ONE - On-device Neural Engine
|
#include <ITensorRegistry.h>
Public Member Functions | |
virtual | ~ITensorRegistry ()=default |
Deconstruct itself. | |
virtual ITensor * | getITensor (const ir::OperandIndex &)=0 |
Returns pointer of ITensor among native and migrant tensors. | |
virtual ITensor * | getNativeITensor (const ir::OperandIndex &)=0 |
Returns pointer of ITensor among native tensors. | |
virtual bool | setMigrantTensor (const ir::OperandIndex &, IPortableTensor *) |
Set the Migrant Tensor which are from other backends. | |
Definition at line 31 of file ITensorRegistry.h.
|
virtualdefault |
Deconstruct itself.
|
pure virtual |
Returns pointer of ITensor among native and migrant tensors.
Native Tensor is a tensor that is managed by this backend Migrant Tensor is a tensor that is imported from another backend
Implemented in onert::backend::acl_common::AclTensorRegistry< T_AclTensorManager >, onert::backend::PortableTensorRegistryTemplate< T_Tensor >, onert::backend::builtin::TensorRegistry, onert::backend::train::PortableTensorRegistryTemplate< Tensor, TrainableTensor, BackPropTensor, GradientTensor >, and onert::backend::builtin::train::TensorRegistry.
Referenced by onert::exec::MinMaxRecorder::handleSubgraphBegin().
|
pure virtual |
Returns pointer of ITensor among native tensors.
Unlike getITensor
, this function only searches from native tensors
Implemented in onert::backend::acl_common::AclTensorRegistry< T_AclTensorManager >, onert::backend::PortableTensorRegistryTemplate< T_Tensor >, onert::backend::builtin::TensorRegistry, onert::backend::train::PortableTensorRegistryTemplate< Tensor, TrainableTensor, BackPropTensor, GradientTensor >, and onert::backend::builtin::train::TensorRegistry.
Referenced by onert::backend::basic::initConsts().
|
inlinevirtual |
Set the Migrant Tensor which are from other backends.
Reimplemented in onert::backend::PortableTensorRegistryTemplate< T_Tensor >, onert::backend::builtin::TensorRegistry, onert::backend::train::PortableTensorRegistryTemplate< Tensor, TrainableTensor, BackPropTensor, GradientTensor >, and onert::backend::builtin::train::TensorRegistry.
Definition at line 61 of file ITensorRegistry.h.