lib/torch/native/native_functions.yaml in torch-rb-0.2.7 vs lib/torch/native/native_functions.yaml in torch-rb-0.3.0

- old
+ new

@@ -1,46 +1,54 @@ # See README.md in this directory for more guidance +# *********NB: _cast_* operators are DEPRECATED and will be removed +# eventually. These were previously used before TorchScript IR supported +# representing ScalarType's. They are now superseded by usage of +# `aten::to()`. The ops remain here for backward compatibility purposes. -# Temporary type cast operators. These are needed to trace type-casts now since -# Type's are not supported in the IR. Instead, we call down to these -# specialized operators for each datatype. -# TODO: remove when we have Type support in the IR +# DEPRECATED. DO NOT USE - func: _cast_Byte(Tensor self, bool non_blocking=False) -> Tensor use_c10_dispatcher: full variants: function +# DEPRECATED. DO NOT USE - func: _cast_Char(Tensor self, bool non_blocking=False) -> Tensor use_c10_dispatcher: full variants: function +# DEPRECATED. DO NOT USE - func: _cast_Double(Tensor self, bool non_blocking=False) -> Tensor use_c10_dispatcher: full variants: function +# DEPRECATED. DO NOT USE - func: _cast_Float(Tensor self, bool non_blocking=False) -> Tensor use_c10_dispatcher: full variants: function +# DEPRECATED. DO NOT USE - func: _cast_Int(Tensor self, bool non_blocking=False) -> Tensor use_c10_dispatcher: full variants: function +# DEPRECATED. DO NOT USE - func: _cast_Long(Tensor self, bool non_blocking=False) -> Tensor use_c10_dispatcher: full variants: function +# DEPRECATED. DO NOT USE - func: _cast_Short(Tensor self, bool non_blocking=False) -> Tensor use_c10_dispatcher: full variants: function +# DEPRECATED. DO NOT USE - func: _cast_Half(Tensor self, bool non_blocking=False) -> Tensor use_c10_dispatcher: full variants: function # Computes the gradient of current tensor w.r.t. graph leaves. -- func: backward(Tensor self, Tensor? gradient=None, bool keep_graph=False, bool create_graph=False) -> () +- func: backward(Tensor self, Tensor? gradient=None, bool? retain_graph=None, bool create_graph=False) -> () manual_kernel_registration: True variants: method # DEPRECATED. Sets the tensor data held by this `Variable` to be the same as # `new_data`. It requires that `new_data` and `Variable` have compatible tensor @@ -77,18 +85,17 @@ # - func: output_nr(Tensor self) -> int use_c10_dispatcher: full manual_kernel_registration: True variants: method - supports_named_tensor: True - func: _version(Tensor self) -> int use_c10_dispatcher: full manual_kernel_registration: True variants: method -- func: requires_grad_(Tensor(a!) self, bool _requires_grad=True) -> Tensor(a!) +- func: requires_grad_(Tensor(a!) self, bool requires_grad=True) -> Tensor(a!) manual_kernel_registration: True variants: method # Enables .grad attribute for non-leaf Tensors. - func: retain_grad(Tensor(a!) self) -> () @@ -96,56 +103,51 @@ manual_kernel_registration: True variants: method - func: rename_(Tensor(a!) self, Dimname[]? names) -> Tensor(a!) variants: method - supports_named_tensor: True - func: rename(Tensor(a) self, Dimname[]? names) -> Tensor(a) variants: method - supports_named_tensor: True - func: align_to(Tensor(a) self, Dimname[] names) -> Tensor(a) variants: method - supports_named_tensor: True - func: align_to.ellipsis_idx(Tensor(a) self, Dimname[] order, int ellipsis_idx) -> Tensor(a) variants: method - supports_named_tensor: True - func: align_as(Tensor self, Tensor other) -> Tensor + use_c10_dispatcher: full variants: method - supports_named_tensor: True - func: align_tensors(Tensor[] tensors) -> Tensor[] - supports_named_tensor: True + use_c10_dispatcher: full - func: refine_names(Tensor(a) self, Dimname[] names) -> Tensor(a) variants: method - supports_named_tensor: True - func: unflatten.Dimname(Tensor self, Dimname dim, int[] sizes, Dimname[] names) -> Tensor variants: method - supports_named_tensor: True - func: unflatten.int(Tensor self, int dim, int[] sizes, Dimname[] names) -> Tensor variants: method - supports_named_tensor: True - - func: _use_cudnn_ctc_loss(Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, int blank) -> bool + use_c10_dispatcher: full dispatch: CUDA: _use_cudnn_ctc_loss - func: _cudnn_ctc_loss(Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, int blank, bool deterministic, bool zero_infinity) -> (Tensor, Tensor) + use_c10_dispatcher: full dispatch: CUDA: _cudnn_ctc_loss - func: _use_cudnn_rnn_flatten_weight() -> bool use_c10_dispatcher: full - func: _cudnn_rnn_flatten_weight(Tensor[] weight_arr, int weight_stride0, int input_size, int mode, int hidden_size, int num_layers, bool batch_first, bool bidirectional) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: _cudnn_rnn_flatten_weight - func: _cudnn_rnn(Tensor input, Tensor[] weight, int weight_stride0, Tensor? weight_buf, Tensor hx, Tensor? cx, int mode, int hidden_size, int num_layers, bool batch_first, float dropout, bool train, bool bidirectional, int[] batch_sizes, Tensor? dropout_state) -> (Tensor, Tensor, Tensor, Tensor, Tensor) dispatch: @@ -165,11 +167,10 @@ - func: _fused_dropout(Tensor self, float p, Generator? generator=None) -> (Tensor, Tensor) variants: function dispatch: CUDA: fused_dropout_cuda - supports_named_tensor: True - func: _masked_scale(Tensor self, Tensor mask, float scale) -> Tensor use_c10_dispatcher: full variants: function dispatch: @@ -177,29 +178,24 @@ - func: _sobol_engine_draw(Tensor quasi, int n, Tensor sobolstate, int dimension, int num_generated, ScalarType? dtype) -> (Tensor, Tensor) - func: _sobol_engine_ff_(Tensor(a!) self, int n, Tensor sobolstate, int dimension, int num_generated) -> Tensor(a!) - - func: _sobol_engine_scramble_(Tensor(a!) self, Tensor ltm, int dimension) -> Tensor(a!) - - func: _sobol_engine_initialize_state_(Tensor(a!) self, int dimension) -> Tensor(a!) - - func: _reshape_from_tensor(Tensor self, Tensor shape) -> Tensor use_c10_dispatcher: full - func: _shape_as_tensor(Tensor self) -> Tensor use_c10_dispatcher: full - func: dropout(Tensor input, float p, bool train) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True - func: dropout_(Tensor(a!) self, float p, bool train) -> Tensor(a!) - supports_named_tensor: True - func: feature_dropout(Tensor input, float p, bool train) -> Tensor use_c10_dispatcher: full - func: feature_dropout_(Tensor(a!) self, float p, bool train) -> Tensor(a!) @@ -207,134 +203,138 @@ - func: alpha_dropout(Tensor input, float p, bool train) -> Tensor use_c10_dispatcher: full - func: alpha_dropout_(Tensor(a!) self, float p, bool train) -> Tensor(a!) - - func: feature_alpha_dropout(Tensor input, float p, bool train) -> Tensor use_c10_dispatcher: full - func: feature_alpha_dropout_(Tensor(a!) self, float p, bool train) -> Tensor(a!) - - func: abs(Tensor self) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: abs_(Tensor(a!) self) -> Tensor(a!) variants: function, method - supports_named_tensor: True - func: abs.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True +- func: absolute(Tensor self) -> Tensor + use_c10_dispatcher: full + variants: function, method + dispatch: + CPU: abs + CUDA: abs + +- func: absolute_(Tensor(a!) self) -> Tensor(a!) + variants: function, method + dispatch: + CPU: abs_ + CUDA: abs_ + +- func: absolute.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) + dispatch: + CPU: abs_out + CUDA: abs_out + - func: angle(Tensor self) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: angle.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True -- func: real(Tensor self) -> Tensor +- func: view_as_real(Tensor(a) self) -> Tensor(a) use_c10_dispatcher: full variants: function - supports_named_tensor: True -- func: imag(Tensor self) -> Tensor +- func: view_as_complex(Tensor(a) self) -> Tensor(a) use_c10_dispatcher: full variants: function - supports_named_tensor: True +- func: real(Tensor(a) self) -> Tensor(a) + use_c10_dispatcher: full + variants: function + +- func: imag(Tensor(a) self) -> Tensor(a) + use_c10_dispatcher: full + variants: function + - func: conj(Tensor self) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: conj.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: acos(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: acos_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: acos.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: avg_pool1d(Tensor self, int[1] kernel_size, int[1] stride=[], int[1] padding=0, bool ceil_mode=False, bool count_include_pad=True) -> Tensor + use_c10_dispatcher: full - func: adaptive_avg_pool1d(Tensor self, int[1] output_size) -> Tensor + use_c10_dispatcher: full # Return: (Tensor output, Tensor indices) - func: adaptive_max_pool1d(Tensor self, int[1] output_size) -> (Tensor, Tensor) + use_c10_dispatcher: full - func: add.Tensor(Tensor self, Tensor other, *, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: function, method dispatch: CPU: add CUDA: add SparseCPU: add_sparse SparseCUDA: add_sparse MkldnnCPU: mkldnn_add - supports_named_tensor: True + Vulkan: vulkan_add - func: add_.Tensor(Tensor(a!) self, Tensor other, *, Scalar alpha=1) -> Tensor(a!) variants: method dispatch: CPU: add_ CUDA: add_ SparseCPU: add_sparse_ SparseCUDA: add_sparse_ MkldnnCPU: mkldnn_add_ - supports_named_tensor: True - func: add.out(Tensor self, Tensor other, *, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: add_out CUDA: add_out SparseCPU: add_out_sparse_cpu SparseCUDA: add_out_sparse_cuda MkldnnCPU: mkldnn_add_out - supports_named_tensor: True # For C++ only, until we have conversion from C++ numbers to Tensor - func: add.Scalar(Tensor self, Scalar other, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: add_.Scalar(Tensor(a!) self, Scalar other, Scalar alpha=1) -> Tensor(a!) variants: method - supports_named_tensor: True - func: addmv(Tensor self, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: function, method - dispatch: - CPU: legacy::cpu::_th_addmv - CUDA: legacy::cuda::_th_addmv - supports_named_tensor: True - func: addmv_(Tensor(a!) self, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1) -> Tensor(a!) variants: function, method - dispatch: - CPU: legacy::cpu::_th_addmv_ - CUDA: legacy::cuda::_th_addmv_ - supports_named_tensor: True - func: addmv.out(Tensor self, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!) + +- func: _addmv_impl_(Tensor(a!) self, Tensor self2, Tensor mat, Tensor vec, *, Scalar beta=1, Scalar alpha=1) -> Tensor(a!) dispatch: - CPU: legacy::cpu::_th_addmv_out - CUDA: legacy::cuda::_th_addmv_out - supports_named_tensor: True + CPU: addmv_impl_cpu + CUDA: addmv_impl_cuda - func: addr(Tensor self, Tensor vec1, Tensor vec2, *, Scalar beta=1, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: function, method @@ -342,13 +342,15 @@ variants: method - func: addr.out(Tensor self, Tensor vec1, Tensor vec2, *, Scalar beta=1, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!) - func: affine_grid_generator(Tensor theta, int[] size, bool align_corners) -> Tensor + use_c10_dispatcher: full variants: function - func: affine_grid_generator_backward(Tensor grad, int[] size, bool align_corners) -> Tensor + use_c10_dispatcher: full variants: function - func: all.dim(Tensor self, int dim, bool keepdim=False) -> Tensor use_c10_dispatcher: full variants: function, method @@ -395,63 +397,90 @@ # (so that it can be traced directly). - func: _dim_arange(Tensor like, int dim) -> Tensor use_c10_dispatcher: full - func: argmax(Tensor self, int? dim=None, bool keepdim=False) -> Tensor + use_c10_dispatcher: full variants: function, method dispatch: CPU: argmax CUDA: argmax - func: argmin(Tensor self, int? dim=None, bool keepdim=False) -> Tensor + use_c10_dispatcher: full variants: function, method dispatch: CPU: argmin CUDA: argmin +- func: acosh(Tensor self) -> Tensor + use_c10_dispatcher: full + supports_named_tensor: True + variants: function, method + +- func: acosh_(Tensor(a!) self) -> Tensor(a!) + supports_named_tensor: True + variants: function, method + +- func: acosh.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) + supports_named_tensor: True + +- func: asinh(Tensor self) -> Tensor + use_c10_dispatcher: full + supports_named_tensor: True + variants: function, method + +- func: asinh_(Tensor(a!) self) -> Tensor(a!) + supports_named_tensor: True + variants: function, method + +- func: asinh.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) + supports_named_tensor: True + +- func: atanh(Tensor self) -> Tensor + use_c10_dispatcher: full + supports_named_tensor: True + variants: function, method + +- func: atanh_(Tensor(a!) self) -> Tensor(a!) + supports_named_tensor: True + variants: function, method + +- func: atanh.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) + supports_named_tensor: True + - func: as_strided(Tensor(a) self, int[] size, int[] stride, int? storage_offset=None) -> Tensor(a) + use_c10_dispatcher: full variants: function, method dispatch: CPU: as_strided_tensorimpl CUDA: as_strided_tensorimpl QuantizedCPU: as_strided_qtensorimpl + QuantizedCUDA: as_strided_qtensorimpl device_guard: False - supports_named_tensor: True - func: as_strided_(Tensor(a!) self, int[] size, int[] stride, int? storage_offset=None) -> Tensor(a!) variants: function, method device_guard: False - func: asin(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: asin_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: asin.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: atan(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: atan_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _atan__cpu - CUDA: _atan__cuda - func: atan.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _atan_out_cpu - CUDA: _atan_out_cuda - func: baddbmm(Tensor self, Tensor batch1, Tensor batch2, *, Scalar beta=1, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: function, method dispatch: @@ -489,29 +518,19 @@ - func: _batch_norm_impl_index_backward(int impl_index, Tensor input, Tensor grad_output, Tensor? weight, Tensor? running_mean, Tensor? running_var, Tensor? save_mean, Tensor? save_var_transform, bool train, float eps, bool[3] output_mask, Tensor reservedSpace) -> (Tensor, Tensor, Tensor) # Sample bernoulli with values in `self` as probability. - func: bernoulli(Tensor self, *, Generator? generator=None) -> Tensor variants: function, method - supports_named_tensor: True - func: bernoulli.out(Tensor self, *, Generator? generator=None, Tensor(a!) out) -> Tensor(a!) variants: function - supports_named_tensor: True - func: bernoulli_.Tensor(Tensor(a!) self, Tensor p, *, Generator? generator=None) -> Tensor(a!) variants: method - dispatch: - CPU: bernoulli_tensor_cpu_ - CUDA: bernoulli_tensor_cuda_ - supports_named_tensor: True - func: bernoulli_.float(Tensor(a!) self, float p=0.5, *, Generator? generator=None) -> Tensor(a!) variants: method - dispatch: - CPU: bernoulli_scalar_cpu_ - CUDA: bernoulli_scalar_cuda_ - supports_named_tensor: True # This out-of-place version isn't used explicitly, but needed by jit. # There is no default valid on `p` here because it would introduce ambiguity # with `bernoulli(Tensor self, *, Generator? generator=None)` declaration. - func: bernoulli.p(Tensor self, float p, *, Generator? generator=None) -> Tensor @@ -559,78 +578,67 @@ CPU: _bincount_cpu CUDA: _bincount_cuda - func: bitwise_not(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: bitwise_not_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: method - func: bitwise_not.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: bitwise_not_out CUDA: bitwise_not_out - func: logical_not(Tensor self) -> Tensor - supports_named_tensor: True + use_c10_dispatcher: full variants: function, method - func: logical_not_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: method - func: logical_not.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: logical_not_out CUDA: logical_not_out - func: logical_xor(Tensor self, Tensor other) -> Tensor + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: logical_xor_(Tensor(a!) self, Tensor other) -> Tensor(a!) variants: method - supports_named_tensor: True - func: logical_xor.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: logical_xor_out CUDA: logical_xor_out - supports_named_tensor: True - func: logical_and(Tensor self, Tensor other) -> Tensor + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: logical_and_(Tensor(a!) self, Tensor other) -> Tensor(a!) variants: method - supports_named_tensor: True - func: logical_and.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: logical_and_out CUDA: logical_and_out - supports_named_tensor: True - func: logical_or(Tensor self, Tensor other) -> Tensor + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: logical_or_(Tensor(a!) self, Tensor other) -> Tensor(a!) variants: method - supports_named_tensor: True - func: logical_or.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: logical_or_out CUDA: logical_or_out - supports_named_tensor: True - func: blackman_window(int window_length, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor - func: blackman_window.periodic(int window_length, bool periodic, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor @@ -638,131 +646,119 @@ use_c10_dispatcher: full variants: function, method dispatch: CPU: bmm_cpu CUDA: bmm_cuda - supports_named_tensor: True + SparseCPU: bmm_sparse_cpu + SparseCUDA: bmm_sparse_cuda +- func: _bmm(Tensor self, Tensor mat2, *, bool deterministic=False) -> Tensor + use_c10_dispatcher: full + variants: function + dispatch: + SparseCUDA: _bmm_sparse_cuda + - func: bmm.out(Tensor self, Tensor mat2, *, Tensor(a!) out) -> Tensor(a!) variants: function dispatch: CPU: bmm_out_cpu CUDA: bmm_out_cuda - supports_named_tensor: True + SparseCPU: bmm_out_sparse_cpu + SparseCUDA: bmm_out_sparse_cuda +- func: _bmm.out(Tensor self, Tensor mat2, *, bool deterministic=False, Tensor(a!) out) -> Tensor(a!) + variants: function + dispatch: + SparseCUDA: _bmm_out_sparse_cuda + - func: broadcast_tensors(Tensor[] tensors) -> Tensor[] + use_c10_dispatcher: full device_guard: False - func: cat(Tensor[] tensors, int dim=0) -> Tensor - supports_named_tensor: True + use_c10_dispatcher: full - func: cat.out(Tensor[] tensors, int dim=0, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: cat.names(Tensor[] tensors, Dimname dim) -> Tensor - supports_named_tensor: True - func: cat.names_out(Tensor[] tensors, Dimname dim, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True +- func: block_diag(Tensor[] tensors) -> Tensor + use_c10_dispatcher: full + variants: function + - func: ceil(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: ceil_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: ceil.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: ceil_out CUDA: ceil_out - func: chain_matmul(Tensor[] matrices) -> Tensor + use_c10_dispatcher: full variants: function - func: chunk(Tensor(a) self, int chunks, int dim=0) -> Tensor(a)[] + use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: clamp(Tensor self, Scalar? min=None, Scalar? max=None) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method dispatch: CPU: clamp CUDA: clamp QuantizedCPU: quantized_clamp + Vulkan: vulkan_clamp - func: clamp_(Tensor(a!) self, Scalar? min=None, Scalar? max=None) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _clamp__cpu - CUDA: _clamp__cuda - func: clamp.out(Tensor self, Scalar? min=None, Scalar? max=None, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _clamp_out_cpu - CUDA: _clamp_out_cuda - func: clamp_max(Tensor self, Scalar max) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: clamp_max_(Tensor(a!) self, Scalar max) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _clamp_max__cpu - CUDA: _clamp_max__cuda - func: clamp_max.out(Tensor self, Scalar max, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _clamp_max_out_cpu - CUDA: _clamp_max_out_cuda - func: clamp_min(Tensor self, Scalar min) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: clamp_min_(Tensor(a!) self, Scalar min) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _clamp_min__cpu - CUDA: _clamp_min__cuda - func: clamp_min.out(Tensor self, Scalar min, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _clamp_min_out_cpu - CUDA: _clamp_min_out_cuda - func: cudnn_is_acceptable(Tensor self) -> bool use_c10_dispatcher: full device_guard: False - func: constant_pad_nd(Tensor self, int[] pad, Scalar value=0) -> Tensor + use_c10_dispatcher: full variants: function - func: contiguous(Tensor self, *, MemoryFormat memory_format=contiguous_format) -> Tensor variants: method - supports_named_tensor: True - func: convolution(Tensor input, Tensor weight, Tensor? bias, int[] stride, int[] padding, int[] dilation, bool transposed, int[] output_padding, int groups) -> Tensor - func: convolution_overrideable(Tensor input, Tensor weight, Tensor? bias, int[] stride, int[] padding, int[] dilation, bool transposed, int[] output_padding, int groups) -> Tensor - func: convolution_backward_overrideable(Tensor grad_output, Tensor input, Tensor weight, int[] stride, int[] padding, int[] dilation, bool transposed, int[] output_padding, int groups, bool[3] output_mask) -> (Tensor grad_input, Tensor grad_weight, Tensor grad_bias) + use_c10_dispatcher: full - func: _convolution(Tensor input, Tensor weight, Tensor? bias, int[] stride, int[] padding, int[] dilation, bool transposed, int[] output_padding, int groups, bool benchmark, bool deterministic, bool cudnn_enabled) -> Tensor - func: _convolution_nogroup(Tensor input, Tensor weight, Tensor? bias, int[] stride, int[] padding, int[] dilation, bool transposed, int[] output_padding) -> Tensor @@ -776,10 +772,11 @@ - func: conv_tbc(Tensor self, Tensor weight, Tensor bias, int pad=0) -> Tensor use_c10_dispatcher: full - func: conv_tbc_backward(Tensor self, Tensor input, Tensor weight, Tensor bias, int pad) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full # NB: we inherit the goofy argument order from PyTorch torch.nn.functional - func: conv_transpose1d(Tensor input, Tensor weight, Tensor? bias=None, int[1] stride=1, int[1] padding=0, int[1] output_padding=0, int groups=1, int[1] dilation=1) -> Tensor - func: conv_transpose2d.input(Tensor input, Tensor weight, Tensor? bias=None, int[2] stride=1, int[2] padding=0, int[2] output_padding=0, int groups=1, int[2] dilation=1) -> Tensor @@ -788,51 +785,32 @@ - func: copy_(Tensor(a!) self, Tensor src, bool non_blocking=False) -> Tensor(a!) manual_kernel_registration: True variants: method device_guard: False - supports_named_tensor: True - func: _copy_from(Tensor self, Tensor dst, bool non_blocking=False) -> Tensor use_c10_dispatcher: full dispatch: {} - func: cos(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: cos_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _cos__cpu - CUDA: _cos__cuda - func: cos.