArray	operator+ (const NDArray &lhs, const NDArray &rhs)
	elementwise add More...

NDArray	operator+ (const NDArray &lhs, const real_t &rhs)
	elementwise add More...

NDArray	operator- (const NDArray &lhs, const NDArray &rhs)
	elementwise subtraction More...

NDArray	operator- (const NDArray &lhs, const real_t &rhs)
	elementwise subtraction More...

NDArray	operator* (const NDArray &lhs, const NDArray &rhs)
	elementwise multiplication More...

NDArray	operator* (const NDArray &lhs, const real_t &rhs)
	elementwise multiplication More...

NDArray	operator/ (const NDArray &lhs, const NDArray &rhs)
	elementwise division More...

NDArray	operator/ (const NDArray &lhs, const real_t &rhs)
	elementwise division More...

void	RandomSeed (uint32_t seed)
	Seed the random number generator. More...

void	SampleUniform (real_t begin, real_t end, NDArray *out)
	Sample uniform distribution for each elements of out. More...

void	SampleGaussian (real_t mu, real_t sigma, NDArray *out)
	Sample gaussian distribution for each elements of out. More...

void	SampleGamma (real_t alpha, real_t beta, NDArray *out)
	Sample gamma distribution for each elements of out. More...

void	SampleExponential (real_t lambda, NDArray *out)
	Sample exponential distribution for each elements of out. More...

void	SamplePoisson (real_t lambda, NDArray *out)
	Sample Poisson distribution for each elements of out. More...

void	SampleNegBinomial (int32_t k, real_t p, NDArray *out)
	Sample negative binomial distribution for each elements of out. More...

void	SampleGenNegBinomial (real_t mu, real_t alpha, NDArray *out)
	Sample generalized negative binomial distribution for each elements of out. More...

Detailed Description

namespace of mxnet

Macros/inlines to assist CLion to parse Cuda files (*.cu, *.cuh)

Typedef Documentation

using mxnet::CachedOpPtr = typedef std::shared_ptr<Imperative::CachedOp>

typedef mshadow::cpu mxnet::cpu

mxnet cpu

typedef std::function<IIterator<DataBatch> *()> mxnet::DataIteratorFactory

typedef the factory function of data iterator

using mxnet::DispatchModeVector = typedef std::vector<DispatchMode>

The result holder of dispatch mode of each Node in the graph.

*+ *
Note
Stored under graph.attrs["dispatch_mode"], provided by Pass "InferStorageType"
*
*
+ * Graph g = ApplyPass(src_graph, "InferStorageType");
+ * const DispatchModeVector& dispatch_modes = g.GetAttr<DispatchModeVector>("dispatch_mode");
+ * // get dispatch mode by entry node id
+ * int node_type = dispatch_modes[nid];
+ *
*
*
See also
FInferStorageType

using mxnet::FCompute = typedef std::function<void (const nnvm::NodeAttrs& attrs, const OpContext& ctx, const std::vector<TBlob>& inputs, const std::vector<OpReqType>& req, const std::vector<TBlob>& outputs)>

Resiger a compute function for simple stateless forward only operator.

Note: Register under "FCompute<cpu>" and "FCompute<gpu>"

using mxnet::FComputeEx = typedef std::function<void (const nnvm::NodeAttrs& attrs, const OpContext& ctx, const std::vector<NDArray>& inputs, const std::vector<OpReqType>& req, const std::vector<NDArray>& outputs)>

Resiger an NDArray compute function for simple stateless forward only operator.

Note: Register under "FComputeEx<xpu>" and "FComputeEx<xpu>" Dispatched only when inferred dispatch_mode is FDispatchComputeEx

using mxnet::FCreateOpState = typedef std::function<OpStatePtr (const NodeAttrs& attrs, Context ctx, const std::vector<TShape>& in_shape, const std::vector<int>& in_type)>

Create a Layer style, forward/backward operator. This is easy to write code that contains state. OpStatePtr is a pointer type, it's content is mutable even if OpStatePtr is constant.

This is not the only way to register an op execution function. More simpler or specialized operator form can be registered

Note: Register under "FCreateLayerOp"

using mxnet::FExecType = typedef std::function<ExecType (const NodeAttrs& attrs)>

Execution mode of this operator.

using mxnet::FInferStorageType = typedef std::function<bool (const NodeAttrs& attrs, const int dev_mask, DispatchMode* dispatch_mode, std::vector<int>* in_attrs, std::vector<int>* out_attrs)>

Resiger a storage and dispatch mode inference function based on storage types of the inputs and outputs, and the dev_mask for the operator.

Note: Register under "FInferStorageType"

using mxnet::FNDArrayFunction = typedef std::function<void (const nnvm::NodeAttrs& attrs, const std::vector<NDArray>& inputs, std::vector<NDArray>* outputs)>

Register an operator called as a NDArray function.

