Caffe2 - C++ API
A deep learning, cross platform ML framework
Public Member Functions | Data Fields
caffe2::ScatterAssignOp< Context > Class Template Reference

Update slices of the tensor in-place by overriding. More...

#include <utility_ops.h>

Inheritance diagram for caffe2::ScatterAssignOp< Context >:
caffe2::Operator< Context > caffe2::OperatorBase caffe2::Observable< OperatorBase >

Public Member Functions

 ScatterAssignOp (const OperatorDef &operator_def, Workspace *ws)
bool RunOnDevice () override
- Public Member Functions inherited from caffe2::Operator< Context >
 Operator (const OperatorDef &operator_def, Workspace *ws)
const Tensor< Context > & Input (int idx)
Tensor< Context > * Output (int idx)
void WaitEvent (const Event &ev, int stream_id=-1) final
void WaitEvents (const std::vector< const Event * > &events, int stream_id=-1) final
bool Run (int stream_id=0) final
bool RunAsync (int stream_id=0) final
bool IsStreamFree (int stream_id) const override
bool HasAsyncPart () const override
bool SupportsAsyncScheduling () const override
const Context * getContext () const
- Public Member Functions inherited from caffe2::OperatorBase
 OperatorBase (const OperatorDef &operator_def, Workspace *ws)
bool HasArgument (const string &name) const
 Checks if the operator has an argument of the given name.
template<typename T >
GetSingleArgument (const string &name, const T &default_value) const
template<typename T >
bool HasSingleArgumentOfType (const string &name) const
template<typename T >
vector< T > GetRepeatedArgument (const string &name, const vector< T > &default_value={}) const
template<typename T >
const T & Input (int idx)
template<typename T >
T * Output (int idx)
template<typename T >
T * Output (int idx, T *allocated)
const BlobInputBlob (int idx)
BlobOutputBlob (int idx)
template<typename T >
bool InputIsType (int idx)
template<typename T >
bool OutputIsType (int idx)
int InputSize () const
int OutputSize () const
const vector< const Blob * > & Inputs () const
const vector< Blob * > & Outputs ()
vector< TensorShape > InputTensorShapes ()
void Wait (const OperatorBase &other, int stream_id=-1)
virtual void Finish ()
virtual void AddRelatedBlobInfo (EnforceNotMet *err)
const OperatorDef & debug_def () const
void set_debug_def (const std::shared_ptr< const OperatorDef > &operator_def)
bool has_debug_def () const
void RecordLastFailedOpNetPosition ()
int net_position () const
void set_net_position (int idx)
const DeviceOption & device_option () const
const Eventevent () const
Eventevent ()
void ResetEvent ()
void DisableEvent ()
bool IsEventDisabled () const
const std::string & type () const
void annotate_engine (const std::string &engine)
const std::string & engine () const
- Public Member Functions inherited from caffe2::Observable< OperatorBase >
const ObserverAttachObserver (std::unique_ptr< Observer > observer)
std::unique_ptr< ObserverDetachObserver (const Observer *observer_ptr)
 Returns a unique_ptr to the removed observer. More...
virtual size_t NumObservers ()
void StartAllObservers ()
void StopAllObservers ()

Data Fields


Additional Inherited Members

- Public Types inherited from caffe2::Observable< OperatorBase >
using Observer = ObserverBase< OperatorBase >
- Static Public Attributes inherited from caffe2::OperatorBase
static constexpr int kNoNetPositionSet = -1
- Protected Member Functions inherited from caffe2::Operator< Context >
void RecordEvent (const char *err_msg=nullptr) final
std::string getErrorMsg ()
- Protected Member Functions inherited from caffe2::OperatorBase
- Protected Attributes inherited from caffe2::Operator< Context >
Context context_
- Protected Attributes inherited from caffe2::OperatorBase
std::unique_ptr< Eventevent_
- Protected Attributes inherited from caffe2::Observable< OperatorBase >
std::vector< std::unique_ptr< Observer > > observers_list_

Detailed Description

template<class Context>
class caffe2::ScatterAssignOp< Context >

Update slices of the tensor in-place by overriding.

Input: DATA - tensor to be updated INDICES - 1-D list of indices on the first dimension of X_0 that need to be updated SLICES - update slices, has to have shape of len(INDICES) + shape(X_0)[1:]

Output: DATA - has to be exactly the same tensor as the input 0

Note: The op pretty much ignores the exact shapes of the input arguments and cares only about sizes. It's done for performance consideration to avoid unnecessary reshapes. Only first dimension of X_0 is important, let's call it N. If M is the total size of X_0 and K is the size of INDICES then X_i is assumed to be of shape K x (M / N) regardless of the real shape.

Note: Each update in INDICES is applied independently which means that if duplicated elements are present in INDICES arbitrary one will win.

For now really works only on CPU because of INDICES access

Definition at line 565 of file utility_ops.h.

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