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _cos_out_cpu - CUDA: _cos_out_cuda - func: cosh(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: cosh_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _cosh__cpu - CUDA: _cosh__cuda - func: cosh.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _cosh_out_cpu - CUDA: _cosh_out_cuda - func: cosine_embedding_loss(Tensor input1, Tensor input2, Tensor target, float margin=0.0, int reduction=Mean) -> Tensor use_c10_dispatcher: full - func: cudnn_affine_grid_generator(Tensor theta, int N, int C, int H, int W) -> Tensor grid @@ -858,137 +836,135 @@ - func: cudnn_convolution.deprecated(Tensor self, Tensor weight, Tensor? bias, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor dispatch: CUDA: cudnn_convolution_deprecated - func: cudnn_convolution(Tensor self, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: cudnn_convolution - func: cudnn_convolution_backward_input(int[] self_size, Tensor grad_output, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: cudnn_convolution_backward_input - func: cudnn_convolution_backward(Tensor self, Tensor grad_output, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic, bool[2] output_mask) -> (Tensor, Tensor) + use_c10_dispatcher: full dispatch: CUDA: cudnn_convolution_backward - func: cudnn_convolution_backward_weight(int[] weight_size, Tensor grad_output, Tensor self, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: cudnn_convolution_backward_weight - func: cudnn_convolution_transpose.deprecated(Tensor self, Tensor weight, Tensor? bias, int[] padding, int[] output_padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor dispatch: CUDA: cudnn_convolution_transpose_deprecated - func: cudnn_convolution_transpose(Tensor self, Tensor weight, int[] padding, int[] output_padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: cudnn_convolution_transpose # NB: output_padding not strictly needed here, but it's helpful for the float # backwards - func: cudnn_convolution_transpose_backward(Tensor self, Tensor grad_output, Tensor weight, int[] padding, int[] output_padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic, bool[2] output_mask) -> (Tensor, Tensor) + use_c10_dispatcher: full dispatch: CUDA: cudnn_convolution_transpose_backward - func: cudnn_convolution_transpose_backward_input(Tensor grad_output, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: cudnn_convolution_transpose_backward_input - func: cudnn_convolution_transpose_backward_weight(int[] weight_size, Tensor grad_output, Tensor self, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: cudnn_convolution_transpose_backward_weight # NB: input is special cased in a way I don't quite understand - func: cudnn_grid_sampler(Tensor self, Tensor grid) -> Tensor output use_c10_dispatcher: full dispatch: CUDA: cudnn_grid_sampler_forward - func: cudnn_grid_sampler_backward(Tensor self, Tensor grid, Tensor grad_output) -> (Tensor grad_self, Tensor grad_grid) + use_c10_dispatcher: full dispatch: CUDA: cudnn_grid_sampler_backward - func: cummax(Tensor self, int dim) -> (Tensor values, Tensor indices) - supports_named_tensor: True + use_c10_dispatcher: full variants: function, method - func: cummax.out(Tensor self, int dim, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: cummax.dimname(Tensor self, Dimname dim) -> (Tensor values, Tensor indices) - supports_named_tensor: True variants: function, method - func: cummax.dimname_out(Tensor self, Dimname dim, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: _cummax_helper(Tensor self, Tensor(a!) values, Tensor(b!) indices, int dim) -> () variants: function dispatch: CPU: cummax_helper_cpu CUDA: cummax_helper_cuda - func: cummin(Tensor self, int dim) -> (Tensor values, Tensor indices) - supports_named_tensor: True + use_c10_dispatcher: full variants: function, method - func: cummin.out(Tensor self, int dim, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: cummin.dimname(Tensor self, Dimname dim) -> (Tensor values, Tensor indices) - supports_named_tensor: True variants: function, method - func: cummin.dimname_out(Tensor self, Dimname dim, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: _cummin_helper(Tensor self, Tensor(a!) values, Tensor(b!) indices, int dim) -> () variants: function dispatch: CPU: cummin_helper_cpu CUDA: cummin_helper_cuda - func: cumprod(Tensor self, int dim, *, ScalarType? dtype=None) -> Tensor - supports_named_tensor: True variants: function, method - func: cumprod.out(Tensor self, int dim, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: cumprod.dimname(Tensor self, Dimname dim, *, ScalarType? dtype=None) -> Tensor - supports_named_tensor: True variants: function, method - func: cumprod.dimname_out(Tensor self, Dimname dim, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: cumsum(Tensor self, int dim, *, ScalarType? dtype=None) -> Tensor - supports_named_tensor: True variants: function, method - func: cumsum.out(Tensor self, int dim, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: cumsum.dimname(Tensor self, Dimname dim, *, ScalarType? dtype=None) -> Tensor - supports_named_tensor: True variants: function, method - func: cumsum.dimname_out(Tensor self, Dimname dim, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: ctc_loss.IntList(Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, int blank=0, int reduction=Mean, bool zero_infinity=False) -> Tensor + use_c10_dispatcher: full # convenience function that converts to intlists for you - func: ctc_loss.Tensor(Tensor log_probs, Tensor targets, Tensor input_lengths, Tensor target_lengths, int blank=0, int reduction=Mean, bool zero_infinity=False) -> Tensor use_c10_dispatcher: full - func: _ctc_loss(Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, int blank=0, bool zero_infinity=False) -> (Tensor, Tensor) + use_c10_dispatcher: full dispatch: CPU: ctc_loss_cpu CUDA: ctc_loss_gpu - func: _ctc_loss_backward(Tensor grad, Tensor log_probs, Tensor targets, int[] input_lengths, int[] target_lengths, Tensor neg_log_likelihood, Tensor log_alpha, int blank, bool zero_infinity=False) -> Tensor + use_c10_dispatcher: full dispatch: CPU: ctc_loss_backward_cpu CUDA: ctc_loss_backward_gpu - func: det(Tensor self) -> Tensor @@ -1002,16 +978,15 @@ - func: diagflat(Tensor self, int offset=0) -> Tensor use_c10_dispatcher: full variants: function, method - func: diagonal(Tensor(a) self, int offset=0, int dim1=0, int dim2=1) -> Tensor(a) + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: diagonal.Dimname(Tensor(a) self, *, Dimname outdim, Dimname dim1, Dimname dim2, int offset=0) -> Tensor(a) variants: function, method - supports_named_tensor: True - func: fill_diagonal_(Tensor(a!) self, Scalar fill_value, bool wrap=False) -> Tensor(a!) variants: method - func: div.Tensor(Tensor self, Tensor other) -> Tensor @@ -1020,51 +995,45 @@ dispatch: CPU: div CUDA: div SparseCPU: div_sparse SparseCUDA: div_sparse - supports_named_tensor: True - func: div_.Tensor(Tensor(a!) self, Tensor other) -> Tensor(a!) variants: method dispatch: CPU: div_ CUDA: div_ SparseCPU: div_sparse_ SparseCUDA: div_sparse_ - supports_named_tensor: True - func: div.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: div_out CUDA: div_out SparseCPU: div_out_sparse_zerodim SparseCUDA: div_out_sparse_zerodim - supports_named_tensor: True # For C++ only, until we have conversion from C++ numbers to Tensor - func: div.Scalar(Tensor self, Scalar other) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: div_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!) variants: method - supports_named_tensor: True - func: dot(Tensor self, Tensor tensor) -> Tensor use_c10_dispatcher: full variants: function, method dispatch: CPU: legacy::cpu::_th_dot CUDA: legacy::cuda::_th_dot - supports_named_tensor: True - func: dot.out(Tensor self, Tensor tensor, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: einsum(str equation, Tensor[] tensors) -> Tensor + use_c10_dispatcher: full - func: embedding(Tensor weight, Tensor indices, int padding_idx=-1, bool scale_grad_by_freq=False, bool sparse=False) -> Tensor use_c10_dispatcher: full - func: embedding_backward(Tensor grad, Tensor indices, int num_weights, int padding_idx, bool scale_grad_by_freq, bool sparse) -> Tensor @@ -1113,20 +1082,23 @@ use_c10_dispatcher: full dispatch: CPU: _embedding_bag_per_sample_weights_backward_cpu CUDA: _embedding_bag_per_sample_weights_backward_cuda +- func: empty_meta(int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor + - func: empty.names(int[] size, *, Dimname[]? names, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor device_guard: False - func: empty.memory_format(int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor dispatch: CPU: empty_cpu CUDA: empty_cuda MkldnnCPU: empty_mkldnn SparseCPU: empty_sparse SparseCUDA: empty_sparse + Vulkan: empty_vulkan - func: new_empty(Tensor self, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor variants: method - func: new_full(Tensor self, int[] size, Scalar fill_value, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor @@ -1137,11 +1109,12 @@ # other overrides are to provide a more helpful error message that dtype is required - func: _empty_affine_quantized(int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, float scale=1, int zero_point=0, MemoryFormat? memory_format=contiguous_format) -> Tensor dispatch: CPU: empty_affine_quantized_other_backends_stub - QuantizedCPU: empty_affine_quantized_cpu + QuantizedCPU: empty_affine_quantized + QuantizedCUDA: empty_affine_quantized # it's a factory function receiving a tensor argument, thus overriding explicitly # other overrides are to provide a more helpful error message that dtype is required - func: _empty_per_channel_affine_quantized(int[] size, *, Tensor scales, Tensor zero_points, int axis, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=contiguous_format) -> Tensor category_override: factory @@ -1149,99 +1122,74 @@ CPU: empty_per_channel_affine_quantized_other_backends_stub QuantizedCPU: empty_per_channel_affine_quantized_cpu - func: resize_(Tensor(a!) self, int[] size, *, MemoryFormat? memory_format=None) -> Tensor(a!) manual_kernel_registration: True - supports_named_tensor: True variants: method device_guard: False +- func: empty_quantized(int[] size, Tensor qtensor) -> Tensor + variants: function + dispatch: + QuantizedCPU: empty_quantized + QuantizedCUDA: empty_quantized + - func: empty.out(int[] size, *, MemoryFormat? memory_format=None, Tensor(a!) out) -> Tensor(a!) device_guard: False - func: empty_like(Tensor self, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor device_guard: False - supports_named_tensor: True - func: empty_strided(int[] size, int[] stride, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor dispatch: CPU: empty_strided_cpu CUDA: empty_strided_cuda + Vulkan: empty_strided_vulkan - func: erf(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: erf_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _erf__cpu - CUDA: _erf__cuda - func: erf.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _erf_out_cpu - CUDA: _erf_out_cuda - func: erfc(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: erfc_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _erfc__cpu - CUDA: _erfc__cuda - func: erfc.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _erfc_out_cpu - CUDA: _erfc_out_cuda - func: exp(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: exp_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _exp__cpu - CUDA: _exp__cuda - func: exp.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _exp_out_cpu - CUDA: _exp_out_cuda - func: expm1(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: expm1_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: expm1.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: expm1_out CUDA: expm1_out - func: expand(Tensor(a) self, int[] size, *, bool implicit=False) -> Tensor(a) + use_c10_dispatcher: full variants: method # This is method-only to match the previous tensor API. In the future we could make this a function too. device_guard: False - supports_named_tensor: True - func: expand_as(Tensor self, Tensor other) -> Tensor use_c10_dispatcher: full variants: method # This is method-only to match the previous tensor API. In the future we could make this a function too. device_guard: False @@ -1261,102 +1209,86 @@ CUDA: eye_out_cuda - func: flatten.using_ints(Tensor self, int start_dim=0, int end_dim=-1) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: flatten.named_out_dim(Tensor self, int start_dim, int end_dim, Dimname out_dim) -> Tensor variants: function, method - supports_named_tensor: True - func: flatten.using_names(Tensor self, Dimname start_dim, Dimname end_dim, Dimname out_dim) -> Tensor variants: function, method - supports_named_tensor: True - func: flatten.DimnameList(Tensor self, Dimname[] dims, Dimname out_dim) -> Tensor variants: function, method - supports_named_tensor: True - func: fill_.Scalar(Tensor(a!) self, Scalar value) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: fill_.Tensor(Tensor(a!) self, Tensor value) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: floor(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: floor_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: floor.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: floor_out CUDA: floor_out - func: floor_divide(Tensor self, Tensor other) -> Tensor + use_c10_dispatcher: full variants: function, method dispatch: CPU: floor_divide CUDA: floor_divide SparseCPU: floor_divide_sparse SparseCUDA: floor_divide_sparse - supports_named_tensor: True - func: floor_divide_.Tensor(Tensor(a!) self, Tensor other) -> Tensor(a!) variants: method dispatch: CPU: floor_divide_ CUDA: floor_divide_ SparseCPU: floor_divide_sparse_ SparseCUDA: floor_divide_sparse_ - supports_named_tensor: True - func: floor_divide.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: floor_divide_out CUDA: floor_divide_out SparseCPU: floor_divide_out_sparse_zerodim SparseCUDA: floor_divide_out_sparse_zerodim - supports_named_tensor: True - func: floor_divide.Scalar(Tensor self, Scalar other) -> Tensor + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: floor_divide_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!) variants: method - supports_named_tensor: True - func: frac(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: frac_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: frac.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: full.names(int[] size, Scalar fill_value, *, Dimname[]? names, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor device_guard: False - func: full(int[] size, Scalar fill_value, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor - func: full.out(int[] size, Scalar fill_value, *, Tensor(a!) out) -> Tensor(a!) - func: full_like(Tensor self, Scalar fill_value, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor - supports_named_tensor: True - func: from_file(str filename, bool? shared=None, int? size=0, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor dispatch: CPU: from_file @@ -1379,10 +1311,11 @@ dispatch: CPU: grid_sampler_2d_cpu CUDA: grid_sampler_2d_cuda - func: grid_sampler_2d_backward(Tensor grad_output, Tensor input, Tensor grid, int interpolation_mode, int padding_mode, bool align_corners) -> (Tensor, Tensor) + use_c10_dispatcher: full dispatch: CPU: grid_sampler_2d_backward_cpu CUDA: grid_sampler_2d_backward_cuda - func: grid_sampler_3d(Tensor input, Tensor grid, int interpolation_mode, int padding_mode, bool align_corners) -> Tensor @@ -1390,10 +1323,11 @@ dispatch: CPU: grid_sampler_3d_cpu CUDA: grid_sampler_3d_cuda - func: grid_sampler_3d_backward(Tensor grad_output, Tensor input, Tensor grid, int interpolation_mode, int padding_mode, bool align_corners) -> (Tensor, Tensor) + use_c10_dispatcher: full dispatch: CPU: grid_sampler_3d_backward_cpu CUDA: grid_sampler_3d_backward_cuda - func: hann_window(int window_length, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor @@ -1417,10 +1351,20 @@ - func: ger.out(Tensor self, Tensor vec2, *, Tensor(a!) out) -> Tensor(a!) - func: group_norm(Tensor input, int num_groups, Tensor? weight=None, Tensor? bias=None, float eps=1e-05, bool cudnn_enabled=True) -> Tensor +- func: native_group_norm(Tensor input, Tensor? weight, Tensor? bias, int N, int C, int HxW, int group, float eps) -> (Tensor, Tensor, Tensor) + dispatch: + CPU: native_group_norm + CUDA: native_group_norm + +- func: native_group_norm_backward(Tensor grad_out, Tensor input, Tensor mean, Tensor rstd, Tensor? weight, int N, int C, int HxW, int group, bool[3] output_mask) -> (Tensor, Tensor, Tensor) + dispatch: + CPU: native_group_norm_backward + CUDA: native_group_norm_backward + # FFT - func: fft(Tensor self, int signal_ndim, bool normalized=False) -> Tensor use_c10_dispatcher: full variants: function, method @@ -1432,13 +1376,15 @@ - func: rfft(Tensor self, int signal_ndim, bool normalized=False, bool onesided=True) -> Tensor use_c10_dispatcher: full variants: function, method - func: irfft(Tensor self, int signal_ndim, bool normalized=False, bool onesided=True, int[] signal_sizes=[]) -> Tensor + use_c10_dispatcher: full variants: function, method - func: _fft_with_size(Tensor self, int signal_ndim, bool complex_input, bool complex_output, bool inverse, int[] checked_signal_sizes, bool normalized, bool onesided, int[] output_sizes) -> Tensor + use_c10_dispatcher: full variants: function dispatch: CPU: _fft_mkl CUDA: _fft_cufft @@ -1447,14 +1393,14 @@ - func: _cufft_get_plan_cache_max_size(int device_index) -> int use_c10_dispatcher: full - func: _cufft_set_plan_cache_max_size(int device_index, int max_size) -> () - use_c10_dispatcher: unboxed_only + use_c10_dispatcher: full - func: _cufft_clear_plan_cache(int device_index) -> () - use_c10_dispatcher: unboxed_only + use_c10_dispatcher: full - func: index.Tensor(Tensor self, Tensor?