Note: Register under "FNDArrayFunction"

using mxnet::FResourceRequest = typedef std::function< std::vector<ResourceRequest> (const NodeAttrs& n)>

The resource request from the operator.

Note: Register under "FResourceRequest"

using mxnet::FStatefulCompute = typedef std::function<void (const OpStatePtr& state, const OpContext& ctx, const std::vector<TBlob>& inputs, const std::vector<OpReqType>& req, const std::vector<TBlob>& outputs)>

Resiger a compute function for stateful operator. OpStatePtr is a pointer type, it's content is mutable even if OpStatePtr is constant.

Note: Register under "FStatefulCompute<cpu>" and "FStatefulCompute<gpu>"

using mxnet::FStatefulComputeEx = typedef std::function<void (const OpStatePtr& state, const OpContext& ctx, const std::vector<NDArray>& inputs, const std::vector<OpReqType>& req, const std::vector<NDArray>& outputs)>

Resiger a compute function for stateful operator using NDArray interface. OpStatePtr is a pointer type, it's content is mutable even if OpStatePtr is constant.

Note: Register under "FStatefulComputeEx<cpu>" and "FStatefulComputeEx<gpu>"

typedef mshadow::gpu mxnet::gpu

mxnet gpu

typedef mshadow::index_t mxnet::index_t

index type usually use unsigned

typedef std::function<void (NDArray **used_vars, real_t *scalars, NDArray **mutate_vars, int num_params, char **param_keys, char **param_vals)> mxnet::NDArrayAPIFunction

definition of NDArray function

using mxnet::Op = typedef nnvm::Op

operator structure from NNVM

typedef std::function<OperatorProperty *()> mxnet::OperatorPropertyFactory

typedef the factory function of operator property

typedef mshadow::default_real_t mxnet::real_t

data type that will be used to store ndarray

using mxnet::StorageTypeVector = typedef std::vector<int>

The result holder of storage type of each NodeEntry in the graph.

Note: Stored under graph.attrs["storage_type"], provided by Pass "InferStorageType"

Graph g = ApplyPass(src_graph, "InferStorageType");
const StorageVector& stypes = g.GetAttr<StorageTypeVector>("storage_type");
// get storage type by entry id
int entry_type = stypes[g.indexed_graph().entry_id(my_entry)];

See also: FInferStorageType

using mxnet::TShape = typedef nnvm::TShape

Shape data structure used to record shape information.

Enumeration Type Documentation

enum mxnet::DispatchMode

strong

the dispatch mode of the operator

Enumerator
kUndefined
kFCompute
kFComputeEx
kFComputeFallback
kVariable

enum mxnet::ExecType

strong

the execution type of the operator

Enumerator
kSync	Forward/Backward are synchronize calls.
kAsync	Forward/Backward are asynchronize, will call OpContext.async_on_complete when operation finishes.
kCrossDeviceCopy	Cross device copy operation, this is a special operator That indicates copy across devices, the input and output can sit on different device. In current implementation, copy operator is specially handled by executor. This flag is used for special case treatment and future extension of different copy ops.

enum mxnet::FnProperty

strong

Function property, used to hint what action is pushed to engine.

Enumerator
kNormal	Normal operation.
kCopyFromGPU	Copy operation from GPU to other devices.
kCopyToGPU	Copy operation from CPU to other devices.
kCPUPrioritized	Prioritized sync operation on CPU.
kAsync	Asynchronous function call.
kDeleteVar	Delete variable call.

enum mxnet::NDArrayFormatErr

Enumerator
kNormalErr
kCSRShapeErr
kCSRIndPtrErr
kCSRIdxErr
kRSPShapeErr
kRSPIdxErr

enum mxnet::NDArrayFunctionTypeMask

mask information on how functions can be exposed

Enumerator

kNDArrayArgBeforeScalar

all the use_vars should go before scalar

kScalarArgBeforeNDArray

all the scalar should go before use_vars

kAcceptEmptyMutateTarget

whether this function allows the handles in the target to be empty NDArray that are not yet initialized, and will initialize them when the function is invoked.

most function should support this, except copy between different devices, which requires the NDArray to be pre-initialized with context

enum mxnet::NDArrayStorageType

Enumerator
kUndefinedStorage
kDefaultStorage
kRowSparseStorage
kCSRStorage

enum mxnet::OpReqType

operation request type to Forward and Backward

Enumerator
kNullOp	no operation, do not write anything
kWriteTo	write gradient to provided space
kWriteInplace	perform an inplace write, Target shares memory with one of input arguments. This option only happen when
kAddTo	add to the provided space

Function Documentation

void mxnet::CopyFromTo	(	const NDArray &	from,
		const NDArray *	to,
		int	priority = `0`
	)

issue an copy operation from one NDArray to another the two ndarray can sit on different devices this operation will be scheduled by the engine

Parameters

from	the ndarray we want to copy data from
to	the target ndarray
priority	Priority of the action.

