org.bytedeco.pytorch

Class MultiheadAttentionImpl

java.lang.Object

org.bytedeco.javacpp.Pointer

org.bytedeco.pytorch.Module

org.bytedeco.pytorch.MultiheadAttentionImplCloneable

org.bytedeco.pytorch.MultiheadAttentionImpl

All Implemented Interfaces:

AutoCloseable

@Namespace(value="torch::nn")
 @NoOffset
 @Properties(inherit=torch.class)
public class MultiheadAttentionImpl
extends MultiheadAttentionImplCloneable

Applies the MultiheadAttention function element-wise. See https://pytorch.org/docs/master/nn.html#torch.nn.MultiheadAttention to learn about the exact behavior of this module. See the documentation for torch::nn::MultiheadAttentionOptions class to learn what constructor arguments are supported for this module. Example:

  MultiheadAttention model(MultiheadAttentionOptions(20, 10).bias(false));
  

Nested Class Summary

Nested classes/interfaces inherited from class org.bytedeco.javacpp.Pointer

Pointer.CustomDeallocator, Pointer.Deallocator, Pointer.NativeDeallocator, Pointer.ReferenceCounter

Field Summary

Fields inherited from class org.bytedeco.javacpp.Pointer

address, capacity, limit, position

Constructor Summary

Constructors

Constructor and Description
MultiheadAttentionImpl(long embed_dim, long num_heads)
MultiheadAttentionImpl(Module pointer) Downcast constructor.
MultiheadAttentionImpl(MultiheadAttentionOptions options_)
MultiheadAttentionImpl(Pointer p) Pointer cast constructor.

Method Summary

Modifier and Type	Method and Description
boolean	_qkv_same_embed_dim()
MultiheadAttentionImpl	_qkv_same_embed_dim(boolean setter)
void	_reset_parameters()
Tensor	bias_k()
MultiheadAttentionImpl	bias_k(Tensor setter)
Tensor	bias_v()
MultiheadAttentionImpl	bias_v(Tensor setter)
T_TensorTensor_T	forward(Tensor query, Tensor key, Tensor value)
T_TensorTensor_T	forward(Tensor query, Tensor key, Tensor value, Tensor key_padding_mask, boolean need_weights, Tensor attn_mask, boolean average_attn_weights)
long	head_dim()
MultiheadAttentionImpl	head_dim(long setter)
Tensor	in_proj_bias()
MultiheadAttentionImpl	in_proj_bias(Tensor setter)
Tensor	in_proj_weight()
MultiheadAttentionImpl	in_proj_weight(Tensor setter)
Tensor	k_proj_weight()
MultiheadAttentionImpl	k_proj_weight(Tensor setter)
MultiheadAttentionOptions	options() The options with which this Module was constructed.
MultiheadAttentionImpl	options(MultiheadAttentionOptions setter)
Tensor	q_proj_weight()
MultiheadAttentionImpl	q_proj_weight(Tensor setter)
void	reset() reset() must perform initialization of all members with reference semantics, most importantly parameters, buffers and submodules.
Tensor	v_proj_weight()
MultiheadAttentionImpl	v_proj_weight(Tensor setter)

Methods inherited from class org.bytedeco.pytorch.MultiheadAttentionImplCloneable

asModule, asModule, clone, clone

Methods inherited from class org.bytedeco.pytorch.Module

apply, apply, apply, apply, apply, apply, apply, apply, buffers, buffers, children, eval, is_serializable, is_training, load, modules, modules, name, named_buffers, named_buffers, named_children, named_modules, named_modules, named_modules, named_parameters, named_parameters, parameters, parameters, pretty_print, register_buffer, register_buffer, register_module, register_module, register_parameter, register_parameter, register_parameter, register_parameter, save, shiftLeft, to, to, to, train, unregister_module, unregister_module, zero_grad