[] indices) -> Tensor variants: function, method # NB: This function is special-cased in tools/autograd/gen_variable_type.py # NB: The following functions are declared in aten/src/ATen/templates/TensorBody.h and defined in aten/src/ATen/TensorIndexing.cpp: @@ -1508,13 +1454,12 @@ use_c10_dispatcher: full variants: function, method - func: isnan(Tensor self) -> Tensor use_c10_dispatcher: full - variants: function + variants: function, method device_guard: False - supports_named_tensor: True dispatch: CPU: isnan CUDA: isnan SparseCPU: isnan_sparse SparseCUDA: isnan_sparse @@ -1526,61 +1471,53 @@ - func: is_floating_point(Tensor self) -> bool use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: is_complex(Tensor self) -> bool use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: is_nonzero(Tensor self) -> bool use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: is_same_size(Tensor self, Tensor other) -> bool use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: is_signed(Tensor self) -> bool use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True -- func: kl_div(Tensor self, Tensor target, int reduction=Mean) -> Tensor +- func: kl_div(Tensor self, Tensor target, int reduction=Mean, *, bool log_target=False) -> Tensor use_c10_dispatcher: full -- func: kl_div_backward(Tensor grad_output, Tensor self, Tensor target, int reduction=Mean) -> Tensor +- func: kl_div_backward(Tensor grad_output, Tensor self, Tensor target, int reduction=Mean, *, bool log_target=False) -> Tensor use_c10_dispatcher: full dispatch: CPU: kl_div_backward_cpu CUDA: kl_div_backward_cuda - func: kthvalue(Tensor self, int k, int dim=-1, bool keepdim=False) -> (Tensor values, Tensor indices) - supports_named_tensor: True + use_c10_dispatcher: full variants: function, method - func: kthvalue.values(Tensor self, int k, int dim=-1, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True dispatch: CPU: kthvalue_out_cpu CUDA: kthvalue_out_cuda - func: kthvalue.dimname(Tensor self, int k, Dimname dim, bool keepdim=False) -> (Tensor values, Tensor indices) - supports_named_tensor: True variants: function, method - func: kthvalue.dimname_out(Tensor self, int k, Dimname dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: layer_norm(Tensor input, int[] normalized_shape, Tensor? weight=None, Tensor? bias=None, float eps=1e-05, bool cudnn_enable=True) -> Tensor - func: native_layer_norm(Tensor input, Tensor? weight, Tensor? bias, int M, int N, float eps) -> (Tensor, Tensor, Tensor) dispatch: @@ -1599,20 +1536,23 @@ python_module: nn dispatch: MkldnnCPU: mkldnn_linear - func: fbgemm_linear_int8_weight_fp32_activation(Tensor input, Tensor weight, Tensor packed, Tensor col_offsets, Scalar weight_scale, Scalar weight_zero_point, Tensor bias) -> Tensor + use_c10_dispatcher: full - func: fbgemm_linear_int8_weight(Tensor input, Tensor weight, Tensor packed, Tensor col_offsets, Scalar weight_scale, Scalar weight_zero_point, Tensor bias) -> Tensor use_c10_dispatcher: full - func: fbgemm_linear_quantize_weight(Tensor input) -> (Tensor, Tensor, float, int) + use_c10_dispatcher: full - func: fbgemm_pack_gemm_matrix_fp16(Tensor input) -> Tensor use_c10_dispatcher: full - func: fbgemm_linear_fp16_weight_fp32_activation(Tensor input, Tensor packed_weight, Tensor bias) -> Tensor + use_c10_dispatcher: full - func: fbgemm_linear_fp16_weight(Tensor input, Tensor packed_weight, Tensor bias) -> Tensor use_c10_dispatcher: full - func: fbgemm_pack_quantized_matrix(Tensor input) -> Tensor @@ -1628,75 +1568,75 @@ CPU: linspace_cpu_out CUDA: linspace_cuda_out - func: log(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: log_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: log.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: log_out CUDA: log_out - func: log10(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: log10_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: log10.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: log10_out CUDA: log10_out - func: log1p(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: log1p_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method dispatch: CPU: log1p_ CUDA: log1p_ SparseCPU: log1p_sparse_ SparseCUDA: log1p_sparse_ - func: log1p.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: log1p_out CUDA: log1p_out SparseCPU: log1p_out_sparse SparseCUDA: log1p_out_sparse - func: log2(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: log2_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: log2.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: log2_out CUDA: log2_out +- func: logaddexp.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) + +- func: logaddexp(Tensor self, Tensor other) -> Tensor + use_c10_dispatcher: full + variants: method, function + +- func: logaddexp2.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) + +- func: logaddexp2(Tensor self, Tensor other) -> Tensor + use_c10_dispatcher: full + variants: method, function + - func: logdet(Tensor self) -> Tensor use_c10_dispatcher: full variants: function, method - func: logspace(Scalar start, Scalar end, int steps=100, float base=10.0, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor @@ -1707,15 +1647,13 @@ CUDA: logspace_cuda_out # log_softmax allows positional dtype, unlike most operators, because kwonly is BC-breaking when loading jit models. - func: log_softmax.int(Tensor self, int dim, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: log_softmax.Dimname(Tensor self, Dimname dim, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: _log_softmax(Tensor self, int dim, bool half_to_float) -> Tensor use_c10_dispatcher: full dispatch: CPU: log_softmax_cpu @@ -1725,34 +1663,50 @@ use_c10_dispatcher: full dispatch: CPU: log_softmax_backward_cpu CUDA: log_softmax_backward_cuda +- func: _logcumsumexp(Tensor self, int dim) -> Tensor + use_c10_dispatcher: full + dispatch: + CPU: _logcumsumexp_cpu + CUDA: _logcumsumexp_cuda + +- func: _logcumsumexp.out(Tensor self, int dim, *, Tensor(a!) out) -> Tensor(a!) + dispatch: + CPU: _logcumsumexp_out_cpu + CUDA: _logcumsumexp_out_cuda + +- func: logcumsumexp(Tensor self, int dim) -> Tensor + variants: function, method + +- func: logcumsumexp.out(Tensor self, int dim, *, Tensor(a!) out) -> Tensor(a!) + +- func: logcumsumexp.dimname(Tensor self, Dimname dim) -> Tensor + variants: function, method + +- func: logcumsumexp.dimname_out(Tensor self, Dimname dim, *, Tensor(a!) out) -> Tensor(a!) + - func: logsumexp(Tensor self, int[1] dim, bool keepdim=False) -> Tensor - supports_named_tensor: True + use_c10_dispatcher: full variants: function, method - func: logsumexp.out(Tensor self, int[1] dim, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: logsumexp.names(Tensor self, Dimname[1] dim, bool keepdim=False) -> Tensor - supports_named_tensor: True variants: function, method - func: logsumexp.names_out(Tensor self, Dimname[1] dim, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: margin_ranking_loss(Tensor input1, Tensor input2, Tensor target, float margin=0.0, int reduction=Mean) -> Tensor use_c10_dispatcher: full - func: matmul(Tensor self, Tensor other) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: matmul.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: matrix_rank.tol(Tensor self, float tol, bool symmetric=False) -> Tensor use_c10_dispatcher: full - func: matrix_rank(Tensor self, bool symmetric=False) -> Tensor @@ -1761,125 +1715,119 @@ - func: matrix_power(Tensor self, int n) -> Tensor use_c10_dispatcher: full variants: function, method - func: max.dim(Tensor self, int dim, bool keepdim=False) -> (Tensor values, Tensor indices) + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: max.dim_max(Tensor self, int dim, bool keepdim=False, *, Tensor(a!) max, Tensor(b!) max_values) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: max_values(Tensor self, int[1] dim, bool keepdim=False) -> Tensor + use_c10_dispatcher: full variants: function, method - func: max.names_dim(Tensor self, Dimname dim, bool keepdim=False) -> (Tensor values, Tensor indices) variants: function, method - supports_named_tensor: True - func: max.names_dim_max(Tensor self, Dimname dim, bool keepdim=False, *, Tensor(a!) max, Tensor(b!) max_values) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: max_values.names(Tensor self, Dimname[1] dim, bool keepdim=False) -> Tensor variants: function, method # Return: (Tensor output, Tensor indices) - func: max_pool1d_with_indices(Tensor self, int[1] kernel_size, int[1] stride=[], int[1] padding=0, int[1] dilation=1, bool ceil_mode=False) -> (Tensor, Tensor) - supports_named_tensor: True + use_c10_dispatcher: full - func: max_pool1d(Tensor self, int[1] kernel_size, int[1] stride=[], int[1] padding=0, int[1] dilation=1, bool ceil_mode=False) -> Tensor - supports_named_tensor: True + use_c10_dispatcher: full - func: max_pool2d(Tensor self, int[2] kernel_size, int[2] stride=[], int[2] padding=0, int[2] dilation=1, bool ceil_mode=False) -> Tensor - supports_named_tensor: True + use_c10_dispatcher: full - func: mkldnn_max_pool2d(Tensor self, int[2] kernel_size, int[2] stride=[], int[2] padding=0, int[2] dilation=1, bool ceil_mode=False) -> Tensor + use_c10_dispatcher: full requires_tensor: True dispatch: MkldnnCPU: mkldnn_max_pool2d - func: quantized_max_pool2d(Tensor self, int[2] kernel_size, int[2] stride=[], int[2] padding=0, int[2] dilation=1, bool ceil_mode=False) -> Tensor + use_c10_dispatcher: full requires_tensor: True dispatch: QuantizedCPU: quantized_max_pool2d - func: max_pool3d(Tensor self, int[3] kernel_size, int[3] stride=[], int[3] padding=0, int[3] dilation=1, bool ceil_mode=False) -> Tensor - supports_named_tensor: True + use_c10_dispatcher: full # The CPU and GPU dispatch variants are named weirdly here because otherwise there # are namespacing issues in C++ - func: mean(Tensor self, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True dispatch: CPU: mean_cpu_gpu CUDA: mean_cpu_gpu QuantizedCPU: quantized_mean_cpu - func: mean.dim(Tensor self, int[1] dim, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True dispatch: CPU: mean_cpu_gpu CUDA: mean_cpu_gpu QuantizedCPU: quantized_mean_cpu + Vulkan: mean_vulkan - func: mean.out(Tensor self, int[1] dim, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: mean_out_cpu_gpu CUDA: mean_out_cpu_gpu QuantizedCPU: quantized_mean_out_cpu - func: mean.names_dim(Tensor self, Dimname[1] dim, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: mean.names_out(Tensor self, Dimname[1] dim, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: median.dim(Tensor self, int dim, bool keepdim=False) -> (Tensor values, Tensor indices) - supports_named_tensor: True + use_c10_dispatcher: full variants: function, method - func: median.dim_values(Tensor self, int dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: median.names_dim(Tensor self, Dimname dim, bool keepdim=False) -> (Tensor values, Tensor indices) - supports_named_tensor: True variants: function, method - func: median.names_dim_values(Tensor self, Dimname dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: min.dim(Tensor self, int dim, bool keepdim=False) -> (Tensor values, Tensor indices) + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: min.dim_min(Tensor self, int dim, bool keepdim=False, *, Tensor(a!) min, Tensor(b!) min_indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: min_values(Tensor self, int[1] dim, bool keepdim=False) -> Tensor + use_c10_dispatcher: full variants: function, method - func: min.names_dim(Tensor self, Dimname dim, bool keepdim=False) -> (Tensor values, Tensor indices) variants: function, method - supports_named_tensor: True - func: min.names_dim_min(Tensor self, Dimname dim, bool keepdim=False, *, Tensor(a!) min, Tensor(b!) min_indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: min_values.names(Tensor self, Dimname[1] dim, bool keepdim=False) -> Tensor variants: function, method - func: mkldnn_convolution(Tensor self, Tensor weight, Tensor? bias, int[] padding, int[] stride, int[] dilation, int groups) -> Tensor - func: mkldnn_convolution_backward_input(int[] self_size, Tensor grad_output, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool bias_defined) -> Tensor + use_c10_dispatcher: full - func: mkldnn_convolution_backward_weights(int[] weight_size, Tensor grad_output, Tensor self, int[] padding, int[] stride, int[] dilation, int groups, bool bias_defined) -> (Tensor, Tensor) + use_c10_dispatcher: full - func: mkldnn_convolution_backward(Tensor self, Tensor grad_output, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool[3] output_mask) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full - func: miopen_batch_norm(Tensor input, Tensor weight, Tensor? bias, Tensor? running_mean, Tensor? running_var, bool training, float exponential_average_factor, float epsilon) -> (Tensor, Tensor, Tensor) dispatch: CUDA: miopen_batch_norm @@ -1890,57 +1838,66 @@ - func: miopen_convolution(Tensor self, Tensor weight, Tensor? bias, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor dispatch: CUDA: miopen_convolution - func: miopen_convolution_backward_input(int[] self_size, Tensor grad_output, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: miopen_convolution_backward_input - func: miopen_convolution_backward(Tensor self, Tensor grad_output, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic, bool[3] output_mask) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full dispatch: CUDA: miopen_convolution_backward - func: miopen_convolution_backward_bias(Tensor grad_output) -> Tensor use_c10_dispatcher: full dispatch: CUDA: miopen_convolution_backward_bias - func: miopen_convolution_backward_weight(int[] weight_size, Tensor grad_output, Tensor self, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: miopen_convolution_backward_weight - func: miopen_convolution_transpose(Tensor self, Tensor weight, Tensor? bias, int[] padding, int[] output_padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor dispatch: CUDA: miopen_convolution_transpose # NB: output_padding not strictly needed here, but it's helpful for the float # backwards - func: miopen_convolution_transpose_backward(Tensor self, Tensor grad_output, Tensor weight, int[] padding, int[] output_padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic, bool[3] output_mask) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full dispatch: CUDA: miopen_convolution_transpose_backward - func: miopen_convolution_transpose_backward_input(Tensor grad_output, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: miopen_convolution_transpose_backward_input - func: miopen_convolution_transpose_backward_weight(int[] weight_size, Tensor grad_output, Tensor self, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: miopen_convolution_transpose_backward_weight - func: miopen_depthwise_convolution(Tensor self, Tensor weight, Tensor? bias, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor dispatch: CUDA: miopen_depthwise_convolution - func: miopen_depthwise_convolution_backward_input(int[] self_size, Tensor grad_output, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: miopen_depthwise_convolution_backward_input - func: miopen_depthwise_convolution_backward(Tensor self, Tensor grad_output, Tensor weight, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic, bool[3] output_mask) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full dispatch: CUDA: miopen_depthwise_convolution_backward - func: miopen_depthwise_convolution_backward_weight(int[] weight_size, Tensor grad_output, Tensor self, int[] padding, int[] stride, int[] dilation, int groups, bool benchmark, bool deterministic) -> Tensor + use_c10_dispatcher: full dispatch: CUDA: miopen_depthwise_convolution_backward_weight - func: miopen_rnn(Tensor input, Tensor[] weight, int weight_stride0, Tensor hx, Tensor? cx, int mode, int hidden_size, int num_layers, bool batch_first, float dropout, bool train, bool bidirectional, int[] batch_sizes, Tensor? dropout_state) -> (Tensor, Tensor, Tensor, Tensor, Tensor) dispatch: @@ -1953,69 +1910,61 @@ - func: mm(Tensor self, Tensor mat2) -> Tensor use_c10_dispatcher: full variants: function, method dispatch: CPU: mm_cpu - CUDA: legacy::cuda::_th_mm + CUDA: mm_cuda SparseCPU: _sparse_mm SparseCUDA: _sparse_mm - supports_named_tensor: True - func: mm.out(Tensor self, Tensor mat2, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: mm_cpu_out - CUDA: legacy::cuda::_th_mm_out + CUDA: mm_out_cuda SparseCPU: _sparse_mm_out SparseCUDA: _sparse_mm_out - supports_named_tensor: True - func: _sparse_mm(Tensor sparse, Tensor dense) -> Tensor use_c10_dispatcher: full - func: mode(Tensor self, int dim=-1, bool keepdim=False) -> (Tensor values, Tensor indices) - supports_named_tensor: True + use_c10_dispatcher: full variants: function, method - func: mode.values(Tensor self, int dim=-1, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: mode.dimname(Tensor self, Dimname dim, bool keepdim=False) -> (Tensor values, Tensor indices) variants: function, method - supports_named_tensor: True - func: mode.dimname_out(Tensor self, Dimname dim, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) - supports_named_tensor: True - func: mul.Tensor(Tensor self, Tensor other) -> Tensor use_c10_dispatcher: full variants: function, method dispatch: CPU: mul CUDA: mul SparseCPU: mul_sparse SparseCUDA: mul_sparse MkldnnCPU: mkldnn_mul - supports_named_tensor: True - func: mul_.Tensor(Tensor(a!) self, Tensor other) -> Tensor(a!) variants: method dispatch: CPU: mul_ CUDA: mul_ SparseCPU: mul_sparse_ SparseCUDA: mul_sparse_ MkldnnCPU: mkldnn_mul_ - supports_named_tensor: True - func: mul.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: mul_out CUDA: mul_out SparseCPU: mul_out_sparse_cpu SparseCUDA: mul_out_sparse_cuda MkldnnCPU: mkldnn_mul_out - supports_named_tensor: True # For C++ only, until we have conversion from C++ numbers to Tensor - func: mul.Scalar(Tensor self, Scalar other) -> Tensor use_c10_dispatcher: full variants: function, method @@ -2025,19 +1974,16 @@ - func: mv(Tensor self, Tensor vec) -> Tensor use_c10_dispatcher: full variants: function, method dispatch: - CPU: mv_cpu - CUDA: legacy::cuda::_th_mv - supports_named_tensor: True + CPU: mv + CUDA: mv + SparseCPU: mv_sparse + SparseCUDA: mv_sparse - func: mv.out(Tensor self, Tensor vec, *, Tensor(a!) out) -> Tensor(a!) - dispatch: - CPU: mv_cpu_out - CUDA: legacy::cuda::_th_mv_out - supports_named_tensor: True - func: mvlgamma(Tensor self, int p) -> Tensor use_c10_dispatcher: full variants: function, method @@ -2052,18 +1998,18 @@ CUDA: narrow_copy_dense SparseCPU: narrow_copy_sparse SparseCUDA: narrow_copy_sparse - func: narrow(Tensor(a) self, int dim, int start, int length) -> Tensor(a) + use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: narrow.Tensor(Tensor(a) self, int dim, Tensor start, int length) -> Tensor(a) + use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: native_batch_norm(Tensor input, Tensor? weight, Tensor? bias, Tensor? running_mean, Tensor? running_var, bool training, float momentum, float eps) -> (Tensor, Tensor, Tensor) dispatch: CPU: batch_norm_cpu CUDA: batch_norm_cuda @@ -2072,10 +2018,11 @@ - func: native_batch_norm.out(Tensor input, Tensor? weight, Tensor? bias, Tensor? running_mean, Tensor? running_var, bool training, float momentum, float eps, *, Tensor(a!) out, Tensor(b!) save_mean, Tensor(c!) save_invstd) -> (Tensor(a!), Tensor(b!), Tensor(c!)) dispatch: CUDA: batch_norm_cuda_out - func: batch_norm_stats(Tensor input, float eps) -> (Tensor, Tensor) + use_c10_dispatcher: full dispatch: CUDA: batch_norm_stats_cuda - func: batch_norm_elemt(Tensor input, Tensor? weight, Tensor? bias, Tensor mean, Tensor invstd, float eps) -> Tensor dispatch: @@ -2088,11 +2035,11 @@ # for backward compatibility - func: batch_norm_gather_stats(Tensor input, Tensor mean, Tensor invstd, Tensor? running_mean, Tensor? running_var, float momentum, float eps, int count) -> (Tensor, Tensor) dispatch: CUDA: batch_norm_gather_stats_cuda -- func: batch_norm_gather_stats_with_counts(Tensor input, Tensor mean, Tensor invstd, Tensor? running_mean, Tensor? running_var, float momentum, float eps, int[] counts) -> (Tensor, Tensor) +- func: batch_norm_gather_stats_with_counts(Tensor input, Tensor mean, Tensor invstd, Tensor? running_mean, Tensor? running_var, float momentum, float eps, Tensor counts) -> (Tensor, Tensor) dispatch: CUDA: batch_norm_gather_stats_with_counts_cuda - func: native_batch_norm_backward(Tensor grad_out, Tensor input, Tensor? weight, Tensor? running_mean, Tensor? running_var, Tensor? save_mean, Tensor? save_invstd, bool train, float eps, bool[3] output_mask) -> (Tensor, Tensor, Tensor) dispatch: @@ -2110,44 +2057,51 @@ - func: batch_norm_update_stats(Tensor input, Tensor? running_mean, Tensor? running_var, float momentum) -> (Tensor, Tensor) dispatch: CPU: batch_norm_update_stats_cpu CUDA: batch_norm_update_stats_cuda +- func: is_vulkan_available() -> bool + use_c10_dispatcher: full + - func: _nnpack_available() -> bool use_c10_dispatcher: full - func: _nnpack_spatial_convolution(Tensor input, Tensor weight, Tensor? bias, int[2] padding, int[2] stride=1) -> Tensor variants: function - func: _nnpack_spatial_convolution_backward(Tensor input, Tensor grad_output, Tensor weight, int[2] padding, bool[3] output_mask) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full variants: function - func: _nnpack_spatial_convolution_backward_input(Tensor input, Tensor grad_output, Tensor weight, int[2] padding) -> Tensor + use_c10_dispatcher: full variants: function - func: _nnpack_spatial_convolution_backward_weight(Tensor input, int[] weightsize, Tensor grad_output, int[2] padding) -> Tensor + use_c10_dispatcher: full variants: function - func: ones.names(int[] size, *, Dimname[]? names, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor device_guard: False - func: ones(int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor - func: ones.out(int[] size, *, Tensor(a!) out) -> Tensor(a!) - func: ones_like(Tensor self, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor - supports_named_tensor: True - func: pairwise_distance(Tensor x1, Tensor x2, float p=2, float eps=1e-06, bool keepdim=False) -> Tensor use_c10_dispatcher: full - func: cdist(Tensor x1, Tensor x2, float p=2, int? compute_mode=None) -> Tensor - supports_named_tensor: True + use_c10_dispatcher: full +- func: _euclidean_dist(Tensor x1, Tensor x2) -> Tensor + use_c10_dispatcher: full + - func: _cdist_forward(Tensor x1, Tensor x2, float p, int? compute_mode) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True - func: _cdist_backward(Tensor grad, Tensor x1, Tensor x2, float p, Tensor cdist) -> Tensor use_c10_dispatcher: full - func: pdist(Tensor self, float p=2) -> Tensor @@ -2162,29 +2116,36 @@ - func: cosine_similarity(Tensor x1, Tensor x2, int dim=1, float eps=1e-08) -> Tensor use_c10_dispatcher: full variants: function - func: permute(Tensor(a) self, int[] dims) -> Tensor(a) + use_c10_dispatcher: full variants: method # This is method-only to match the previous tensor API. In the future we could make this a function too. # Only exposed from C++ -- in Python, # we expose it as an attribute `T`, not a function. # # I'd like to name this "T" in C++ too, but # calling a native function "T" causes undefined # behavior on Windows, for reasons I don't understand # (maybe related to capital letter collation somehow...) - func: numpy_T(Tensor(a) self) -> Tensor(a) + use_c10_dispatcher: full variants: method - func: pixel_shuffle(Tensor self, int upscale_factor) -> Tensor use_c10_dispatcher: full +- func: channel_shuffle(Tensor self, int groups) -> Tensor + use_c10_dispatcher: full + dispatch: + CPU: channel_shuffle + QuantizedCPU: quantized_channel_shuffle + - func: is_pinned(Tensor self) -> bool use_c10_dispatcher: full variants: method - supports_named_tensor: True - func: pin_memory(Tensor self) -> Tensor use_c10_dispatcher: full variants: method @@ -2194,10 +2155,34 @@ - func: poisson_nll_loss(Tensor input, Tensor target, bool log_input, bool full, float eps, int reduction) -> Tensor use_c10_dispatcher: full variants: function +- func: rad2deg(Tensor self) -> Tensor + use_c10_dispatcher: full + variants: function, method + supports_named_tensor: True + +- func: rad2deg_(Tensor(a!) self) -> Tensor(a!) + variants: function, method + supports_named_tensor: True + +- func: rad2deg.