Note: The function name explicitly marks the order of from and to due to different possible convention carried by copy function.

void mxnet::CopyFromTo	(	const NDArray &	from,
		const NDArray &	to,
		int	priority = `0`
	)

issue an copy operation from one NDArray to another the two ndarray can sit on different devices this operation will be scheduled by the engine

Parameters

from	the ndarray we want to copy data from
to	the target ndarray
priority	Priority of the action.

Note: The function name explicitly marks the order of from and to due to different possible convention carried by copy function.

void mxnet::ElementwiseSum	(	const std::vector< NDArray > &	source,
		NDArray *	out,
		int	priority = `0`
	)

Perform elementwise sum over each data from source, store result into out.

Parameters

source	the ndarray we want to sum
out	the target ndarray
priority	Priority of the action.

size_t mxnet::num_aux_data ( NDArrayStorageType stype )

Returns: the number of aux data used for given storage type

NDArray mxnet::operator*	(	const NDArray &	lhs,
		const NDArray &	rhs
	)

elementwise multiplication

Parameters

lhs	left operand
rhs	right operand

Returns: a new result ndarray

NDArray mxnet::operator*	(	const NDArray &	lhs,
		const real_t &	rhs
	)

elementwise multiplication

Parameters

lhs	left operand
rhs	right operand

Returns: a new result ndarray

NDArray mxnet::operator+	(	const NDArray &	lhs,
		const NDArray &	rhs
	)

elementwise add

Parameters

lhs	left operand
rhs	right operand

Returns: a new result ndarray

NDArray mxnet::operator+	(	const NDArray &	lhs,
		const real_t &	rhs
	)

elementwise add

Parameters

lhs	left operand
rhs	right operand

Returns: a new result ndarray

NDArray mxnet::operator-	(	const NDArray &	lhs,
		const NDArray &	rhs
	)

elementwise subtraction

Parameters

lhs	left operand
rhs	right operand

Returns: a new result ndarray

NDArray mxnet::operator-	(	const NDArray &	lhs,
		const real_t &	rhs
	)

elementwise subtraction

Parameters

lhs	left operand
rhs	right operand

Returns: a new result ndarray

NDArray mxnet::operator/	(	const NDArray &	lhs,
		const NDArray &	rhs
	)

elementwise division

Parameters

lhs	left operand
rhs	right operand

Returns: a new result ndarray

NDArray mxnet::operator/	(	const NDArray &	lhs,
		const real_t &	rhs
	)

elementwise division

Parameters

lhs	left operand
rhs	right operand

Returns: a new result ndarray

void mxnet::RandomSeed ( uint32_t seed )

Seed the random number generator.

Parameters

seed	the seed to set to global random number generators.

void mxnet::SampleExponential	(	real_t	lambda,
		NDArray *	out
	)

Sample exponential distribution for each elements of out.

Parameters

lambda	parameter (rate) of the exponential distribution
out	output NDArray.

void mxnet::SampleGamma	(	real_t	alpha,
		real_t	beta,
		NDArray *	out
	)

Sample gamma distribution for each elements of out.

Parameters

alpha	parameter (shape) of the gamma distribution
beta	parameter (scale) of the gamma distribution
out	output NDArray.

void mxnet::SampleGaussian	(	real_t	mu,
		real_t	sigma,
		NDArray *	out
	)

Sample gaussian distribution for each elements of out.

Parameters

mu	mean of gaussian distribution.
sigma	standard deviation of gaussian distribution.
out	output NDArray.

void mxnet::SampleGenNegBinomial	(	real_t	mu,
		real_t	alpha,
		NDArray *	out
	)

Sample generalized negative binomial distribution for each elements of out.

Parameters

mu	parameter (mean) of the distribution
alpha	parameter (over dispersion) of the distribution
out	output NDArray.

void mxnet::SampleNegBinomial	(	int32_t	k,
		real_t	p,
		NDArray *	out
	)

Sample negative binomial distribution for each elements of out.

Parameters

k	failure limit
p	success probability
out	output NDArray.

void mxnet::SamplePoisson	(	real_t	lambda,
		NDArray *	out
	)

Sample Poisson distribution for each elements of out.

Parameters

lambda	parameter (rate) of the Poisson distribution
out	output NDArray.

void mxnet::SampleUniform	(	real_t	begin,
		real_t	end,
		NDArray *	out
	)

Sample uniform distribution for each elements of out.

Parameters

begin	lower bound of distribution.
end	upper bound of distribution.
out	output NDArray.

Namespaces

Classes

Typedefs

Enumerations

Functions

Detailed Description

Typedef Documentation

Enumeration Type Documentation

Function Documentation