Methods inherited from class org.bytedeco.javacpp.Pointer

address, asBuffer, asByteBuffer, availablePhysicalBytes, calloc, capacity, capacity, close, deallocate, deallocate, deallocateReferences, deallocator, deallocator, equals, fill, formatBytes, free, getDirectBufferAddress, getPointer, getPointer, getPointer, getPointer, hashCode, interruptDeallocatorThread, isNull, isNull, limit, limit, malloc, maxBytes, maxPhysicalBytes, memchr, memcmp, memcpy, memmove, memset, offsetAddress, offsetof, offsetof, parseBytes, physicalBytes, physicalBytesInaccurate, position, position, put, realloc, referenceCount, releaseReference, retainReference, setNull, sizeof, sizeof, toString, totalBytes, totalCount, totalPhysicalBytes, withDeallocator, zero

Methods inherited from class java.lang.Object

finalize, getClass, notify, notifyAll, wait, wait, wait

Constructor Detail

MultiheadAttentionImpl

public MultiheadAttentionImpl(Pointer p)

Pointer cast constructor. Invokes Pointer(Pointer).

MultiheadAttentionImpl

public MultiheadAttentionImpl(Module pointer)

Downcast constructor.

MultiheadAttentionImpl

public MultiheadAttentionImpl(@Cast(value="int64_t")
                              long embed_dim,
                              @Cast(value="int64_t")
                              long num_heads)

MultiheadAttentionImpl

public MultiheadAttentionImpl(@Const @ByRef
                              MultiheadAttentionOptions options_)

Method Detail

forward

@ByVal
public T_TensorTensor_T forward(@Const @ByRef
                                       Tensor query,
                                       @Const @ByRef
                                       Tensor key,
                                       @Const @ByRef
                                       Tensor value,
                                       @Const @ByRef(nullValue="torch::Tensor{}")
                                       Tensor key_padding_mask,
                                       @Cast(value="bool")
                                       boolean need_weights,
                                       @Const @ByRef(nullValue="torch::Tensor{}")
                                       Tensor attn_mask,
                                       @Cast(value="bool")
                                       boolean average_attn_weights)

forward

@ByVal
public T_TensorTensor_T forward(@Const @ByRef
                                       Tensor query,
                                       @Const @ByRef
                                       Tensor key,
                                       @Const @ByRef
                                       Tensor value)

reset

public void reset()

Description copied from class: MultiheadAttentionImplCloneable

reset() must perform initialization of all members with reference semantics, most importantly parameters, buffers and submodules.

Overrides:

reset in class MultiheadAttentionImplCloneable

_reset_parameters

public void _reset_parameters()

options

@ByRef
public MultiheadAttentionOptions options()

The options with which this Module was constructed.

options

public MultiheadAttentionImpl options(MultiheadAttentionOptions setter)

_qkv_same_embed_dim

@Cast(value="bool")
public boolean _qkv_same_embed_dim()

_qkv_same_embed_dim

public MultiheadAttentionImpl _qkv_same_embed_dim(boolean setter)

in_proj_weight

@ByRef
public Tensor in_proj_weight()

in_proj_weight

public MultiheadAttentionImpl in_proj_weight(Tensor setter)

in_proj_bias

@ByRef
public Tensor in_proj_bias()

in_proj_bias

public MultiheadAttentionImpl in_proj_bias(Tensor setter)

bias_k

@ByRef
public Tensor bias_k()

bias_k

public MultiheadAttentionImpl bias_k(Tensor setter)

bias_v

@ByRef
public Tensor bias_v()

bias_v

public MultiheadAttentionImpl bias_v(Tensor setter)

q_proj_weight

@ByRef
public Tensor q_proj_weight()

q_proj_weight

public MultiheadAttentionImpl q_proj_weight(Tensor setter)

k_proj_weight

@ByRef
public Tensor k_proj_weight()

k_proj_weight

public MultiheadAttentionImpl k_proj_weight(Tensor setter)

v_proj_weight

@ByRef
public Tensor v_proj_weight()

v_proj_weight

public MultiheadAttentionImpl v_proj_weight(Tensor setter)

head_dim

@Cast(value="int64_t")
public long head_dim()

head_dim

public MultiheadAttentionImpl head_dim(long setter)