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) + supports_named_tensor: True + +- func: deg2rad(Tensor self) -> Tensor + use_c10_dispatcher: full + variants: function, method + supports_named_tensor: True + +- func: deg2rad_(Tensor(a!) self) -> Tensor(a!) + variants: function, method + supports_named_tensor: True + +- func: deg2rad.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) + supports_named_tensor: True + - func: scalar_tensor(Scalar s, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor - func: rand.names(int[] size, *, Dimname[]? names, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor device_guard: False @@ -2211,11 +2196,10 @@ - func: rand.out(int[] size, *, Tensor(a!) out) -> Tensor(a!) - func: rand.generator_out(int[] size, *, Generator? generator, Tensor(a!) out) -> Tensor(a!) - func: rand_like(Tensor self, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor - supports_named_tensor: True - func: randint(int high, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor - func: randint.generator(int high, int[] size, *, Generator? generator, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor @@ -2248,11 +2232,10 @@ - func: randn.out(int[] size, *, Tensor(a!) out) -> Tensor(a!) - func: randn.generator_out(int[] size, *, Generator? generator, Tensor(a!) out) -> Tensor(a!) - func: randn_like(Tensor self, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor - supports_named_tensor: True - func: randperm(int n, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor - func: randperm.generator(int n, *, Generator? generator, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor @@ -2272,36 +2255,31 @@ CPU: range_cpu_out CUDA: range_cuda_out - func: reciprocal(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: reciprocal_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: reciprocal.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: neg(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: neg_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: neg.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: neg_out CUDA: neg_out - func: repeat(Tensor self, int[] repeats) -> Tensor + use_c10_dispatcher: full variants: method # This is method-only to match the previous tensor API. In the future we could make this a function too. - func: repeat_interleave.Tensor(Tensor repeats) -> Tensor use_c10_dispatcher: full variants: function @@ -2316,15 +2294,16 @@ - func: repeat_interleave.self_int(Tensor self, int repeats, int? dim=None) -> Tensor use_c10_dispatcher: full variants: function, method - func: reshape(Tensor self, int[] shape) -> Tensor + use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: _mkldnn_reshape(Tensor self, int[] shape) -> Tensor + use_c10_dispatcher: full device_guard: False requires_tensor: True dispatch: MkldnnCPU: mkldnn_reshape @@ -2333,19 +2312,16 @@ variants: method device_guard: False - func: round(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: round_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: round.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: round_out CUDA: round_out - func: rrelu(Tensor self, Scalar lower=0.125, Scalar upper=0.3333333333333333, bool training=False, Generator? generator=None) -> Tensor @@ -2358,14 +2334,12 @@ dispatch: CPU: relu CUDA: relu MkldnnCPU: mkldnn_relu QuantizedCPU: quantized_relu - supports_named_tensor: True - func: relu_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method dispatch: CPU: relu_ CUDA: relu_ MkldnnCPU: mkldnn_relu_ @@ -2377,10 +2351,11 @@ dispatch: CPU: prelu_cpu CUDA: prelu_cuda - func: prelu_backward(Tensor grad_output, Tensor self, Tensor weight) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function, method dispatch: CPU: prelu_backward_cpu CUDA: prelu_backward_cuda @@ -2406,32 +2381,28 @@ use_c10_dispatcher: full variants: function, method - func: rsqrt(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: rsqrt_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: rsqrt.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: rsqrt_out CUDA: rsqrt_out - func: select.Dimname(Tensor(a) self, Dimname dim, int index) -> Tensor(a) variants: function, method device_guard: False - supports_named_tensor: True - func: select.int(Tensor(a) self, int dim, int index) -> Tensor(a) + use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: selu(Tensor self) -> Tensor use_c10_dispatcher: full - func: selu_(Tensor(a!) self) -> Tensor(a!) @@ -2439,58 +2410,48 @@ - func: celu(Tensor self, Scalar alpha=1.0) -> Tensor use_c10_dispatcher: full - func: celu_(Tensor(a!) self, Scalar alpha=1.0) -> Tensor(a!) - - func: sigmoid(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method dispatch: CPU: sigmoid CUDA: sigmoid QuantizedCPU: quantized_sigmoid MkldnnCPU: mkldnn_sigmoid - func: sigmoid_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method dispatch: CPU: sigmoid_ CUDA: sigmoid_ MkldnnCPU: mkldnn_sigmoid_ - func: sigmoid.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: sin(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: sin_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: sin.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: sin_out CUDA: sin_out - func: sinh(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: sinh_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: sinh.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True # Returns a copy of this `Variable` that is detached from its autograd graph. # This method is OK to call if the `Variable` is a view. # # NOTE: Previously, if we change the tensor metadata (e.g. sizes / strides / @@ -2502,52 +2463,47 @@ # changing metadata of the detached tensor and expecting the original tensor to also # be updated. - func: detach(Tensor self) -> Tensor use_c10_dispatcher: full manual_kernel_registration: True - supports_named_tensor: True variants: function, method # Like `detach()`, but modifies this `Variable` in-place. This method may # only be called on non-view `Variable`s. You can use `is_view()` to check # this. If this `Variable` is a view, throws an `std::runtime_error()`. - func: detach_(Tensor(a!) self) -> Tensor(a!) manual_kernel_registration: True - supports_named_tensor: True variants: function, method - func: size.int(Tensor self, int dim) -> int use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: size.Dimname(Tensor self, Dimname dim) -> int variants: function, method device_guard: False - supports_named_tensor: True - func: slice.Tensor(Tensor(a) self, int dim=0, int start=0, int end=9223372036854775807, int step=1) -> Tensor(a) + use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: slogdet(Tensor self) -> (Tensor sign, Tensor logabsdet) + use_c10_dispatcher: full variants: function, method - func: smm(Tensor self, Tensor mat2) -> Tensor use_c10_dispatcher: full variants: function, method # softmax allows positional dtype, unlike most operators, because kwonly is BC-breaking when loading jit models. - func: softmax.int(Tensor self, int dim, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: softmax.Dimname(Tensor self, Dimname dim, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: _softmax(Tensor self, int dim, bool half_to_float) -> Tensor use_c10_dispatcher: full dispatch: CPU: softmax_cpu @@ -2559,31 +2515,30 @@ dispatch: CPU: softmax_backward_cpu CUDA: softmax_backward_cuda - func: split.Tensor(Tensor(a) self, int split_size, int dim=0) -> Tensor(a)[] + use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: split_with_sizes(Tensor self, int[] split_sizes, int dim=0) -> Tensor[] + use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: squeeze(Tensor(a) self) -> Tensor(a) - supports_named_tensor: True + use_c10_dispatcher: full variants: function, method device_guard: False - func: squeeze.dim(Tensor(a) self, int dim) -> Tensor(a) - supports_named_tensor: True + use_c10_dispatcher: full variants: function, method device_guard: False - func: squeeze.dimname(Tensor(a) self, Dimname dim) -> Tensor(a) - supports_named_tensor: True variants: function, method device_guard: False - func: squeeze_(Tensor(a!) self) -> Tensor(a!) variants: method @@ -2607,194 +2562,157 @@ CUDA: _sspaddmm_out_only_sparse_cuda SparseCPU: _sspaddmm_out_cpu SparseCUDA: _sspaddmm_out_cuda - func: stack(Tensor[] tensors, int dim=0) -> Tensor + use_c10_dispatcher: full - func: stack.out(Tensor[] tensors, int dim=0, *, Tensor(a!) out) -> Tensor(a!) # The signature is designed to be consistent with librosa except that it is # missing the `pad_mode` and `center` arguments, which are taken care of at # `torch.functional.py`. They shall be moved here once we have mapping between # Python strings and C++ Enum in codegen. - func: stft(Tensor self, int n_fft, int? hop_length=None, int? win_length=None, Tensor? window=None, bool normalized=False, bool onesided=True) -> Tensor variants: function, method +- func: istft(Tensor self, int n_fft, int? hop_length=None, int? win_length=None, Tensor? window=None, bool center=True, bool normalized=False, bool onesided=True, int? length=None) -> Tensor + variants: function, method + - func: stride.int(Tensor self, int dim) -> int use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: stride.Dimname(Tensor self, Dimname dim) -> int variants: function, method device_guard: False - supports_named_tensor: True - func: sum(Tensor self, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: sum.dim_IntList(Tensor self, int[1] dim, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: sum.dim_DimnameList(Tensor self, Dimname[1] dim, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: sum.IntList_out(Tensor self, int[1] dim, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: sum.DimnameList_out(Tensor self, Dimname[1] dim, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: sum_to_size(Tensor self, int[] size) -> Tensor + use_c10_dispatcher: full variants: method device_guard: False - func: sqrt(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: sqrt_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: sqrt.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: square(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: square_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: std(Tensor self, bool unbiased=True) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: std.dim(Tensor self, int[1] dim, bool unbiased=True, bool keepdim=False) -> Tensor + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: std_mean(Tensor self, bool unbiased=True) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function - supports_named_tensor: True - func: std_mean.dim(Tensor self, int[1] dim, bool unbiased=True, bool keepdim=False) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function - supports_named_tensor: True - func: std_mean.names_dim(Tensor self, Dimname[1] dim, bool unbiased=True, bool keepdim=False) -> (Tensor, Tensor) variants: function - supports_named_tensor: True - func: std.out(Tensor self, int[1] dim, bool unbiased=True, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: std.names_dim(Tensor self, Dimname[1] dim, bool unbiased=True, bool keepdim=False) -> Tensor variants: function, method - supports_named_tensor: True - func: std.names_out(Tensor self, Dimname[1] dim, bool unbiased=True, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: prod(Tensor self, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: prod.dim_int(Tensor self, int dim, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: prod.int_out(Tensor self, int dim, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: prod.dim_Dimname(Tensor self, Dimname dim, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor variants: function, method - supports_named_tensor: True - func: prod.Dimname_out(Tensor self, Dimname dim, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - - func: t(Tensor(a) self) -> Tensor(a) + use_c10_dispatcher: full device_guard: False variants: function, method - supports_named_tensor: True - func: t_(Tensor(a!) self) -> Tensor(a!) device_guard: False variants: method - func: tan(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: tan_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _tan__cpu - CUDA: _tan__cuda - func: tan.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _tan_out_cpu - CUDA: _tan_out_cuda - func: tanh(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method dispatch: CPU: tanh CUDA: tanh QuantizedCPU: quantized_tanh - func: tanh_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - dispatch: - CPU: _tanh__cpu - CUDA: _tanh__cuda - func: tanh.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - dispatch: - CPU: _tanh_out_cpu - CUDA: _tanh_out_cuda - func: tensordot(Tensor self, Tensor other, int[] dims_self, int[] dims_other) -> Tensor + use_c10_dispatcher: full variants: function # TODO: namespace threshold in 'nn' - func: threshold(Tensor self, Scalar threshold, Scalar value) -> Tensor use_c10_dispatcher: full variants: function - supports_named_tensor: True dispatch: CPU: threshold CUDA: threshold_cuda + QuantizedCPU: quantized_threshold - func: threshold_(Tensor(a!) self, Scalar threshold, Scalar value) -> Tensor(a!) variants: function - supports_named_tensor: True dispatch: CPU: threshold_ CUDA: threshold__cuda - func: threshold.out(Tensor self, Scalar threshold, Scalar value, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: threshold_out CUDA: threshold_out_cuda - func: threshold_backward(Tensor grad_output, Tensor self, Scalar threshold) -> Tensor @@ -2803,18 +2721,17 @@ dispatch: CPU: threshold_backward CUDA: threshold_backward_cuda - func: transpose.int(Tensor(a) self, int dim0, int dim1) -> Tensor(a) + use_c10_dispatcher: full variants: function, method device_guard: False - supports_named_tensor: True - func: transpose.Dimname(Tensor(a) self, Dimname dim0, Dimname dim1) -> Tensor(a) variants: function, method device_guard: False - supports_named_tensor: True - func: _mkldnn_transpose(Tensor self, int dim0, int dim1) -> Tensor use_c10_dispatcher: full device_guard: False requires_tensor: True @@ -2835,33 +2752,45 @@ use_c10_dispatcher: full python_module: nn variants: function - func: flip(Tensor self, int[] dims) -> Tensor + use_c10_dispatcher: full variants: function, method dispatch: CPU: flip_cpu CUDA: flip_cuda +- func: fliplr(Tensor self) -> Tensor + use_c10_dispatcher: full + variants: function, method + +- func: flipud(Tensor self) -> Tensor + use_c10_dispatcher: full + variants: function, method + - func: roll(Tensor self, int[1] shifts, int[1] dims=[]) -> Tensor + use_c10_dispatcher: full variants: function, method dispatch: CPU: roll_cpu CUDA: roll_cuda # default int[] value [0,1] should not add space after comma, since native_parse.py uses ', ' to split args - func: rot90(Tensor self, int k=1, int[] dims=[0,1]) -> Tensor + use_c10_dispatcher: full variants: function, method - func: trapz.x(Tensor y, Tensor x, *, int dim=-1) -> Tensor use_c10_dispatcher: full - func: trapz.dx(Tensor y, *, float dx=1, int dim=-1) -> Tensor use_c10_dispatcher: full - func: _trilinear(Tensor i1, Tensor i2, Tensor i3, int[] expand1, int[] expand2, int[] expand3, int[] sumdim, int unroll_dim=1) -> Tensor + use_c10_dispatcher: full - func: triplet_margin_loss(Tensor anchor, Tensor positive, Tensor negative, float margin=1.0, float p=2, float eps=1e-06, bool swap=False, int reduction=Mean) -> Tensor use_c10_dispatcher: full - func: true_divide.Tensor(Tensor self, Tensor other) -> Tensor @@ -2870,49 +2799,41 @@ dispatch: CPU: true_divide CUDA: true_divide SparseCPU: true_divide_sparse SparseCUDA: true_divide_sparse - supports_named_tensor: True - func: true_divide_.Tensor(Tensor(a!) self, Tensor other) -> Tensor(a!) variants: method dispatch: CPU: true_divide_ CUDA: true_divide_ SparseCPU: true_divide_sparse_ SparseCUDA: true_divide_sparse_ - supports_named_tensor: True - func: true_divide.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: true_divide_out CUDA: true_divide_out SparseCPU: true_divide_out_sparse_zerodim SparseCUDA: true_divide_out_sparse_zerodim - supports_named_tensor: True - func: true_divide.Scalar(Tensor self, Scalar other) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: true_divide_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!) variants: method - supports_named_tensor: True - func: trunc(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: trunc_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: function, method - func: trunc.