org.bytedeco.pytorch

Class NestedTensorImpl

java.lang.Object

org.bytedeco.javacpp.Pointer

org.bytedeco.pytorch.TensorImpl

org.bytedeco.pytorch.NestedTensorImpl

All Implemented Interfaces:

AutoCloseable

@Namespace(value="at::native")
 @NoOffset
 @Properties(inherit=torch.class)
public class NestedTensorImpl
extends TensorImpl

Nested Class Summary

Nested classes/interfaces inherited from class org.bytedeco.pytorch.TensorImpl

TensorImpl.ImplType, TensorImpl.LongIdentity, TensorImpl.SizesStridesPolicy, TensorImpl.SymIntIdentity

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
NestedTensorImpl(int impl_type, Tensor base_tensor, Tensor nested_sizes, Tensor nested_strides, Tensor storage_offsets)
NestedTensorImpl(Pointer p) Pointer cast constructor.
NestedTensorImpl(Storage storage, DispatchKeySet key_set, TypeMeta data_type, Tensor nested_sizes, Tensor nested_strides, Tensor storage_offsets)
NestedTensorImpl(TensorImpl.ImplType impl_type, Tensor base_tensor, Tensor nested_sizes, Tensor nested_strides, Tensor storage_offsets)
NestedTensorImpl(Tensor buffer, Tensor nested_sizes)
NestedTensorImpl(Tensor buffer, Tensor nested_sizes, Tensor nested_strides, Tensor storage_offsets)

Method Summary

Modifier and Type	Method and Description
long	get_buffer_size()
Tensor	get_buffer() Return a view of the nested tensor as a 1 dimensional contiguous tensor.
Tensor	get_nested_sizes()
Tensor	get_nested_strides()
Tensor	get_storage_offsets()
Tensor	get_unsafe_storage_as_tensor() If possible use get_buffer() instead.
LongOptional	opt_size(long d)
long	size(long d) Return the size of a tensor at some dimension, wrapping the dimension if necessary.

Methods inherited from class org.bytedeco.pytorch.TensorImpl

_change_backend_component_keys, _fw_grad, _generic_sizes, _generic_sizes, _generic_storage_offset, _generic_storage_offset, _generic_strides, _generic_strides, _is_zerotensor, _set_conj, _set_fw_grad, _set_neg, _set_zero, allow_tensor_metadata_change, autograd_meta, bump_version, data, device_type, device, dim_default, dim, dtype_initialized, dtype, empty_tensor_restride_symint, empty_tensor_restride_symint, empty_tensor_restride, empty_tensor_restride, Extend, FreeMemory, generic_set_sizes_contiguous, generic_set_sizes_contiguous, generic_set_sizes_contiguous, get_backend_meta_intrusive_ptr, get_backend_meta, get_device, grad, has_compatible_shallow_copy_type, has_named_tensor_meta, has_storage, has_symbolic_sizes_strides, is_conj, is_contiguous_default, is_contiguous, is_contiguous, is_cpu, is_cuda, is_empty, is_hip, is_hpu, is_inference, is_ipu, is_lazy, is_meta, is_metal, is_mkldnn, is_mps, is_mtia, is_neg, is_nested, is_non_overlapping_and_dense_default, is_non_overlapping_and_dense, is_ort, is_python_dispatch, is_quantized, is_sparse_csr, is_sparse, is_strides_like_channels_last_3d, is_strides_like_channels_last, is_strides_like_default, is_strides_like, is_ve, is_vulkan, is_wrapped_number, is_xla, is_xpu, itemsize, key_set, layout, mutable_data, mutable_grad, named_tensor_meta, numel_default, numel, pyobj_slot, raw_mutable_data, release_resources, release_storage_and_set_meta_custom_data_ptr_error_msg_, remove_autograd_key, requires_grad, ReserveSpace, Reshape, set_allow_tensor_metadata_change, set_autograd_meta, set_backend_meta, set_custom_device, set_custom_layout, set_custom_sizes_strides, set_custom_sizes_strides, set_named_tensor_meta, set_python_custom_device, set_python_custom_layout, set_python_custom_sizes_strides, set_python_custom_sizes_strides, set_python_dispatch, set_requires_grad, set_size, set_sizes_and_strides, set_sizes_and_strides, set_sizes_and_strides, set_sizes_and_strides, set_sizes_and_strides, set_sizes_and_strides, set_sizes_contiguous, set_sizes_contiguous, set_storage_access_should_throw, set_storage_and_dtype, set_storage_keep_dtype, set_storage_offset, set_stride, set_version_counter, set_wrapped_number, shallow_copy_and_detach, shallow_copy_from, ShareData, ShareExternalPointer, sizes_default, sizes, storage_initialized, storage_offset_default, storage_offset, storage, stride, strides_default, strides, support_as_strided, sym_numel_default, sym_numel, sym_size, sym_sizes_default, sym_sizes, sym_storage_offset_default, sym_storage_offset, sym_strides_default, sym_strides, unique_version, unsafe_storage, version_counter