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: trunc_out CUDA: trunc_out - func: type_as(Tensor self, Tensor other) -> Tensor @@ -2922,83 +2843,88 @@ - func: _has_compatible_shallow_copy_type(Tensor self, Tensor from) -> bool use_c10_dispatcher: full variants: function - func: _unique(Tensor self, bool sorted=True, bool return_inverse=False) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: _unique_cpu CUDA: _unique_cuda - func: unique_dim(Tensor self, int dim, bool sorted=True, bool return_inverse=False, bool return_counts=False) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: unique_dim_cpu CUDA: unique_dim_cuda - func: unique_consecutive(Tensor self, bool return_inverse=False, bool return_counts=False, int? dim=None) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: unique_consecutive_cpu CUDA: unique_consecutive_cuda - func: unique_dim_consecutive(Tensor self, int dim, bool return_inverse=False, bool return_counts=False) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: unique_dim_consecutive_cpu CUDA: unique_dim_consecutive_cuda # _unique and _unique_dim are fragile and modifying them easily cause internal break # the below operator is a temporary hack for adding return_counts support # Please don't rely on these two operators, they will be removed soon - func: _unique2(Tensor self, bool sorted=True, bool return_inverse=False, bool return_counts=False) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: _unique2_cpu CUDA: _unique2_cuda - func: _unsafe_view(Tensor self, int[] size) -> Tensor + use_c10_dispatcher: full - func: unsqueeze(Tensor(a) self, int dim) -> Tensor(a) + use_c10_dispatcher: full variants: function, method device_guard: False - func: unsqueeze_(Tensor(a!) self, int dim) -> Tensor(a!) variants: method device_guard: False +- func: vander(Tensor x, int? N=None, bool increasing=False) -> Tensor + use_c10_dispatcher: full + - func: var(Tensor self, bool unbiased=True) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: var.dim(Tensor self, int[1] dim, bool unbiased=True, bool keepdim=False) -> Tensor + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: var.out(Tensor self, int[1] dim, bool unbiased=True, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: var.names_dim(Tensor self, Dimname[1] dim, bool unbiased=True, bool keepdim=False) -> Tensor variants: function, method - supports_named_tensor: True - func: var.names_out(Tensor self, Dimname[1] dim, bool unbiased=True, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: var_mean(Tensor self, bool unbiased=True) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function - supports_named_tensor: True - func: var_mean.dim(Tensor self, int[1] dim, bool unbiased=True, bool keepdim=False) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function - supports_named_tensor: True - func: var_mean.names_dim(Tensor self, Dimname[1] dim, bool unbiased=True, bool keepdim=False) -> (Tensor, Tensor) variants: function - supports_named_tensor: True - func: view_as(Tensor self, Tensor other) -> Tensor use_c10_dispatcher: full variants: method device_guard: False @@ -3009,47 +2935,51 @@ - func: where.self(Tensor condition, Tensor self, Tensor other) -> Tensor use_c10_dispatcher: full variants: function, method - func: where(Tensor condition) -> Tensor[] + use_c10_dispatcher: full variants: function - func: _s_where(Tensor condition, Tensor self, Tensor other) -> Tensor use_c10_dispatcher: full variants: function - func: norm_except_dim(Tensor v, int pow=2, int dim=0) -> Tensor + use_c10_dispatcher: full variants: function # VariableType::_weight_norm does not want to be given a gap in the autograd graph, # so we don't define "dispatch" variants for it. - func: _weight_norm(Tensor v, Tensor g, int dim=0) -> Tensor use_c10_dispatcher: full variants: function - func: _weight_norm_cuda_interface(Tensor v, Tensor g, int dim=0) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CUDA: weight_norm_cuda - func: _weight_norm_cuda_interface_backward(Tensor grad_w, Tensor saved_v, Tensor saved_g, Tensor saved_norms, int dim) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CUDA: weight_norm_cuda_backward - func: _weight_norm_differentiable_backward(Tensor grad_w, Tensor saved_v, Tensor saved_g, Tensor saved_norms, int dim) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function - func: zeros.names(int[] size, *, Dimname[]? names, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor device_guard: False - func: zeros(int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor - func: zeros.out(int[] size, *, Tensor(a!) out) -> Tensor(a!) - func: zeros_like(Tensor self, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None, MemoryFormat? memory_format=None) -> Tensor - supports_named_tensor: True - func: _standard_gamma_grad(Tensor self, Tensor output) -> Tensor use_c10_dispatcher: full variants: function dispatch: @@ -3077,10 +3007,15 @@ - func: poisson(Tensor self, Generator? generator=None) -> Tensor dispatch: CPU: _s_poisson_cpu CUDA: _s_poisson_cuda +- func: binomial(Tensor count, Tensor prob, Generator? generator=None) -> Tensor + dispatch: + CPU: _s_binomial_cpu + CUDA: _s_binomial_cuda + # When more variants get ported to native, this dispatch will get more # complicated - func: native_norm(Tensor self, Scalar p=2) -> Tensor use_c10_dispatcher: full @@ -3093,18 +3028,50 @@ use_c10_dispatcher: full - func: _sparse_sum.dtype(Tensor self, *, ScalarType dtype) -> Tensor - func: _sparse_sum.dim(Tensor self, int[1] dim) -> Tensor + use_c10_dispatcher: full - func: _sparse_sum.dim_dtype(Tensor self, int[1] dim, *, ScalarType dtype) -> Tensor - func: _sparse_sum_backward(Tensor grad, Tensor self, int[] dim) -> Tensor + use_c10_dispatcher: full dispatch: SparseCPU: _sparse_sum_backward_cpu SparseCUDA: _sparse_sum_backward_cuda +- func: _sparse_softmax.int(Tensor self, int dim, ScalarType? dtype=None) -> Tensor + variants: function + +- func: _sparse_softmax.Dimname(Tensor self, Dimname dim, *, ScalarType? dtype=None) -> Tensor + variants: function + +- func: _sparse_softmax(Tensor self, int dim, bool half_to_float) -> Tensor + use_c10_dispatcher: full + dispatch: + SparseCPU: softmax_sparse_cpu + +- func: _sparse_softmax_backward_data(Tensor grad_output, Tensor output, int dim, Tensor self) -> Tensor + dispatch: + SparseCPU: softmax_backward_sparse_cpu + +- func: _sparse_log_softmax.int(Tensor self, int dim, ScalarType? dtype=None) -> Tensor + variants: function + +- func: _sparse_log_softmax.Dimname(Tensor self, Dimname dim, *, ScalarType? dtype=None) -> Tensor + variants: function + +- func: _sparse_log_softmax(Tensor self, int dim, bool half_to_float) -> Tensor + use_c10_dispatcher: full + dispatch: + SparseCPU: log_softmax_sparse_cpu + +- func: _sparse_log_softmax_backward_data(Tensor grad_output, Tensor output, int dim, Tensor self) -> Tensor + dispatch: + SparseCPU: log_softmax_backward_sparse_cpu + - func: norm.ScalarOpt_dtype(Tensor self, Scalar? p, *, ScalarType dtype) -> Tensor variants: function, method - func: norm.Scalar(Tensor self, Scalar p=2) -> Tensor use_c10_dispatcher: full @@ -3112,10 +3079,11 @@ - func: norm.ScalarOpt_dim_dtype(Tensor self, Scalar? p, int[1] dim, bool keepdim, *, ScalarType dtype) -> Tensor variants: function, method - func: norm.ScalarOpt_dim(Tensor self, Scalar? p, int[1] dim, bool keepdim=False) -> Tensor + use_c10_dispatcher: full variants: function, method - func: norm.dtype_out(Tensor self, Scalar? p, int[1] dim, bool keepdim, *, ScalarType dtype, Tensor(a!) out) -> Tensor(a!) - func: norm.out(Tensor self, Scalar? p, int[1] dim, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!) @@ -3133,10 +3101,11 @@ - func: frobenius_norm(Tensor self) -> Tensor use_c10_dispatcher: full variants: function - func: frobenius_norm.dim(Tensor self, int[1] dim, bool keepdim=False) -> Tensor + use_c10_dispatcher: full variants: function - func: frobenius_norm.out(Tensor self, int[1] dim, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!) variants: function @@ -3146,10 +3115,11 @@ - func: nuclear_norm.out(Tensor self, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!) variants: function - func: nuclear_norm.dim(Tensor self, int[2] dim, bool keepdim=False) -> Tensor + use_c10_dispatcher: full variants: function - func: nuclear_norm.dim_out(Tensor self, int[2] dim, bool keepdim=False, *, Tensor(a!) out) -> Tensor(a!) variants: function @@ -3160,37 +3130,33 @@ CUDA: clone SparseCPU: clone_sparse SparseCUDA: clone_sparse MkldnnCPU: mkldnn_clone QuantizedCPU: quantized_clone - supports_named_tensor: True + QuantizedCUDA: quantized_clone - func: resize_as_(Tensor(a!) self, Tensor the_template, *, MemoryFormat? memory_format=None) -> Tensor(a!) manual_kernel_registration: True - supports_named_tensor: True variants: function, method - func: pow.Tensor_Scalar_out(Tensor self, Scalar exponent, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: pow_out CUDA: pow_out SparseCPU: pow_out_sparse_scalar SparseCUDA: pow_out_sparse_scalar - func: pow.Tensor_Scalar(Tensor self, Scalar exponent) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True dispatch: CPU: pow CUDA: pow SparseCPU: pow_sparse_scalar SparseCUDA: pow_sparse_scalar - func: zero_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: method, function dispatch: CPU: zero_ CUDA: zero_ SparseCPU: zero_sparse_ @@ -3201,88 +3167,77 @@ dispatch: CPU: sub_out CUDA: sub_out SparseCPU: sub_out_sparse SparseCUDA: sub_out_sparse - supports_named_tensor: True - func: sub.Tensor(Tensor self, Tensor other, *, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: function, method dispatch: CPU: sub CUDA: sub SparseCPU: sub_sparse SparseCUDA: sub_sparse - supports_named_tensor: True - func: sub_.Tensor(Tensor(a!) self, Tensor other, *, Scalar alpha=1) -> Tensor(a!) variants: method dispatch: CPU: sub_ CUDA: sub_ SparseCPU: sub_sparse_ SparseCUDA: sub_sparse_ - supports_named_tensor: True # For C++ only, until we have conversion from C++ numbers to Tensor - func: sub.Scalar(Tensor self, Scalar other, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: sub_.Scalar(Tensor(a!) self, Scalar other, Scalar alpha=1) -> Tensor(a!) variants: method - supports_named_tensor: True - func: rsub.Tensor(Tensor self, Tensor other, *, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: function - supports_named_tensor: True # For C++ only, until we have conversion from C++ numbers to Tensor - func: rsub.Scalar(Tensor self, Scalar other, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: function - supports_named_tensor: True # Functionally the same as addmm, but we give it a different derivative formula # that doesn't propagate gradients to non-present entries on sparse. - func: _sparse_addmm(Tensor self, Tensor sparse, Tensor dense, *, Scalar beta=1, Scalar alpha=1) -> Tensor use_c10_dispatcher: full - named_guard: False - func: addmm.out(Tensor self, Tensor mat1, Tensor mat2, *, Scalar beta=1, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!) dispatch: - CPU: legacy::cpu::_th_addmm_out - CUDA: legacy::cuda::_th_addmm_out + CPU: addmm_cpu_out + CUDA: addmm_out_cuda SparseCPU: addmm_out_sparse_dense_cpu SparseCUDA: addmm_out_sparse_dense_cuda - supports_named_tensor: True - func: addmm(Tensor self, Tensor mat1, Tensor mat2, *, Scalar beta=1, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: function, method dispatch: - CPU: legacy::cpu::_th_addmm - CUDA: legacy::cuda::_th_addmm + CPU: addmm_cpu + CUDA: addmm_cuda SparseCPU: addmm_sparse_dense_cpu SparseCUDA: addmm_sparse_dense_cuda - supports_named_tensor: True + Vulkan: vulkan_addmm - func: addmm_(Tensor(a!) self, Tensor mat1, Tensor mat2, *, Scalar beta=1, Scalar alpha=1) -> Tensor(a!) variants: method dispatch: CPU: legacy::cpu::_th_addmm_ - CUDA: legacy::cuda::_th_addmm_ + CUDA: addmm__cuda # Warning! For whatever reason, the inplace sparse addmm is NON # broadcasting SparseCPU: s_addmm_sparse_dense_cpu_ SparseCUDA: s_addmm_sparse_dense_cuda_ - supports_named_tensor: True - # NOTE [ Sparse: autograd and API ] # # # Sparse Tensor Constructors # ~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -3394,11 +3349,10 @@ # `indices()` and `_indices()`. We mark their outputs as non-differentiable, so # the view relation is not tracked by autograd, but the version counter is still # shared. In other words, their outputs are non-differentiable views of the # sparse tensor. - # FIXME: would be nicer if TensorOptions was optional based; not adding default arguments for options given # the default would never make sense. - func: sparse_coo_tensor.size(int[] size, *, ScalarType dtype, Layout layout, Device device, bool pin_memory=False) -> Tensor - func: sparse_coo_tensor.indices(Tensor indices, Tensor values, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor @@ -3431,20 +3385,18 @@ dispatch: SparseCPU: sparse_resize_and_clear_ SparseCUDA: sparse_resize_and_clear_ requires_tensor: True - - func: sparse_mask(Tensor self, Tensor mask) -> Tensor use_c10_dispatcher: full variants: method dispatch: SparseCPU: sparse_mask_cpu SparseCUDA: sparse_mask_cuda requires_tensor: True - - func: to_dense(Tensor self) -> Tensor use_c10_dispatcher: full variants: method dispatch: SparseCPU: sparse_to_dense @@ -3472,11 +3424,10 @@ SparseCPU: sparse_dim_sparse SparseCUDA: sparse_dim_sparse requires_tensor: True device_guard: False - - func: dense_dim(Tensor self) -> int use_c10_dispatcher: full variants: method dispatch: SparseCPU: dense_dim_sparse @@ -3492,50 +3443,47 @@ SparseCPU: dense_dim_sparse SparseCUDA: dense_dim_sparse requires_tensor: True device_guard: False - - func: _nnz(Tensor self) -> int use_c10_dispatcher: full variants: method dispatch: SparseCPU: _nnz_sparse SparseCUDA: _nnz_sparse requires_tensor: True device_guard: False - - func: coalesce(Tensor self) -> Tensor use_c10_dispatcher: full variants: method dispatch: SparseCPU: coalesce_sparse_cpu SparseCUDA: coalesce_sparse_cuda requires_tensor: True - - func: is_coalesced(Tensor self) -> bool use_c10_dispatcher: full variants: method dispatch: SparseCPU: is_coalesced_sparse SparseCUDA: is_coalesced_sparse requires_tensor: True device_guard: False - supports_named_tensor: True - - func: _indices(Tensor(a) self) -> Tensor(a) + use_c10_dispatcher: full variants: method dispatch: SparseCPU: _indices_sparse SparseCUDA: _indices_sparse requires_tensor: True device_guard: False - func: _values(Tensor(a) self) -> Tensor(a) + use_c10_dispatcher: full variants: method dispatch: SparseCPU: _values_sparse SparseCUDA: _values_sparse requires_tensor: True @@ -3551,26 +3499,27 @@ SparseCUDA: _coalesced_sparse_ requires_tensor: True device_guard: False - func: indices(Tensor(a) self) -> Tensor(a) + use_c10_dispatcher: full variants: method dispatch: SparseCPU: indices_sparse SparseCUDA: indices_sparse requires_tensor: True device_guard: False - func: values(Tensor(a) self) -> Tensor(a) + use_c10_dispatcher: full variants: method dispatch: SparseCPU: values_sparse SparseCUDA: values_sparse requires_tensor: True device_guard: False - - func: hspmm.out(Tensor mat1, Tensor mat2, *, Tensor(a!) out) -> Tensor(a!) dispatch: SparseCPU: hspmm_out_sparse_cpu SparseCUDA: hspmm_out_sparse_cuda requires_tensor: True @@ -3588,16 +3537,15 @@ SparseCPU: copy_sparse_ SparseCUDA: copy_sparse_ requires_tensor: True - func: unbind.int(Tensor(a) self, int dim=0) -> Tensor(a)[] + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: unbind.Dimname(Tensor(a) self, Dimname dim) -> Tensor(a)[] variants: function, method - supports_named_tensor: True - func: to_sparse.sparse_dim(Tensor self, int sparse_dim) -> Tensor use_c10_dispatcher: full variants: method dispatch: @@ -3616,10 +3564,11 @@ variants: method dispatch: CPU: dense_to_mkldnn - func: mkldnn_reorder_conv2d_weight(Tensor self, int[2] padding=0, int[2] stride=1, int[2] dilation=1, int groups=1) -> Tensor + use_c10_dispatcher: full variants: function python_module: nn dispatch: MkldnnCPU: mkldnn_reorder_conv2d_weight @@ -3627,70 +3576,92 @@ use_c10_dispatcher: full - func: quantize_per_tensor(Tensor self, float scale, int zero_point, ScalarType dtype) -> Tensor variants: function dispatch: - CPU: quantize_per_tensor_cpu + CPU: quantize_per_tensor + CUDA: quantize_per_tensor +- func: quantize_per_tensor.tensors(Tensor[] tensors, Tensor scales, Tensor zero_points, ScalarType dtype) -> Tensor[] + variants: function + dispatch: + CPU: quantize_per_tensor_list_cpu + - func: quantize_per_channel(Tensor self, Tensor scales, Tensor zero_points, int axis, ScalarType dtype) -> Tensor variants: function dispatch: CPU: quantize_per_channel_cpu -- func: dequantize(Tensor self) -> Tensor +- func: dequantize.self(Tensor self) -> Tensor use_c10_dispatcher: full variants: function, method dispatch: QuantizedCPU: dequantize_quant + QuantizedCUDA: dequantize_quant +- func: dequantize.tensors(Tensor[] tensors) -> Tensor[] + use_c10_dispatcher: full + variants: function + dispatch: + QuantizedCPU: dequantize_tensors_quant + - func: q_scale(Tensor self) -> float use_c10_dispatcher: full variants: function, method dispatch: QuantizedCPU: q_scale_quant + QuantizedCUDA: q_scale_quant - func: q_zero_point(Tensor self) -> int use_c10_dispatcher: full variants: function, method dispatch: QuantizedCPU: q_zero_point_quant + QuantizedCUDA: q_zero_point_quant - func: q_per_channel_scales(Tensor self) -> Tensor + use_c10_dispatcher: full variants: function, method dispatch: QuantizedCPU: q_per_channel_scales_quant - func: q_per_channel_zero_points(Tensor self) -> Tensor + use_c10_dispatcher: full variants: function, method dispatch: QuantizedCPU: q_per_channel_zero_points_quant - func: q_per_channel_axis(Tensor self) -> int + use_c10_dispatcher: full variants: function, method dispatch: QuantizedCPU: q_per_channel_axis_quant - func: int_repr(Tensor self) -> Tensor use_c10_dispatcher: full variants: function, method dispatch: - QuantizedCPU: int_repr_quant + QuantizedCPU: int_repr_quant_cpu + QuantizedCUDA: int_repr_quant_cuda - func: _make_per_tensor_quantized_tensor(Tensor self, float scale, int zero_point) -> Tensor use_c10_dispatcher: full dispatch: CPU: make_per_tensor_quantized_tensor_cpu + CUDA: make_per_tensor_quantized_tensor_cuda - func: _make_per_channel_quantized_tensor(Tensor self, Tensor scale, Tensor zero_point, int axis) -> Tensor + use_c10_dispatcher: full dispatch: CPU: make_per_channel_quantized_tensor_cpu - func: qscheme(Tensor self) -> QScheme use_c10_dispatcher: full variants: method dispatch: QuantizedCPU: qscheme_quant + QuantizedCUDA: qscheme_quant - func: fake_quantize_per_tensor_affine(Tensor self, float scale, int zero_point, int quant_min, int quant_max) -> Tensor use_c10_dispatcher: full variants: function @@ -3704,45 +3675,47 @@ - func: fake_quantize_per_channel_affine_backward(Tensor grad, Tensor self, Tensor scale, Tensor zero_point, int axis, int quant_min, int quant_max) -> Tensor use_c10_dispatcher: full variants: function +- func: _choose_qparams_per_tensor(Tensor self, bool reduce_range=False) -> (float, int) + use_c10_dispatcher: full + variants: function + # to(Device) must not exist because all constructors of Device also works for # TensorOptions. Otherwise, an ambiguity error is thrown. # See NOTE [ TensorOptions Constructors ]. - func: to.dtype_layout(Tensor self, *, ScalarType dtype, Layout layout, Device device, bool pin_memory=False, bool non_blocking=False, bool copy=False, MemoryFormat? memory_format=None) -> Tensor variants: method device_guard: False - supports_named_tensor: True - func: to.device(Tensor self, Device device, ScalarType dtype, bool non_blocking=False, bool copy=False, MemoryFormat? memory_format=None) -> Tensor variants: method device_guard: False - supports_named_tensor: True - func: to.dtype(Tensor self, ScalarType dtype, bool non_blocking=False, bool copy=False, MemoryFormat? memory_format=None) -> Tensor variants: method device_guard: False - supports_named_tensor: True - func: to.other(Tensor self, Tensor other, bool non_blocking=False, bool copy=False, MemoryFormat? memory_format=None) -> Tensor variants: method device_guard: False - func: meshgrid(Tensor[] tensors) -> Tensor[] + use_c10_dispatcher: full - func: cartesian_prod(Tensor[] tensors) -> Tensor + use_c10_dispatcher: full variants: function - func: combinations(Tensor self, int r=2, bool with_replacement=False) -> Tensor use_c10_dispatcher: full variants: function - func: item(Tensor self) -> Scalar use_c10_dispatcher: full variants: method - supports_named_tensor: True - func: result_type.Tensor(Tensor tensor, Tensor other) -> ScalarType variants: function - func: result_type.Scalar(Tensor tensor, Scalar other) -> ScalarType @@ -3764,11 +3737,10 @@ use_c10_dispatcher: full dispatch: CPU: _local_scalar_dense_cpu CUDA: _local_scalar_dense_cuda variants: function - supports_named_tensor: True # Fused RNN kernels - func: _thnn_fused_lstm_cell(Tensor input_gates, Tensor hidden_gates, Tensor cx, Tensor? input_bias=None, Tensor? hidden_bias=None) -> (Tensor, Tensor, Tensor) dispatch: CUDA: _thnn_fused_lstm_cell_cuda @@ -3782,53 +3754,67 @@ - func: _thnn_fused_gru_cell(Tensor input_gates, Tensor hidden_gates, Tensor hx, Tensor? input_bias=None, Tensor? hidden_bias=None) -> (Tensor, Tensor) dispatch: CUDA: _thnn_fused_gru_cell_cuda - func: _thnn_fused_gru_cell_backward(Tensor grad_hy, Tensor workspace, bool has_bias) -> (Tensor, Tensor, Tensor, Tensor, Tensor) + use_c10_dispatcher: full dispatch: CUDA: _thnn_fused_gru_cell_backward_cuda - func: _thnn_differentiable_gru_cell_backward(Tensor grad_hy, Tensor input_gates, Tensor hidden_gates, Tensor hx, Tensor? input_bias, Tensor? hidden_bias) -> (Tensor, Tensor, Tensor, Tensor, Tensor) # RNN cells and layers - func: lstm.input(Tensor input, Tensor[] hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, bool batch_first) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full - func: lstm.data(Tensor data, Tensor batch_sizes, Tensor[] hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full - func: gru.input(Tensor input, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, bool batch_first) -> (Tensor, Tensor) + use_c10_dispatcher: full - func: gru.data(Tensor data, Tensor batch_sizes, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional) -> (Tensor, Tensor) + use_c10_dispatcher: full - func: rnn_tanh.input(Tensor input, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, bool batch_first) -> (Tensor, Tensor) + use_c10_dispatcher: full - func: rnn_tanh.data(Tensor data, Tensor batch_sizes, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional) -> (Tensor, Tensor) + use_c10_dispatcher: full - func: rnn_relu.input(Tensor input, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, bool batch_first) -> (Tensor, Tensor) + use_c10_dispatcher: full - func: rnn_relu.data(Tensor data, Tensor batch_sizes, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional) -> (Tensor, Tensor) + use_c10_dispatcher: full - func: lstm_cell(Tensor input, Tensor[] hx, Tensor w_ih, Tensor w_hh, Tensor? b_ih=None, Tensor? b_hh=None) -> (Tensor, Tensor) - func: gru_cell(Tensor input, Tensor hx, Tensor w_ih, Tensor w_hh, Tensor? b_ih=None, Tensor? b_hh=None) -> Tensor - func: rnn_tanh_cell(Tensor input, Tensor hx, Tensor w_ih, Tensor w_hh, Tensor? b_ih=None, Tensor? b_hh=None) -> Tensor - func: rnn_relu_cell(Tensor input, Tensor hx, Tensor w_ih, Tensor w_hh, Tensor? b_ih=None, Tensor? b_hh=None) -> Tensor +# Quantized RNN layer registration has been moved to C10 dispatch in `RNN.cpp` + # Quantized RNN layers -- func: quantized_lstm(Tensor input, Tensor[] hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, bool batch_first, *, ScalarType? dtype=None, bool use_dynamic=False) -> (Tensor, Tensor, Tensor) +# - func: quantized_lstm(Tensor input, Tensor[] hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, bool batch_first, *, ScalarType? dtype=None, bool use_dynamic=False) -> (Tensor, Tensor, Tensor) -- func: quantized_lstm.data(Tensor data, Tensor batch_sizes, Tensor[] hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, *, ScalarType? dtype=None, bool use_dynamic=False) -> (Tensor, Tensor, Tensor) +# - func: quantized_lstm.data(Tensor data, Tensor batch_sizes, Tensor[] hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, *, ScalarType? dtype=None, bool use_dynamic=False) -> (Tensor, Tensor, Tensor) # Quantized GRU layers -- func: quantized_gru.input(Tensor input, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, bool batch_first) -> (Tensor, Tensor) +# - func: quantized_gru.input(Tensor input, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, bool batch_first) -> (Tensor, Tensor) +# use_c10_dispatcher: full -- func: quantized_gru.data(Tensor data, Tensor batch_sizes, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional) -> (Tensor, Tensor) +# - func: quantized_gru.data(Tensor data, Tensor batch_sizes, Tensor hx, Tensor[] params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional) -> (Tensor, Tensor) +# use_c10_dispatcher: full # Quantized RNN cells - func: quantized_lstm_cell(Tensor input, Tensor[] hx, Tensor w_ih, Tensor w_hh, Tensor b_ih, Tensor b_hh, Tensor packed_ih, Tensor packed_hh, Tensor col_offsets_ih, Tensor col_offsets_hh, Scalar scale_ih, Scalar scale_hh, Scalar zero_point_ih, Scalar zero_point_hh) -> (Tensor, Tensor) + use_c10_dispatcher: full - func: quantized_gru_cell(Tensor input, Tensor hx, Tensor w_ih, Tensor w_hh, Tensor b_ih, Tensor b_hh, Tensor packed_ih, Tensor packed_hh, Tensor col_offsets_ih, Tensor col_offsets_hh, Scalar scale_ih, Scalar scale_hh, Scalar zero_point_ih, Scalar zero_point_hh) -> Tensor use_c10_dispatcher: full - func: quantized_rnn_relu_cell(Tensor input, Tensor hx, Tensor w_ih, Tensor w_hh, Tensor b_ih, Tensor b_hh, Tensor packed_ih, Tensor packed_hh, Tensor col_offsets_ih, Tensor col_offsets_hh, Scalar scale_ih, Scalar scale_hh, Scalar zero_point_ih, Scalar zero_point_hh) -> Tensor @@ -3837,14 +3823,17 @@ - func: quantized_rnn_tanh_cell(Tensor input, Tensor hx, Tensor w_ih, Tensor w_hh, Tensor b_ih, Tensor b_hh, Tensor packed_ih, Tensor packed_hh, Tensor col_offsets_ih, Tensor col_offsets_hh, Scalar scale_ih, Scalar scale_hh, Scalar zero_point_ih, Scalar zero_point_hh) -> Tensor use_c10_dispatcher: full # PackedSequence utilities - func: _pack_padded_sequence(Tensor input, Tensor lengths, bool batch_first) -> (Tensor, Tensor) + use_c10_dispatcher: full - func: _pack_padded_sequence_backward(Tensor grad, int[] input_size, Tensor batch_sizes, bool batch_first) -> Tensor + use_c10_dispatcher: full - func: _pad_packed_sequence(Tensor data, Tensor batch_sizes, bool batch_first, Scalar padding_value, int total_length) -> (Tensor, Tensor) + use_c10_dispatcher: full # wrappers for legacy TH methods - func: set_.source_Storage(Tensor(a!) self, Storage source) -> Tensor(a!) variants: method @@ -3855,13 +3844,14 @@ - func: set_.source_Storage_storage_offset(Tensor(a!) self, Storage source, int storage_offset, int[] size, int[] stride=[]) -> Tensor(a!) variants: method device_guard: False dispatch: - CPU: legacy::cpu::_th_set_ - CUDA: legacy::cuda::_th_set_ - QuantizedCPU: set_storage + CPU: set_storage_cpu_ + CUDA: set_storage_cuda_ + QuantizedCPU: set_storage_quantized_ + QuantizedCUDA: set_storage_quantized_ - func: set_.source_Tensor(Tensor(a!) self, Tensor source) -> Tensor(a!) variants: method device_guard: False dispatch: @@ -3876,10 +3866,11 @@ - func: set_quantizer_(Tensor(a!) self, ConstQuantizerPtr quantizer) -> Tensor(a!) variants: method dispatch: QuantizedCPU: set_quantizer_ + QuantizedCUDA: set_quantizer_ - func: is_set_to(Tensor self, Tensor tensor) -> bool use_c10_dispatcher: full variants: method device_guard: False @@ -3890,28 +3881,24 @@ - func: masked_fill_.Scalar(Tensor(a!) self, Tensor mask, Scalar value) -> Tensor(a!) variants: method dispatch: CPU: masked_fill__cpu CUDA: masked_fill__cuda - supports_named_tensor: True - func: masked_fill.Scalar(Tensor self, Tensor mask, Scalar value) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: masked_fill_.Tensor(Tensor(a!) self, Tensor mask, Tensor value) -> Tensor(a!) variants: method dispatch: CPU: masked_fill__cpu CUDA: masked_fill__cuda - supports_named_tensor: True - func: masked_fill.Tensor(Tensor self, Tensor mask, Tensor value) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: masked_scatter_(Tensor(a!) self, Tensor mask, Tensor source) -> Tensor(a!) variants: method dispatch: CPU: masked_scatter__cpu @@ -3920,17 +3907,19 @@ - func: masked_scatter(Tensor self, Tensor mask, Tensor source) -> Tensor use_c10_dispatcher: full variants: function, method - func: view(Tensor(a) self, int[] size) -> Tensor(a) + use_c10_dispatcher: full variants: method device_guard: False dispatch: CPU: view CUDA: view MkldnnCPU: mkldnn_view QuantizedCPU: view + QuantizedCUDA: view - func: put_(Tensor(a!) self, Tensor index, Tensor source, bool accumulate=False) -> Tensor(a!) variants: method dispatch: CPU: legacy::cpu::_th_put_ @@ -3949,63 +3938,55 @@ - func: index_add.dimname(Tensor self, Dimname dim, Tensor index, Tensor source) -> Tensor variants: function, method - func: index_fill_.int_Scalar(Tensor(a!) self, int dim, Tensor index, Scalar value) -> Tensor(a!) variants: method - supports_named_tensor: True dispatch: CPU: legacy::cpu::_th_index_fill_ CUDA: legacy::cuda::_th_index_fill_ - func: index_fill.int_Scalar(Tensor self, int dim, Tensor index, Scalar value) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: function, method - func: index_fill_.int_Tensor(Tensor(a!) self, int dim, Tensor index, Tensor value) -> Tensor(a!) variants: method dispatch: CPU: index_fill_ CUDA: index_fill_ - supports_named_tensor: True - func: index_fill.int_Tensor(Tensor self, int dim, Tensor index, Tensor value) -> Tensor use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: index_fill_.Dimname_Scalar(Tensor(a!) self, Dimname dim, Tensor index, Scalar value) -> Tensor(a!) variants: method - supports_named_tensor: True - func: index_fill_.Dimname_Tensor(Tensor(a!) self, Dimname dim, Tensor index, Tensor value) -> Tensor(a!) variants: method - supports_named_tensor: True - func: index_fill.Dimname_Scalar(Tensor self, Dimname dim, Tensor index, Scalar value) -> Tensor variants: function, method - supports_named_tensor: True - func: index_fill.Dimname_Tensor(Tensor self, Dimname dim, Tensor index, Tensor value) -> Tensor variants: function, method - supports_named_tensor: True - func: scatter_.src(Tensor(a!) self, int dim, Tensor index, Tensor src) -> Tensor(a!) variants: method dispatch: - CPU: scatter_cpu_ - CUDA: legacy::cuda::_th_scatter_ + CPU: scatter_ + CUDA: scatter_ - func: scatter.src(Tensor self, int dim, Tensor index, Tensor src) -> Tensor use_c10_dispatcher: full variants: function, method - func: scatter_.value(Tensor(a!) self, int dim, Tensor index, Scalar value) -> Tensor(a!) variants: method dispatch: - CPU: scatter_fill_cpu_ - CUDA: legacy::cuda::_th_scatter_ + CPU: scatter_fill_ + CUDA: scatter_fill_ - func: scatter.value(Tensor self, int dim, Tensor index, Scalar value) -> Tensor use_c10_dispatcher: full variants: function, method @@ -4016,12 +3997,12 @@ variants: function, method - func: scatter_add_(Tensor(a!) self, int dim, Tensor index, Tensor src) -> Tensor(a!) variants: method dispatch: - CPU: scatter_add_cpu_ - CUDA: legacy::cuda::_th_scatter_add_ + CPU: scatter_add_ + CUDA: scatter_add_ - func: scatter_add(Tensor self, int dim, Tensor index, Tensor src) -> Tensor use_c10_dispatcher: full variants: function, method @@ -4075,13 +4056,15 @@ dispatch: CPU: bitwise_and_out CUDA: bitwise_and_out - func: bitwise_and.Scalar(Tensor self, Scalar other) -> Tensor + use_c10_dispatcher: full variants: method, function - func: bitwise_and.Tensor(Tensor self, Tensor other) -> Tensor + use_c10_dispatcher: full variants: method, function - func: bitwise_and_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!) variants: method @@ -4113,13 +4096,15 @@ dispatch: CPU: bitwise_or_out CUDA: bitwise_or_out - func: bitwise_or.Scalar(Tensor self, Scalar other) -> Tensor + use_c10_dispatcher: full variants: method, function - func: bitwise_or.Tensor(Tensor self, Tensor other) -> Tensor + use_c10_dispatcher: full variants: method, function - func: bitwise_or_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!) variants: method @@ -4151,13 +4136,15 @@ dispatch: CPU: bitwise_xor_out CUDA: bitwise_xor_out - func: bitwise_xor.Scalar(Tensor self, Scalar other) -> Tensor + use_c10_dispatcher: full variants: method, function - func: bitwise_xor.Tensor(Tensor self, Tensor other) -> Tensor + use_c10_dispatcher: full variants: method, function - func: bitwise_xor_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!) variants: method @@ -4229,18 +4216,16 @@ dispatch: CPU: __irshift__ CUDA: __irshift__ - func: lgamma_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: method dispatch: CPU: _lgamma__cpu CUDA: _lgamma__cuda - func: atan2_(Tensor(a!) self, Tensor other) -> Tensor(a!) - supports_named_tensor: True variants: method - func: tril_(Tensor(a!) self, int diagonal=0) -> Tensor(a!) variants: method dispatch: @@ -4252,32 +4237,28 @@ dispatch: CPU: triu_cpu_ CUDA: triu_cuda_ - func: digamma_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: method - func: polygamma_(Tensor(a!) self, int n) -> Tensor(a!) - supports_named_tensor: True variants: method - func: renorm_(Tensor(a!) self, Scalar p, int dim, Scalar maxnorm) -> Tensor(a!) variants: method dispatch: CPU: legacy::cpu::_th_renorm_ CUDA: legacy::cuda::_th_renorm_ - func: pow_.Scalar(Tensor(a!) self, Scalar exponent) -> Tensor(a!) - supports_named_tensor: True variants: method dispatch: CPU: pow_ CUDA: pow_ - func: pow_.Tensor(Tensor(a!) self, Tensor exponent) -> Tensor(a!) - supports_named_tensor: True variants: method dispatch: CPU: pow_ CUDA: pow_ @@ -4295,17 +4276,17 @@ - func: fmod_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!) variants: method dispatch: CPU: fmod_ - CUDA: legacy::cuda::_th_fmod_ + CUDA: fmod_cuda_ - func: fmod_.Tensor(Tensor(a!) self, Tensor other) -> Tensor(a!) variants: method dispatch: CPU: fmod_ - CUDA: legacy::cuda::_th_fmod_ + CUDA: fmod_cuda_ - func: remainder_.Scalar(Tensor(a!) self, Scalar other) -> Tensor(a!) variants: method dispatch: CPU: remainder_ @@ -4319,76 +4300,61 @@ - func: addbmm_(Tensor(a!) self, Tensor batch1, Tensor batch2, *, Scalar beta=1, Scalar alpha=1) -> Tensor(a!) variants: method dispatch: CPU: legacy::cpu::_th_addbmm_ - CUDA: legacy::cuda::_th_addbmm_ + CUDA: addbmm__cuda - func: addbmm.out(Tensor self, Tensor batch1, Tensor batch2, *, Scalar beta=1, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!) dispatch: - CPU: legacy::cpu::_th_addbmm_out - CUDA: legacy::cuda::_th_addbmm_out + CPU: addbmm_cpu_out + CUDA: addbmm_out_cuda - func: addbmm(Tensor self, Tensor batch1, Tensor batch2, *, Scalar beta=1, Scalar alpha=1) -> Tensor use_c10_dispatcher: full variants: method, function dispatch: - CPU: legacy::cpu::_th_addbmm - CUDA: legacy::cuda::_th_addbmm + CPU: addbmm_cpu + CUDA: addbmm_cuda - func: addcdiv_(Tensor(a!) self, Tensor tensor1, Tensor tensor2, *, Scalar value=1) -> Tensor(a!) variants: method - supports_named_tensor: True - func: random_.from(Tensor(a!) self, int from, int? to, *, Generator? generator=None) -> Tensor(a!) variants: method - supports_named_tensor: True - func: random_.to(Tensor(a!) self, int to, *, Generator? generator=None) -> Tensor(a!) variants: method - supports_named_tensor: True - func: random_(Tensor(a!) self, *, Generator? generator=None) -> Tensor(a!) variants: method - supports_named_tensor: True - func: uniform_(Tensor(a!) self, float from=0, float to=1, *, Generator? generator=None) -> Tensor(a!) variants: method - dispatch: - CPU: legacy::cpu::_th_uniform_ - CUDA: uniform_cuda_ - supports_named_tensor: True - func: cauchy_(Tensor(a!) self, float median=0, float sigma=1, *, Generator? generator=None) -> Tensor(a!) variants: method - supports_named_tensor: True - func: log_normal_(Tensor(a!) self, float mean=1, float std=2, *, Generator? generator=None) -> Tensor(a!) variants: method - supports_named_tensor: True - func: exponential_(Tensor(a!) self, float lambd=1, *, Generator? generator=None) -> Tensor(a!) variants: method - supports_named_tensor: True - func: geometric_(Tensor(a!) self, float p, *, Generator? generator=None) -> Tensor(a!) variants: method - supports_named_tensor: True # wrappers for TH functions - func: diag.out(Tensor self, int diagonal=0, *, Tensor(a!) out) -> Tensor(a!) dispatch: - CPU: legacy::cpu::_th_diag_out - CUDA: legacy::cuda::_th_diag_out + CPU: diag_cpu_out + CUDA: diag_cuda_out - func: diag(Tensor self, int diagonal=0) -> Tensor use_c10_dispatcher: full variants: method, function - dispatch: - CPU: legacy::cpu::_th_diag - CUDA: legacy::cuda::_th_diag - func: cross.out(Tensor self, Tensor other, int? dim=None, *, Tensor(a!) out) -> Tensor(a!) - func: cross(Tensor self, Tensor other, int? dim=None) -> Tensor use_c10_dispatcher: full @@ -4425,197 +4391,173 @@ - func: trace(Tensor self) -> Tensor use_c10_dispatcher: full variants: method, function dispatch: CPU: legacy::cpu::_th_trace - CUDA: legacy::cuda::_th_trace + CUDA: trace_cuda - func: ne.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: ne_out CUDA: ne_out QuantizedCPU: ne_out_quantized_cpu - func: ne.Scalar(Tensor self, Scalar other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: ne CUDA: ne QuantizedCPU: ne_quantized_cpu - func: ne.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: ne_out CUDA: ne_out QuantizedCPU: ne_out_quantized_cpu - func: ne.Tensor(Tensor self, Tensor other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: ne CUDA: ne QuantizedCPU: ne_quantized_cpu - func: eq.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: eq_out CUDA: eq_out QuantizedCPU: eq_out_quantized_cpu - func: eq.Scalar(Tensor self, Scalar other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: eq CUDA: eq QuantizedCPU: eq_quantized_cpu - func: eq.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: eq_out CUDA: eq_out QuantizedCPU: eq_out_quantized_cpu - func: eq.Tensor(Tensor self, Tensor other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: eq CUDA: eq QuantizedCPU: eq_quantized_cpu - func: ge.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: ge_out CUDA: ge_out QuantizedCPU: ge_out_quantized_cpu - func: ge.Scalar(Tensor self, Scalar other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: ge CUDA: ge QuantizedCPU: ge_quantized_cpu - func: ge.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: ge_out CUDA: ge_out QuantizedCPU: ge_out_quantized_cpu - func: ge.Tensor(Tensor self, Tensor other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: ge CUDA: ge QuantizedCPU: ge_quantized_cpu - func: le.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: le_out CUDA: le_out QuantizedCPU: le_out_quantized_cpu - func: le.Scalar(Tensor self, Scalar other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: le CUDA: le QuantizedCPU: le_quantized_cpu - func: le.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: le_out CUDA: le_out QuantizedCPU: le_out_quantized_cpu - func: le.Tensor(Tensor self, Tensor other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: le CUDA: le QuantizedCPU: le_quantized_cpu - func: gt.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: gt_out CUDA: gt_out QuantizedCPU: gt_out_quantized_cpu - func: gt.Scalar(Tensor self, Scalar other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: gt CUDA: gt QuantizedCPU: gt_quantized_cpu - func: gt.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: gt_out CUDA: gt_out QuantizedCPU: gt_out_quantized_cpu - func: gt.Tensor(Tensor self, Tensor other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: gt CUDA: gt QuantizedCPU: gt_quantized_cpu - func: lt.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: lt_out CUDA: lt_out QuantizedCPU: lt_out_quantized_cpu - func: lt.Scalar(Tensor self, Scalar other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: lt CUDA: lt QuantizedCPU: lt_quantized_cpu - func: lt.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: lt_out CUDA: lt_out QuantizedCPU: lt_out_quantized_cpu - func: lt.Tensor(Tensor self, Tensor other) -> Tensor - supports_named_tensor: True use_c10_dispatcher: full variants: method, function dispatch: CPU: lt CUDA: lt @@ -4654,19 +4596,17 @@ - func: masked_select.out(Tensor self, Tensor mask, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: masked_select_out_cpu CUDA: masked_select_out_cuda - supports_named_tensor: True - func: masked_select(Tensor self, Tensor mask) -> Tensor use_c10_dispatcher: full variants: method, function dispatch: CPU: masked_select_cpu CUDA: masked_select_cuda - supports_named_tensor: True - func: nonzero.