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

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

Constructor Detail

NestedTensorImpl

public NestedTensorImpl(Pointer p)

Pointer cast constructor. Invokes Pointer(Pointer).

NestedTensorImpl

public NestedTensorImpl(@Cast(value={"","c10::Storage&&"}) @StdMove
                        Storage storage,
                        @ByVal
                        DispatchKeySet key_set,
                        @Const @ByVal
                        TypeMeta data_type,
                        @ByVal
                        Tensor nested_sizes,
                        @ByVal
                        Tensor nested_strides,
                        @ByVal
                        Tensor storage_offsets)

NestedTensorImpl

public NestedTensorImpl(@ByVal
                        Tensor buffer,
                        @ByVal
                        Tensor nested_sizes,
                        @ByVal
                        Tensor nested_strides,
                        @ByVal
                        Tensor storage_offsets)

NestedTensorImpl

public NestedTensorImpl(@ByVal
                        Tensor buffer,
                        @ByVal
                        Tensor nested_sizes)

NestedTensorImpl

public NestedTensorImpl(TensorImpl.ImplType impl_type,
                        @Const @ByRef
                        Tensor base_tensor,
                        @ByVal
                        Tensor nested_sizes,
                        @ByVal
                        Tensor nested_strides,
                        @ByVal
                        Tensor storage_offsets)

NestedTensorImpl

public NestedTensorImpl(@Cast(value="c10::TensorImpl::ImplType")
                        int impl_type,
                        @Const @ByRef
                        Tensor base_tensor,
                        @ByVal
                        Tensor nested_sizes,
                        @ByVal
                        Tensor nested_strides,
                        @ByVal
                        Tensor storage_offsets)

Method Detail

get_nested_sizes

@Const
 @ByRef
public Tensor get_nested_sizes()

get_nested_strides

@Const
 @ByRef
public Tensor get_nested_strides()

get_storage_offsets

@Const
 @ByRef
public Tensor get_storage_offsets()

opt_size

@ByVal
public LongOptional opt_size(@Cast(value="int64_t")
                                    long d)

size

@Cast(value="int64_t")
public long size(@Cast(value="int64_t")
                                        long d)

Description copied from class: TensorImpl

Return the size of a tensor at some dimension, wrapping the dimension if necessary. NOTE: if you know wrapping is unnecessary, do sizes()[d] instead; it will be faster

Overrides:

size in class TensorImpl

get_buffer

@ByVal
public Tensor get_buffer()

Return a view of the nested tensor as a 1 dimensional contiguous tensor. The buffer tensor created by this function shares the same storage_impl as the original nested tensor, and therefore can be seen as a view.

Returns:

A newly constructed view tensor

get_unsafe_storage_as_tensor

@ByVal
public Tensor get_unsafe_storage_as_tensor()

If possible use get_buffer() instead. This function returns the storage as a tensor directly, which is not safe to use in general. If using this function, The caller must ensure to account for nested_sizes, nested_strides and storage_offsets.

Returns:

A newly constructed view tensor

get_buffer_size

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