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: legacy::cpu::_th_nonzero_out CUDA: legacy::cuda::_th_nonzero_out @@ -4677,80 +4617,81 @@ dispatch: CPU: legacy::cpu::_th_nonzero CUDA: legacy::cuda::_th_nonzero - func: nonzero_numpy(Tensor self) -> Tensor[] + use_c10_dispatcher: full variants: method, function - func: gather.out(Tensor self, int dim, Tensor index, *, bool sparse_grad=False, Tensor(a!) out) -> Tensor(a!) dispatch: - CPU: gather_out_cpu - CUDA: gather_out_cuda + CPU: gather_out_cpu_cuda + CUDA: gather_out_cpu_cuda - func: gather(Tensor self, int dim, Tensor index, *, bool sparse_grad=False) -> Tensor use_c10_dispatcher: full variants: method, function dispatch: - CPU: gather_cpu - CUDA: gather_cuda + CPU: gather + CUDA: gather - func: gather.dimname_out(Tensor self, Dimname dim, Tensor index, *, bool sparse_grad=False, Tensor(a!) out) -> Tensor(a!) - func: gather.dimname(Tensor self, Dimname dim, Tensor index, *, bool sparse_grad=False) -> Tensor variants: method, function - func: _gather_sparse_backward(Tensor self, int dim, Tensor index, Tensor grad) -> Tensor use_c10_dispatcher: full - func: addcmul.out(Tensor self, Tensor tensor1, Tensor tensor2, *, Scalar value=1, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: addcmul(Tensor self, Tensor tensor1, Tensor tensor2, *, Scalar value=1) -> Tensor use_c10_dispatcher: full variants: method, function - supports_named_tensor: True - func: addcmul_(Tensor(a!) self, Tensor tensor1, Tensor tensor2, *, Scalar value=1) -> Tensor(a!) variants: method - supports_named_tensor: True - func: addcdiv.out(Tensor self, Tensor tensor1, Tensor tensor2, *, Scalar value=1, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: addcdiv(Tensor self, Tensor tensor1, Tensor tensor2, *, Scalar value=1) -> Tensor use_c10_dispatcher: full variants: method, function - supports_named_tensor: True - func: lstsq.X(Tensor self, Tensor A, *, Tensor(a!) X, Tensor(b!) qr) -> (Tensor(a!) solution, Tensor(b!) QR) dispatch: CPU: legacy::cpu::_th_gels_out CUDA: legacy::cuda::_th_gels_out - func: lstsq(Tensor self, Tensor A) -> (Tensor solution, Tensor QR) + use_c10_dispatcher: full variants: method, function dispatch: CPU: legacy::cpu::_th_gels CUDA: legacy::cuda::_th_gels - func: triangular_solve.X(Tensor self, Tensor A, bool upper=True, bool transpose=False, bool unitriangular=False, *, Tensor(a!) X, Tensor(b!) M) -> (Tensor(a!) solution, Tensor(b!) cloned_coefficient) - func: triangular_solve(Tensor self, Tensor A, bool upper=True, bool transpose=False, bool unitriangular=False) -> (Tensor solution, Tensor cloned_coefficient) + use_c10_dispatcher: full variants: method, function - func: _triangular_solve_helper(Tensor self, Tensor A, bool upper, bool transpose, bool unitriangular) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: _triangular_solve_helper_cpu CUDA: _triangular_solve_helper_cuda - func: symeig.e(Tensor self, bool eigenvectors=False, bool upper=True, *, Tensor(a!) e, Tensor(b!) V) -> (Tensor(a!) eigenvalues, Tensor(b!) eigenvectors) - func: symeig(Tensor self, bool eigenvectors=False, bool upper=True) -> (Tensor eigenvalues, Tensor eigenvectors) + use_c10_dispatcher: full variants: method, function - func: _symeig_helper(Tensor self, bool eigenvectors, bool upper) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: _symeig_helper_cpu CUDA: _symeig_helper_cuda @@ -4758,21 +4699,24 @@ dispatch: CPU: legacy::cpu::_th_eig_out CUDA: legacy::cuda::_th_eig_out - func: eig(Tensor self, bool eigenvectors=False) -> (Tensor eigenvalues, Tensor eigenvectors) + use_c10_dispatcher: full variants: method, function dispatch: CPU: legacy::cpu::_th_eig CUDA: legacy::cuda::_th_eig - func: svd.U(Tensor self, bool some=True, bool compute_uv=True, *, Tensor(a!) U, Tensor(b!) S, Tensor(c!) V) -> (Tensor(a!) U, Tensor(b!) S, Tensor(c!) V) - func: svd(Tensor self, bool some=True, bool compute_uv=True) -> (Tensor U, Tensor S, Tensor V) + use_c10_dispatcher: full variants: method, function - func: _svd_helper(Tensor self, bool some, bool compute_uv) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: _svd_helper_cpu CUDA: _svd_helper_cuda @@ -4801,15 +4745,17 @@ dispatch: CPU: _cholesky_solve_helper_cpu CUDA: _cholesky_solve_helper_cuda - func: solve(Tensor self, Tensor A) -> (Tensor solution, Tensor LU) + use_c10_dispatcher: full variants: function, method - func: solve.solution(Tensor self, Tensor A, *, Tensor(a!) solution, Tensor(b!) lu) -> (Tensor(a!) solution, Tensor(b!) LU) - func: _solve_helper(Tensor self, Tensor A) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: _solve_helper_cpu CUDA: _solve_helper_cuda @@ -4826,13 +4772,15 @@ CUDA: legacy::cuda::_th_potri - func: qr.Q(Tensor self, bool some=True, *, Tensor(a!) Q, Tensor(b!) R) -> (Tensor(a!) Q, Tensor(b!) R) - func: qr(Tensor self, bool some=True) -> (Tensor Q, Tensor R) + use_c10_dispatcher: full variants: method, function - func: _qr_helper(Tensor self, bool some) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: _qr_helper_cpu CUDA: _qr_helper_cuda @@ -4840,10 +4788,11 @@ dispatch: CPU: legacy::cpu::_th_geqrf_out CUDA: legacy::cuda::_th_geqrf_out - func: geqrf(Tensor self) -> (Tensor a, Tensor tau) + use_c10_dispatcher: full variants: method, function dispatch: CPU: legacy::cpu::_th_geqrf CUDA: legacy::cuda::_th_geqrf @@ -4866,10 +4815,11 @@ variants: method, function dispatch: CPU: legacy::cpu::_th_ormqr - func: _lu_with_info(Tensor self, bool pivot=True, bool check_errors=True) -> (Tensor, Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: _lu_with_info_cpu CUDA: _lu_with_info_cuda @@ -4897,10 +4847,11 @@ dispatch: CPU: multinomial CUDA: multinomial - func: _multinomial_alias_setup(Tensor probs) -> (Tensor, Tensor) + use_c10_dispatcher: full variants: function dispatch: CPU: legacy::cpu::_th_multinomial_alias_setup CUDA: legacy::cuda::_th_multinomial_alias_setup @@ -4909,84 +4860,71 @@ dispatch: CPU: legacy::cpu::_th_multinomial_alias_draw CUDA: legacy::cuda::_th_multinomial_alias_draw - func: lgamma.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: _lgamma_out_cpu CUDA: _lgamma_out_cuda - func: lgamma(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: method, function dispatch: CPU: lgamma CUDA: lgamma - func: digamma.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: digamma(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: method, function - func: polygamma.out(int n, Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: polygamma(int n, Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: method, function - func: erfinv(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: method, function dispatch: CPU: erfinv CUDA: erfinv - func: erfinv_(Tensor(a!) self) -> Tensor(a!) - supports_named_tensor: True variants: method dispatch: CPU: _erfinv__cpu CUDA: _erfinv__cuda - func: erfinv.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: _erfinv_out_cpu CUDA: _erfinv_out_cuda - func: sign(Tensor self) -> Tensor + use_c10_dispatcher: full variants: function, method - supports_named_tensor: True - func: sign_(Tensor(a!) self) -> Tensor(a!) variants: method - supports_named_tensor: True - func: sign.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: sign_out CUDA: sign_out - func: dist(Tensor self, Tensor other, Scalar p=2) -> Tensor use_c10_dispatcher: full variants: method, function - func: atan2.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True - func: atan2(Tensor self, Tensor other) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: method, function - func: lerp.Scalar_out(Tensor self, Tensor end, Scalar weight, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: lerp_cpu_scalar_out @@ -5024,30 +4962,30 @@ CUDA: _histc_cuda - func: fmod.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: fmod_out - CUDA: legacy::cuda::_th_fmod_out + CUDA: fmod_cuda_out - func: fmod.Scalar(Tensor self, Scalar other) -> Tensor use_c10_dispatcher: full variants: method, function dispatch: CPU: fmod - CUDA: legacy::cuda::_th_fmod + CUDA: fmod_cuda - func: fmod.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: fmod_out - CUDA: legacy::cuda::_th_fmod_out + CUDA: fmod_cuda_out - func: fmod.Tensor(Tensor self, Tensor other) -> Tensor use_c10_dispatcher: full variants: method, function dispatch: CPU: fmod - CUDA: legacy::cuda::_th_fmod + CUDA: fmod_cuda - func: remainder.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: remainder_out CUDA: remainder_out @@ -5080,13 +5018,12 @@ - func: min(Tensor self) -> Tensor use_c10_dispatcher: full variants: method, function dispatch: CPU: min - CUDA: legacy::cuda::_th_min + CUDA: min QuantizedCPU: min_quant - supports_named_tensor: True - func: max.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!) - func: max.other(Tensor self, Tensor other) -> Tensor use_c10_dispatcher: full @@ -5095,28 +5032,27 @@ - func: max(Tensor self) -> Tensor use_c10_dispatcher: full variants: method, function dispatch: CPU: max - CUDA: legacy::cuda::_th_max + CUDA: max QuantizedCPU: max_quant - supports_named_tensor: True - func: median(Tensor self) -> Tensor use_c10_dispatcher: full variants: method, function dispatch: CPU: median_cpu CUDA: median_cuda - supports_named_tensor: True - func: sort.values(Tensor self, int dim=-1, bool descending=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!) values, Tensor(b!) indices) dispatch: CPU: legacy::cpu::_th_sort_out CUDA: legacy::cuda::_th_sort_out - func: sort(Tensor self, int dim=-1, bool descending=False) -> (Tensor values, Tensor indices) + use_c10_dispatcher: full variants: method, function dispatch: CPU: legacy::cpu::_th_sort CUDA: legacy::cuda::_th_sort QuantizedCPU: sort_quant @@ -5137,24 +5073,23 @@ dispatch: CPU: topk_out_cpu CUDA: legacy::cuda::_th_topk_out - func: topk(Tensor self, int k, int dim=-1, bool largest=True, bool sorted=True) -> (Tensor values, Tensor indices) + use_c10_dispatcher: full variants: method, function dispatch: CPU: topk CUDA: topk QuantizedCPU: quantized_topk_cpu - func: all(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: method, function - func: any(Tensor self) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: method, function dispatch: CPU: any CUDA: any SparseCPU: any_sparse @@ -5171,154 +5106,122 @@ dispatch: CPU: legacy::cpu::_th_renorm CUDA: legacy::cuda::_th_renorm - func: unfold(Tensor(a) self, int dimension, int size, int step) -> Tensor(a) + use_c10_dispatcher: full variants: method device_guard: False dispatch: CPU: unfold CUDA: unfold + QuantizedCPU: unfold + QuantizedCUDA: unfold +- func: unfold_backward(Tensor grad_in, int[] input_sizes, int dim, int size, int step) -> Tensor + variants: function + dispatch: + CPU: unfold_backward + CUDA: unfold_backward + - func: equal(Tensor self, Tensor other) -> bool use_c10_dispatcher: full variants: method, function dispatch: CPU: legacy::cpu::_th_equal CUDA: legacy::cuda::_th_equal - QuantizedCPU: quantized_equal - supports_named_tensor: True + QuantizedCPU: quantized_equal_cpu - func: pow.Tensor_Tensor_out(Tensor self, Tensor exponent, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: pow_out CUDA: pow_out - func: pow.Tensor_Tensor(Tensor self, Tensor exponent) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True variants: method, function dispatch: CPU: pow CUDA: pow - func: pow.Scalar_out(Scalar self, Tensor exponent, *, Tensor(a!) out) -> Tensor(a!) - supports_named_tensor: True dispatch: CPU: pow_out CUDA: pow_out - func: pow.Scalar(Scalar self, Tensor exponent) -> Tensor use_c10_dispatcher: full - supports_named_tensor: True dispatch: CPU: pow CUDA: pow - func: normal_(Tensor(a!) self, float mean=0, float std=1, *, Generator? generator=None) -> Tensor(a!) variants: method - dispatch: - CPU: normal_cpu_ - CUDA: normal_cuda_ - supports_named_tensor: True - func: normal.Tensor_float_out(Tensor mean, float std=1, *, Generator? generator=None, Tensor(a!) out) -> Tensor(a!) - dispatch: - CPU: normal_out_cpu - CUDA: normal_out_cuda - func: normal.Tensor_float(Tensor mean, float std=1, *, Generator? generator=None) -> Tensor - dispatch: - CPU: normal_cpu - CUDA: normal_cuda - func: normal.float_Tensor_out(float mean, Tensor std, *, Generator? generator=None, Tensor(a!) out) -> Tensor(a!) - dispatch: - CPU: normal_out_cpu - CUDA: normal_out_cuda - func: normal.float_Tensor(float mean, Tensor std, *, Generator? generator=None) -> Tensor - dispatch: - CPU: normal_cpu - CUDA: normal_cuda - func: normal.Tensor_Tensor_out(Tensor mean, Tensor std, *, Generator? generator=None, Tensor(a!) out) -> Tensor(a!) - dispatch: - CPU: normal_out_cpu - CUDA: normal_out_cuda - func: normal.Tensor_Tensor(Tensor mean, Tensor std, *, Generator? generator=None) -> Tensor - dispatch: - CPU: normal_cpu - CUDA: normal_cuda - func: normal.float_float(float mean, float std, int[] size, *, Generator? generator=None, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor - func: normal.float_float_out(float mean, float std, int[] size, *, Generator? generator=None, Tensor(a!) out) -> Tensor(a!) - func: alias(Tensor(a) self) -> Tensor(a) + use_c10_dispatcher: full variants: method, function - supports_named_tensor: True - func: _addr(Tensor self, Tensor vec1, Tensor vec2, *, Scalar beta=1, Scalar alpha=1) -> Tensor use_c10_dispatcher: full dispatch: CPU: legacy::cpu::_th_addr - CUDA: legacy::cuda::_th_addr + CUDA: addr_cuda - func: _addr_(Tensor(a!) self, Tensor vec1, Tensor vec2, *, Scalar beta=1, Scalar alpha=1) -> Tensor(a!) dispatch: CPU: legacy::cpu::_th_addr_ - CUDA: legacy::cuda::_th_addr_ + CUDA: addr__cuda - func: _addr.out(Tensor self, Tensor vec1, Tensor vec2, *, Scalar beta=1, Scalar alpha=1, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: legacy::cpu::_th_addr_out - CUDA: legacy::cuda::_th_addr_out + CUDA: addr_out_cuda - func: _index_copy_(Tensor(a!) self, int dim, Tensor index, Tensor source) -> Tensor(a!) dispatch: CPU: legacy::cpu::_th_index_copy_ CUDA: legacy::cuda::_th_index_copy_ - func: _cumsum(Tensor self, int dim) -> Tensor use_c10_dispatcher: full dispatch: CPU: _cumsum_cpu - CUDA: legacy::cuda::_th_cumsum + CUDA: _cumsum_cuda - func: _cumsum.out(Tensor self, int dim, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: _cumsum_out_cpu - CUDA: legacy::cuda::_th_cumsum_out + CUDA: _cumsum_out_cuda - func: _cumprod(Tensor self, int dim) -> Tensor use_c10_dispatcher: full dispatch: CPU: _cumprod_cpu - CUDA: legacy::cuda::_th_cumprod + CUDA: _cumprod_cuda - func: _cumprod.out(Tensor self, int dim, *, Tensor(a!) out) -> Tensor(a!) dispatch: CPU: _cumprod_out_cpu - CUDA: legacy::cuda::_th_cumprod_out + CUDA: _cumprod_out_cuda -- func: _var(Tensor self, bool unbiased=True) -> Tensor - use_c10_dispatcher: full - dispatch: - CPU: legacy::cpu::_th_var - CUDA: legacy::cuda::_th_var - supports_named_tensor: True - -- func: _std(Tensor self, bool unbiased=True) -> Tensor - use_c10_dispatcher: full - dispatch: - CPU: legacy::cpu::_th_std - CUDA: legacy::cuda::_th_std - supports_named_tensor: True - - func: _amp_non_finite_check_and_unscale_(Tensor(a!) self, Tensor(b!) found_inf, Tensor inv_scale) -> () variants: function dispatch: CUDA: _amp_non_finite_check_and_unscale_cuda_ @@ -5326,10 +5229,11 @@ variants: function dispatch: CUDA: _amp_update_scale_cuda - func: _cat(Tensor[] tensors, int dim=0) -> Tensor + use_c10_dispatcher: full dispatch: CPU: _cat_cpu CUDA: cat_cuda QuantizedCPU: quantized_cat @@ -5338,39 +5242,54 @@ CPU: _cat_out_cpu CUDA: cat_out_cuda QuantizedCPU: quantized_cat_out - func: _mode(Tensor self, int dim=-1, bool keepdim=False) -> (Tensor, Tensor) + use_c10_dispatcher: full dispatch: CPU: legacy::cpu::_th_mode CUDA: legacy::cuda::_th_mode - func: _mode.values(Tensor self, int dim=-1, bool keepdim=False, *, Tensor(a!) values, Tensor(b!) indices) -> (Tensor(a!), Tensor(b!)) dispatch: CPU: legacy::cpu::_th_mode_out CUDA: legacy::cuda::_th_mode_out -- func: _max(Tensor self, int dim, bool keepdim=False) -> (Tensor, Tensor) +- func: bucketize.Tensor(Tensor self, Tensor boundaries, *, bool out_int32=False, bool right=False) -> Tensor + use_c10_dispatcher: full dispatch: - CPU: legacy::cpu::_th_max - CUDA: legacy::cuda::_th_max + CPU: bucketize_cpu + CUDA: bucketize_cuda -- func: _max.max(Tensor self, int dim, bool keepdim=False, *, Tensor(a!) max, Tensor(b!) max_indices) -> (Tensor(a!), Tensor(b!)) +- func: bucketize.Tensor_out(Tensor self, Tensor boundaries, *, bool out_int32=False, bool right=False, Tensor(a!) out) -> Tensor(a!) dispatch: - CPU: legacy::cpu::_th_max_out - CUDA: legacy::cuda::_th_max_out + CPU: bucketize_out_cpu + CUDA: bucketize_out_cuda -- func: _min(Tensor self, int dim, bool keepdim=False) -> (Tensor, Tensor) +- func: bucketize.Scalar(Scalar self, Tensor boundaries, *, bool out_int32=False, bool right=False) -> Tensor + use_c10_dispatcher: full dispatch: - CPU: legacy::cpu::_th_min - CUDA: legacy::cuda::_th_min + CPU: bucketize_cpu + CUDA: bucketize_cuda -- func: _min.min(Tensor self, int dim, bool keepdim=False, *, Tensor(a!) min, Tensor(b!) min_indices) -> (Tensor(a!), Tensor(b!)) +- func: searchsorted.Tensor(Tensor sorted_sequence, Tensor self, *, bool out_int32=False, bool right=False) -> Tensor + use_c10_dispatcher: full dispatch: - CPU: legacy::cpu::_th_min_out - CUDA: legacy::cuda::_th_min_out + CPU: searchsorted_cpu + CUDA: searchsorted_cuda +- func: searchsorted.Tensor_out(Tensor sorted_sequence, Tensor self, *, bool out_int32=False, bool right=False, Tensor(a!) out) -> Tensor(a!) + dispatch: + CPU: searchsorted_out_cpu + CUDA: searchsorted_out_cuda + +- func: searchsorted.Scalar(Tensor sorted_sequence, Scalar self, *, bool out_int32=False, bool right=False) -> Tensor + use_c10_dispatcher: full + dispatch: + CPU: searchsorted_cpu + CUDA: searchsorted_cuda + ## NN wrappers - func: mse_loss.out(Tensor self, Tensor target, int reduction=Mean, *, Tensor(a!) out) -> Tensor(a!) python_module: nn @@ -5444,10 +5363,11 @@ dispatch: CPU: multilabel_margin_loss_forward_out_cpu CUDA: legacy::cuda::_thnn_multilabel_margin_loss_forward_out - func: multilabel_margin_loss_forward(Tensor self, Tensor target, int reduction) -> (Tensor output, Tensor is_target) + use_c10_dispatcher: full python_module: nn dispatch: CPU: multilabel_margin_loss_forward_cpu CUDA: legacy::cuda::_thnn_multilabel_margin_loss_forward @@ -5558,22 +5478,14 @@ use_c10_dispatcher: full python_module: nn - func: elu.out(Tensor self, Scalar alpha=1, Scalar scale=1, Scalar input_scale=1, *, Tensor(a!) out) -> Tensor(a!) python_module: nn - dispatch: - CPU: elu_out - CUDA: elu_out - QuantizedCPU: quantized_elu_out - func: elu(Tensor self, Scalar alpha=1, Scalar scale=1, Scalar input_scale=1) -> Tensor use_c10_dispatcher: full python_module: nn - dispatch: - CPU: elu - CUDA: elu - QuantizedCPU: quantized_elu - func: elu_backward.grad_input(Tensor grad_output, Scalar alpha, Scalar scale, Scalar input_scale, Tensor output, *, Tensor(a!) grad_input) -> Tensor(a!) python_module: nn dispatch: CPU: elu_backward_out @@ -5583,14 +5495,10 @@ use_c10_dispatcher: full python_module: nn - func: elu_(Tensor(a!) self, Scalar alpha=1, Scalar scale=1, Scalar input_scale=1) -> Tensor(a!) python_module: nn - dispatch: - CPU: elu_ - CUDA: elu_ - QuantizedCPU: quantized_elu_ - func: glu.out(Tensor self, int dim=-1, *, Tensor(a!) out) -> Tensor(a!) python_module: nn dispatch: CPU: glu_out @@ -5620,17 +5528,24 @@ python_module: nn - func: hardsigmoid(Tensor self) -> Tensor use_c10_dispatcher: full python_module: nn + dispatch: + CPU: hardsigmoid + CUDA: hardsigmoid + QuantizedCPU: quantized_hardsigmoid - func: hardsigmoid_(Tensor(a!) self) -> Tensor(a!) python_module: nn - func: hardsigmoid_backward(Tensor grad_output, Tensor self) -> Tensor use_c10_dispatcher: full python_module: nn + dispatch: + CPU: hardsigmoid_backward + CUDA: hardsigmoid_backward - func: hardtanh.out(Tensor self, Scalar min_val=-1, Scalar max_val=1, *, Tensor(a!) out) -> Tensor(a!) python_module: nn dispatch: CPU: hardtanh_out @@ -5659,11 +5574,29 @@ python_module: nn dispatch: CPU: hardtanh_ CUDA: hardtanh_ QuantizedCPU: quantized_hardtanh_ + Vulkan: vulkan_hardtanh_ +- func: hardswish.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!) + python_module: nn + +- func: hardswish(Tensor self) -> Tensor + use_c10_dispatcher: full + python_module: nn + +- func: hardswish_(Tensor(a!) self) -> Tensor(a!) + python_module: nn + +- func: hardswish_backward(Tensor grad_output, Tensor self) -> Tensor + use_c10_dispatcher: full + python_module: nn + dispatch: + CPU: hardswish_backward + CUDA: hardswish_backward + - func: leaky_relu.out(Tensor self, Scalar negative_slope=0.01, *, Tensor(a!) out) -> Tensor(a!) python_module: nn dispatch: CPU: leaky_relu_out CUDA: leaky_relu_out @@ -5700,10 +5633,11 @@ dispatch: CPU: log_sigmoid_forward_out_cpu CUDA: legacy::cuda::_thnn_log_sigmoid_forward_out - func: log_sigmoid_forward(Tensor self) -> (Tensor output, Tensor buffer) + use_c10_dispatcher: full python_module: nn dispatch: CPU: log_sigmoid_forward_cpu CUDA: legacy::cuda::_thnn_log_sigmoid_forward @@ -5782,18 +5716,21 @@ CPU: adaptive_avg_pool2d_out_cpu CUDA: adaptive_avg_pool2d_out_cuda MkldnnCPU: mkldnn_adaptive_avg_pool2d_out - func: adaptive_avg_pool2d(Tensor self, int[2] output_size) -> Tensor + use_c10_dispatcher: full python_module: nn - func: mkldnn_adaptive_avg_pool2d(Tensor self, int[2] output_size) -> Tensor + use_c10_dispatcher: full dispatch: MkldnnCPU: mkldnn_adaptive_avg_pool2d requires_tensor: True - func: _adaptive_avg_pool2d(Tensor self, int[2] output_size) -> Tensor + use_c10_dispatcher: full dispatch: CPU: adaptive_avg_pool2d_cpu CUDA: adaptive_avg_pool2d_cuda QuantizedCPU: quantized_adaptive_avg_pool2d @@ -5809,10 +5746,11 @@ dispatch: CPU: adaptive_avg_pool3d_out_cpu CUDA: adaptive_avg_pool3d_out_cuda - func: adaptive_avg_pool3d(Tensor self, int[3] output_size) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: adaptive_avg_pool3d_cpu CUDA: adaptive_avg_pool3d_cuda @@ -5836,10 +5774,11 @@ CPU: adaptive_max_pool2d_out_cpu CUDA: adaptive_max_pool2d_out_cuda # Return: (Tensor output, Tensor indices) - func: adaptive_max_pool2d(Tensor self, int[2] output_size) -> (Tensor, Tensor) + use_c10_dispatcher: full python_module: nn dispatch: CPU: adaptive_max_pool2d_cpu CUDA: adaptive_max_pool2d_cuda @@ -5863,10 +5802,11 @@ CPU: adaptive_max_pool3d_out_cpu CUDA: adaptive_max_pool3d_out_cuda # Return: (Tensor output, Tensor indices) - func: adaptive_max_pool3d(Tensor self, int[3] output_size) -> (Tensor, Tensor) + use_c10_dispatcher: full python_module: nn dispatch: CPU: adaptive_max_pool3d_cpu CUDA: adaptive_max_pool3d_cuda @@ -5889,10 +5829,11 @@ CPU: avg_pool2d_out_cpu CUDA: avg_pool2d_out_cuda MkldnnCPU: mkldnn_avg_pool2d_out - func: avg_pool2d(Tensor self, int[2] kernel_size, int[2] stride=[], int[2] padding=0, bool ceil_mode=False, bool count_include_pad=True, int? divisor_override=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: avg_pool2d_cpu CUDA: avg_pool2d_cuda MkldnnCPU: mkldnn_avg_pool2d @@ -5903,10 +5844,11 @@ dispatch: CPU: avg_pool2d_backward_out_cpu CUDA: avg_pool2d_backward_out_cuda - func: avg_pool2d_backward(Tensor grad_output, Tensor self, int[2] kernel_size, int[2] stride, int[2] padding, bool ceil_mode, bool count_include_pad, int? divisor_override) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: avg_pool2d_backward_cpu CUDA: avg_pool2d_backward_cuda @@ -5915,10 +5857,11 @@ dispatch: CPU: avg_pool3d_out_cpu CUDA: avg_pool3d_out_cuda - func: avg_pool3d(Tensor self, int[3] kernel_size, int[3] stride=[], int[3] padding=0, bool ceil_mode=False, bool count_include_pad=True, int? divisor_override=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: avg_pool3d_cpu CUDA: avg_pool3d_cuda QuantizedCPU: quantized_avg_pool3d @@ -5928,10 +5871,11 @@ dispatch: CPU: avg_pool3d_backward_out_cpu CUDA: avg_pool3d_backward_out_cuda - func: avg_pool3d_backward(Tensor grad_output, Tensor self, int[3] kernel_size, int[3] stride, int[3] padding, bool ceil_mode, bool count_include_pad, int? divisor_override) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: avg_pool3d_backward_cpu CUDA: avg_pool3d_backward_cuda @@ -5942,10 +5886,11 @@ CPU: fractional_max_pool2d_out_cpu CUDA: fractional_max_pool2d_out_cuda # Return: (Tensor output, Tensor indices) - func: fractional_max_pool2d(Tensor self, int[2] kernel_size, int[2] output_size, Tensor random_samples) -> (Tensor, Tensor) + use_c10_dispatcher: full python_module: nn dispatch: CPU: fractional_max_pool2d_cpu CUDA: fractional_max_pool2d_cuda @@ -5954,10 +5899,11 @@ dispatch: CPU: fractional_max_pool2d_backward_out_cpu CUDA: fractional_max_pool2d_backward_out_cuda - func: fractional_max_pool2d_backward(Tensor grad_output, Tensor self, int[2] kernel_size, int[2] output_size, Tensor indices) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: fractional_max_pool2d_backward_cpu CUDA: fractional_max_pool2d_backward_cuda @@ -5968,10 +5914,11 @@ CPU: fractional_max_pool3d_out_cpu CUDA: fractional_max_pool3d_out_cuda # Return: (Tensor output, Tensor indices) - func: fractional_max_pool3d(Tensor self, int[3] kernel_size, int[3] output_size, Tensor random_samples) -> (Tensor, Tensor) + use_c10_dispatcher: full python_module: nn dispatch: CPU: fractional_max_pool3d_cpu CUDA: fractional_max_pool3d_cuda @@ -5980,10 +5927,11 @@ dispatch: CPU: fractional_max_pool3d_backward_out_cpu CUDA: fractional_max_pool3d_backward_out_cuda - func: fractional_max_pool3d_backward(Tensor grad_output, Tensor self, int[3] kernel_size, int[3] output_size, Tensor indices) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: fractional_max_pool3d_backward_cpu CUDA: fractional_max_pool3d_backward_cuda @@ -5994,23 +5942,24 @@ CPU: max_pool2d_with_indices_out_cpu CUDA: max_pool2d_with_indices_out_cuda # Return: (Tensor output, Tensor indices) - func: max_pool2d_with_indices(Tensor self, int[2] kernel_size, int[2] stride=[], int[2] padding=0, int[2] dilation=1, bool ceil_mode=False) -> (Tensor, Tensor) + use_c10_dispatcher: full python_module: nn dispatch: CPU: max_pool2d_with_indices_cpu CUDA: max_pool2d_with_indices_cuda - supports_named_tensor: True - func: max_pool2d_with_indices_backward.grad_input(Tensor grad_output, Tensor self, int[2] kernel_size, int[2] stride, int[2] padding, int[2] dilation, bool ceil_mode, Tensor indices, *, Tensor(a!) grad_input) -> Tensor(a!) python_module: nn dispatch: CPU: max_pool2d_with_indices_backward_out_cpu CUDA: max_pool2d_with_indices_backward_out_cuda - func: max_pool2d_with_indices_backward(Tensor grad_output, Tensor self, int[2] kernel_size, int[2] stride, int[2] padding, int[2] dilation, bool ceil_mode, Tensor indices) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: max_pool2d_with_indices_backward_cpu CUDA: max_pool2d_with_indices_backward_cuda @@ -6021,23 +5970,24 @@ CPU: max_pool3d_with_indices_out_cpu CUDA: max_pool3d_with_indices_out_cuda # Return: (Tensor output, Tensor indices) - func: max_pool3d_with_indices(Tensor self, int[3] kernel_size, int[3] stride=[], int[3] padding=0, int[3] dilation=1, bool ceil_mode=False) -> (Tensor, Tensor) + use_c10_dispatcher: full python_module: nn dispatch: CPU: max_pool3d_with_indices_cpu CUDA: max_pool3d_with_indices_cuda - supports_named_tensor: True - func: max_pool3d_with_indices_backward.grad_input(Tensor grad_output, Tensor self, int[3] kernel_size, int[3] stride, int[3] padding, int[3] dilation, bool ceil_mode, Tensor indices, *, Tensor(a!) grad_input) -> Tensor(a!) python_module: nn dispatch: CPU: max_pool3d_with_indices_backward_out_cpu CUDA: max_pool3d_with_indices_backward_out_cuda - func: max_pool3d_with_indices_backward(Tensor grad_output, Tensor self, int[3] kernel_size, int[3] stride, int[3] padding, int[3] dilation, bool ceil_mode, Tensor indices) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: max_pool3d_with_indices_backward_cpu CUDA: max_pool3d_with_indices_backward_cuda @@ -6046,10 +5996,11 @@ dispatch: CPU: max_unpooling2d_forward_out_cpu CUDA: max_unpooling2d_forward_out_cuda - func: max_unpool2d(Tensor self, Tensor indices, int[2] output_size) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: max_unpooling2d_forward_cpu CUDA: max_unpooling2d_forward_cuda @@ -6058,10 +6009,11 @@ dispatch: CPU: max_unpooling2d_backward_out_cpu CUDA: max_unpooling2d_backward_out_cuda - func: max_unpool2d_backward(Tensor grad_output, Tensor self, Tensor indices, int[2] output_size) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: max_unpooling2d_backward_cpu CUDA: max_unpooling2d_backward_cuda @@ -6070,10 +6022,11 @@ dispatch: CPU: max_unpooling3d_forward_out_cpu CUDA: max_unpooling3d_forward_out_cuda - func: max_unpool3d(Tensor self, Tensor indices, int[3] output_size, int[3] stride, int[3] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: max_unpooling3d_forward_cpu CUDA: max_unpooling3d_forward_cuda @@ -6082,10 +6035,11 @@ dispatch: CPU: max_unpooling3d_backward_out_cpu CUDA: max_unpooling3d_backward_out_cuda - func: max_unpool3d_backward(Tensor grad_output, Tensor self, Tensor indices, int[3] output_size, int[3] stride, int[3] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: max_unpooling3d_backward_cpu CUDA: max_unpooling3d_backward_cuda @@ -6094,22 +6048,25 @@ dispatch: CPU: reflection_pad1d_out_cpu CUDA: reflection_pad1d_out_cuda - func: reflection_pad1d(Tensor self, int[2] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: reflection_pad1d_cpu CUDA: reflection_pad1d_cuda + QuantizedCPU: reflection_pad1d_cpu - func: reflection_pad1d_backward.grad_input(Tensor grad_output, Tensor self, int[2] padding, *, Tensor(a!) grad_input) -> Tensor(a!) python_module: nn dispatch: CPU: reflection_pad1d_backward_out_cpu CUDA: reflection_pad1d_backward_out_cuda - func: reflection_pad1d_backward(Tensor grad_output, Tensor self, int[2] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: reflection_pad1d_backward_cpu CUDA: reflection_pad1d_backward_cuda @@ -6118,10 +6075,11 @@ dispatch: CPU: reflection_pad2d_out_cpu CUDA: reflection_pad2d_out_cuda - func: reflection_pad2d(Tensor self, int[4] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: reflection_pad2d_cpu CUDA: reflection_pad2d_cuda @@ -6130,10 +6088,11 @@ dispatch: CPU: reflection_pad2d_backward_out_cpu CUDA: reflection_pad2d_backward_out_cuda - func: reflection_pad2d_backward(Tensor grad_output, Tensor self, int[4] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: reflection_pad2d_backward_cpu CUDA: reflection_pad2d_backward_cuda @@ -6142,10 +6101,11 @@ dispatch: CPU: replication_pad1d_out_cpu CUDA: replication_pad1d_out_cuda - func: replication_pad1d(Tensor self, int[2] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: replication_pad1d_cpu CUDA: replication_pad1d_cuda @@ -6154,10 +6114,11 @@ dispatch: CPU: replication_pad1d_backward_out_cpu CUDA: replication_pad1d_backward_out_cuda - func: replication_pad1d_backward(Tensor grad_output, Tensor self, int[2] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: replication_pad1d_backward_cpu CUDA: replication_pad1d_backward_cuda @@ -6166,10 +6127,11 @@ dispatch: CPU: replication_pad2d_out_cpu CUDA: replication_pad2d_out_cuda - func: replication_pad2d(Tensor self, int[4] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: replication_pad2d_cpu CUDA: replication_pad2d_cuda @@ -6178,10 +6140,11 @@ dispatch: CPU: replication_pad2d_backward_out_cpu CUDA: replication_pad2d_backward_out_cuda - func: replication_pad2d_backward(Tensor grad_output, Tensor self, int[4] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: replication_pad2d_backward_cpu CUDA: replication_pad2d_backward_cuda @@ -6190,10 +6153,11 @@ dispatch: CPU: replication_pad3d_out_cpu CUDA: replication_pad3d_out_cuda - func: replication_pad3d(Tensor self, int[6] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: replication_pad3d_cpu CUDA: replication_pad3d_cuda @@ -6202,10 +6166,11 @@ dispatch: CPU: replication_pad3d_backward_out_cpu CUDA: replication_pad3d_backward_out_cuda - func: replication_pad3d_backward(Tensor grad_output, Tensor self, int[6] padding) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: replication_pad3d_backward_cpu CUDA: replication_pad3d_backward_cuda @@ -6214,10 +6179,11 @@ dispatch: CPU: upsample_linear1d_out_cpu CUDA: upsample_linear1d_out_cuda - func: upsample_linear1d(Tensor self, int[1] output_size, bool align_corners, float? scales=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_linear1d_cpu CUDA: upsample_linear1d_cuda @@ -6226,10 +6192,11 @@ dispatch: CPU: upsample_linear1d_backward_out_cpu CUDA: upsample_linear1d_backward_out_cuda - func: upsample_linear1d_backward(Tensor grad_output, int[1] output_size, int[3] input_size, bool align_corners, float? scales=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_linear1d_backward_cpu CUDA: upsample_linear1d_backward_cuda @@ -6238,10 +6205,11 @@ dispatch: CPU: upsample_bilinear2d_out_cpu CUDA: upsample_bilinear2d_out_cuda - func: upsample_bilinear2d(Tensor self, int[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_bilinear2d_cpu CUDA: upsample_bilinear2d_cuda QuantizedCPU: quantized_upsample_bilinear2d_cpu @@ -6251,10 +6219,11 @@ dispatch: CPU: upsample_bilinear2d_backward_out_cpu CUDA: upsample_bilinear2d_backward_out_cuda - func: upsample_bilinear2d_backward(Tensor grad_output, int[2] output_size, int[4] input_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_bilinear2d_backward_cpu CUDA: upsample_bilinear2d_backward_cuda @@ -6263,10 +6232,11 @@ dispatch: CPU: upsample_bicubic2d_out_cpu CUDA: upsample_bicubic2d_out_cuda - func: upsample_bicubic2d(Tensor self, int[2] output_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_bicubic2d_cpu CUDA: upsample_bicubic2d_cuda @@ -6275,10 +6245,11 @@ dispatch: CPU: upsample_bicubic2d_backward_out_cpu CUDA: upsample_bicubic2d_backward_out_cuda - func: upsample_bicubic2d_backward(Tensor grad_output, int[2] output_size, int[4] input_size, bool align_corners, float? scales_h=None, float? scales_w=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_bicubic2d_backward_cpu CUDA: upsample_bicubic2d_backward_cuda @@ -6287,10 +6258,11 @@ dispatch: CPU: upsample_trilinear3d_out_cpu CUDA: upsample_trilinear3d_out_cuda - func: upsample_trilinear3d(Tensor self, int[3] output_size, bool align_corners, float? scales_d=None, float? scales_h=None, float? scales_w=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_trilinear3d_cpu CUDA: upsample_trilinear3d_cuda @@ -6299,10 +6271,11 @@ dispatch: CPU: upsample_trilinear3d_backward_out_cpu CUDA: upsample_trilinear3d_backward_out_cuda - func: upsample_trilinear3d_backward(Tensor grad_output, int[3] output_size, int[5] input_size, bool align_corners, float? scales_d=None, float? scales_h=None, float? scales_w=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_trilinear3d_backward_cpu CUDA: upsample_trilinear3d_backward_cuda @@ -6311,10 +6284,11 @@ dispatch: CPU: upsample_nearest1d_out_cpu CUDA: upsample_nearest1d_out_cuda - func: upsample_nearest1d(Tensor self, int[1] output_size, float? scales=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_nearest1d_cpu CUDA: upsample_nearest1d_cuda @@ -6323,10 +6297,11 @@ dispatch: CPU: upsample_nearest1d_backward_out_cpu CUDA: upsample_nearest1d_backward_out_cuda - func: upsample_nearest1d_backward(Tensor grad_output, int[1] output_size, int[3] input_size, float? scales=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_nearest1d_backward_cpu CUDA: upsample_nearest1d_backward_cuda @@ -6335,23 +6310,26 @@ dispatch: CPU: upsample_nearest2d_out_cpu CUDA: upsample_nearest2d_out_cuda - func: upsample_nearest2d(Tensor self, int[2] output_size, float? scales_h=None, float? scales_w=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_nearest2d_cpu CUDA: upsample_nearest2d_cuda QuantizedCPU: quantized_upsample_nearest2d_cpu + Vulkan: upsample_nearest2d_vulkan - func: upsample_nearest2d_backward.grad_input(Tensor grad_output, int[2] output_size, int[4] input_size, float? scales_h=None, float? scales_w=None, *, Tensor(a!) grad_input) -> Tensor(a!) python_module: nn dispatch: CPU: upsample_nearest2d_backward_out_cpu CUDA: upsample_nearest2d_backward_out_cuda - func: upsample_nearest2d_backward(Tensor grad_output, int[2] output_size, int[4] input_size, float? scales_h=None, float? scales_w=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_nearest2d_backward_cpu CUDA: upsample_nearest2d_backward_cuda @@ -6360,10 +6338,11 @@ dispatch: CPU: upsample_nearest3d_out_cpu CUDA: upsample_nearest3d_out_cuda - func: upsample_nearest3d(Tensor self, int[3] output_size, float? scales_d=None, float? scales_h=None, float? scales_w=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_nearest3d_cpu CUDA: upsample_nearest3d_cuda QuantizedCPU: quantized_upsample_nearest3d_cpu @@ -6373,10 +6352,11 @@ dispatch: CPU: upsample_nearest3d_backward_out_cpu CUDA: upsample_nearest3d_backward_out_cuda - func: upsample_nearest3d_backward(Tensor grad_output, int[3] output_size, int[5] input_size, float? scales_d=None, float? scales_h=None, float? scales_w=None) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: upsample_nearest3d_backward_cpu CUDA: upsample_nearest3d_backward_cuda @@ -6435,10 +6415,11 @@ dispatch: CPU: slow_conv_transpose2d_backward_out_cpu CUDA: slow_conv_transpose2d_backward_out_cuda - func: slow_conv_transpose2d_backward.output_mask(Tensor grad_output, Tensor self, Tensor weight, int[2] kernel_size, int[2] stride, int[2] padding, int[2] output_padding, int[2] dilation, Tensor columns, Tensor ones, bool[3] output_mask) -> (Tensor grad_input, Tensor grad_weight, Tensor grad_bias) + use_c10_dispatcher: full python_module: nn dispatch: CPU: slow_conv_transpose2d_backward_cpu CUDA: slow_conv_transpose2d_backward_cuda @@ -6459,10 +6440,11 @@ dispatch: CPU: slow_conv_transpose3d_backward_out_cpu CUDA: slow_conv_transpose3d_backward_out_cuda - func: slow_conv_transpose3d_backward.output_mask(Tensor grad_output, Tensor self, Tensor weight, int[3] kernel_size, int[3] stride, int[3] padding, int[3] output_padding, int[3] dilation, Tensor finput, Tensor fgrad_input, bool[3] output_mask) -> (Tensor grad_input, Tensor grad_weight, Tensor grad_bias) + use_c10_dispatcher: full python_module: nn dispatch: CPU: slow_conv_transpose3d_backward_cpu CUDA: slow_conv_transpose3d_backward_cuda @@ -6486,17 +6468,18 @@ - func: thnn_conv2d_backward.grad_input(Tensor grad_output, Tensor self, Tensor weight, int[2] kernel_size, int[2] stride, int[2] padding, Tensor finput, Tensor fgrad_input, *, Tensor(a!)? grad_input, Tensor(b!)? grad_weight, Tensor(c!)? grad_bias) -> (Tensor(a!), Tensor(b!), Tensor(c!)) python_module: nn dispatch: CPU: slow_conv2d_backward_out_cpu - CUDA: legacy::cuda::_thnn_conv2d_backward_out + CUDA: slow_conv2d_backward_out_cuda - func: thnn_conv2d_backward.output_mask(Tensor grad_output, Tensor self, Tensor weight, int[2] kernel_size, int[2] stride, int[2] padding, Tensor finput, Tensor fgrad_input, bool[3] output_mask) -> (Tensor grad_input, Tensor grad_weight, Tensor grad_bias) + use_c10_dispatcher: full python_module: nn dispatch: CPU: slow_conv2d_backward_cpu - CUDA: legacy::cuda::_thnn_conv2d_backward + CUDA: slow_conv2d_backward_cuda - func: thnn_conv_depthwise2d.out(Tensor self, Tensor weight, int[2] kernel_size, Tensor? bias=None, int[2] stride=1, int[2] padding=0, int[2] dilation=1, *, Tensor(a!) out) -> Tensor(a!) python_module: nn - func: thnn_conv_depthwise2d(Tensor self, Tensor weight, int[2] kernel_size, Tensor? bias=None, int[2] stride=1, int[2] padding=0, int[2] dilation=1) -> Tensor @@ -6513,16 +6496,17 @@ CUDA: legacy::cuda::_thnn_conv_depthwise2d_forward - func: thnn_conv_depthwise2d_backward.grad_input(Tensor grad_output, Tensor self, Tensor weight, int[2] kernel_size, int[2] stride, int[2] padding, int[2] dilation, *, Tensor(a!)? grad_input, Tensor(b!)? grad_weight) -> (Tensor(a!), Tensor(b!)) python_module: nn dispatch: - CUDA: legacy::cuda::_thnn_conv_depthwise2d_backward_out + CUDA: thnn_conv_depthwise2d_backward_out - func: thnn_conv_depthwise2d_backward.output_mask(Tensor grad_output, Tensor self, Tensor weight, int[2] kernel_size, int[2] stride, int[2] padding, int[2] dilation, bool[2] output_mask) -> (Tensor grad_input, Tensor grad_weight) + use_c10_dispatcher: full python_module: nn dispatch: - CUDA: legacy::cuda::_thnn_conv_depthwise2d_backward + CUDA: thnn_conv_depthwise2d_backward - func: slow_conv3d.out(Tensor self, Tensor weight, int[3] kernel_size, Tensor? bias=None, int[3] stride=1, int[3] padding=0, *, Tensor(a!) out) -> Tensor(a!) python_module: nn - func: slow_conv3d(Tensor self, Tensor weight, int[3] kernel_size, Tensor? bias=None, int[3] stride=1, int[3] padding=0) -> Tensor @@ -6542,10 +6526,11 @@ python_module: nn dispatch: CPU: slow_conv3d_backward_out_cpu - func: slow_conv3d_backward.output_mask(Tensor grad_output, Tensor self, Tensor weight, int[3] kernel_size, int[3] stride, int[3] padding, Tensor finput, Tensor fgrad_input, bool[3] output_mask) -> (Tensor grad_input, Tensor grad_weight, Tensor grad_bias) + use_c10_dispatcher: full python_module: nn dispatch: CPU: slow_conv3d_backward_cpu - func: slow_conv_dilated2d(Tensor self, Tensor weight, int[2] kernel_size, Tensor? bias=None, int[2] stride=1, int[2] padding=0, int[2] dilation=1) -> Tensor @@ -6553,10 +6538,11 @@ dispatch: CPU: slow_conv_dilated2d_cpu CUDA: slow_conv_dilated2d_cuda - func: slow_conv_dilated2d_backward(Tensor grad_output, Tensor self, Tensor weight, int[2] kernel_size, int[2] stride, int[2] padding, int[2] dilation, bool[3] output_mask) -> (Tensor grad_input, Tensor grad_weight, Tensor grad_bias) + use_c10_dispatcher: full python_module: nn dispatch: CPU: slow_conv_dilated2d_backward_cpu CUDA: slow_conv_dilated2d_backward_cuda @@ -6565,10 +6551,11 @@ dispatch: CPU: slow_conv_dilated3d_cpu CUDA: slow_conv_dilated3d_cuda - func: slow_conv_dilated3d_backward(Tensor grad_output, Tensor self, Tensor weight, int[3] kernel_size, int[3] stride, int[3] padding, int[3] dilation, bool[3] output_mask) -> (Tensor grad_input, Tensor grad_weight, Tensor grad_bias) + use_c10_dispatcher: full python_module: nn dispatch: CPU: slow_conv_dilated3d_backward_cpu CUDA: slow_conv_dilated3d_backward_cuda @@ -6577,10 +6564,11 @@ dispatch: CPU: col2im_out_cpu CUDA: col2im_out_cuda - func: col2im(Tensor self, int[2] output_size, int[2] kernel_size, int[2] dilation, int[2] padding, int[2] stride) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: col2im_cpu CUDA: col2im_cuda @@ -6589,10 +6577,11 @@ dispatch: CPU: col2im_backward_out_cpu CUDA: col2im_backward_out_cuda - func: col2im_backward(Tensor grad_output, int[2] kernel_size, int[2] dilation, int[2] padding, int[2] stride) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: col2im_backward_cpu CUDA: col2im_backward_cuda @@ -6601,10 +6590,11 @@ dispatch: CPU: im2col_out_cpu CUDA: im2col_out_cuda - func: im2col(Tensor self, int[2] kernel_size, int[2] dilation, int[2] padding, int[2] stride) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: im2col_cpu CUDA: im2col_cuda @@ -6613,21 +6603,25 @@ dispatch: CPU: im2col_backward_out_cpu CUDA: im2col_backward_out_cuda - func: im2col_backward(Tensor grad_output, int[2] input_size, int[2] kernel_size, int[2] dilation, int[2] padding, int[2] stride) -> Tensor + use_c10_dispatcher: full python_module: nn dispatch: CPU: im2col_backward_cpu CUDA: im2col_backward_cuda - func: isfinite(Tensor self) -> Tensor use_c10_dispatcher: full - variants: function + variants: function, method device_guard: False - supports_named_tensor: True - func: isinf(Tensor self) -> Tensor use_c10_dispatcher: full - variants: function + variants: function, method device_guard: False - supports_named_tensor: True + +# Note: this function is only for testing. +# It is undocumented and should not be used outside of tests. +- func: _test_serialization_subcmul(Tensor self, Tensor other, Scalar alpha=1) -> Tensor + use_c10_dispatcher: full