Home
last modified time | relevance | path

Searched defs:scalar_t (Results 1 – 25 of 68) sorted by relevance

123

/aosp_15_r20/external/pytorch/aten/src/ATen/cuda/
H A DCUDASparseBlas.h17 #define CUSPARSE_CSRGEAM2_BUFFERSIZE_ARGTYPES(scalar_t) \ argument
30 CUSPARSE_CSRGEAM2_BUFFERSIZE_ARGTYPES(scalar_t)) { in csrgeam2_bufferSizeExt()
77 #define CUSPARSE_CSRGEAM2_ARGTYPES(scalar_t) \ argument
89 inline void csrgeam2(CUSPARSE_CSRGEAM2_ARGTYPES(scalar_t)) { in csrgeam2()
107 #define CUSPARSE_BSRMM_ARGTYPES(scalar_t) \ argument
116 inline void bsrmm(CUSPARSE_BSRMM_ARGTYPES(scalar_t)) { in bsrmm()
132 #define CUSPARSE_BSRMV_ARGTYPES(scalar_t) \ argument
140 inline void bsrmv(CUSPARSE_BSRMV_ARGTYPES(scalar_t)) { in bsrmv()
158 #define CUSPARSE_BSRSV2_BUFFER_ARGTYPES(scalar_t) \ argument
166 inline void bsrsv2_bufferSize(CUSPARSE_BSRSV2_BUFFER_ARGTYPES(scalar_t)) { in bsrsv2_bufferSize()
[all …]
/aosp_15_r20/external/pytorch/aten/src/ATen/mkl/
H A DSparseBlas.h19 #define MKL_SPARSE_CREATE_CSR_ARGTYPES(scalar_t) \ argument
25 inline void create_csr(MKL_SPARSE_CREATE_CSR_ARGTYPES(scalar_t)) { in create_csr()
43 #define MKL_SPARSE_CREATE_BSR_ARGTYPES(scalar_t) \ argument
50 inline void create_bsr(MKL_SPARSE_CREATE_BSR_ARGTYPES(scalar_t)) { in create_bsr()
68 #define MKL_SPARSE_MV_ARGTYPES(scalar_t) \ argument
74 inline void mv(MKL_SPARSE_MV_ARGTYPES(scalar_t)) { in mv()
90 #define MKL_SPARSE_ADD_ARGTYPES(scalar_t) \ argument
95 inline void add(MKL_SPARSE_ADD_ARGTYPES(scalar_t)) { in add()
111 #define MKL_SPARSE_EXPORT_CSR_ARGTYPES(scalar_t) \ argument
117 inline void export_csr(MKL_SPARSE_EXPORT_CSR_ARGTYPES(scalar_t)) { in export_csr()
[all …]
/aosp_15_r20/external/pytorch/aten/src/ATen/native/cuda/linalg/
H A DCUDASolver.h264 #define CUDASOLVER_GEQRF_BUFFERSIZE_ARGTYPES(scalar_t) \ argument
268 void geqrf_bufferSize(CUDASOLVER_GEQRF_BUFFERSIZE_ARGTYPES(scalar_t)) { in geqrf_bufferSize()
283 #define CUDASOLVER_GEQRF_ARGTYPES(scalar_t) \ argument
288 void geqrf(CUDASOLVER_GEQRF_ARGTYPES(scalar_t)) { in geqrf()
439 #define CUDASOLVER_SYEVD_BUFFERSIZE_ARGTYPES(scalar_t, value_t) \ argument
444 void syevd_bufferSize(CUDASOLVER_SYEVD_BUFFERSIZE_ARGTYPES(scalar_t, value_t)) { in syevd_bufferSize()
462 #define CUDASOLVER_SYEVD_ARGTYPES(scalar_t, value_t) \ argument
468 void syevd(CUDASOLVER_SYEVD_ARGTYPES(scalar_t, value_t)) { in syevd()
484 #define CUDASOLVER_SYEVJ_BUFFERSIZE_ARGTYPES(scalar_t, value_t) \ argument
490 void syevj_bufferSize(CUDASOLVER_SYEVJ_BUFFERSIZE_ARGTYPES(scalar_t, value_t)) { in syevj_bufferSize()
[all …]
/aosp_15_r20/external/pytorch/aten/src/ATen/native/cuda/
H A DRNN.cu101 TensorInfo<scalar_t, index_type> input, in lstm_cell_forward()
188 TensorInfo<scalar_t, index_type> storage, in lstm_cell_backward()
253 TensorInfo<scalar_t, index_type> Input, in gru_cell_forward()
323 TensorInfo<scalar_t, index_type> gradInInput, in gru_cell_backward()
H A DGridSampler.cu28 TensorInfo<const scalar_t, index_t> input, in grid_sampler_2d_kernel()
30 TensorInfo<scalar_t, index_t> output, in grid_sampler_2d_kernel()
158 TensorInfo<const scalar_t, index_t> input, in grid_sampler_3d_kernel()
160 TensorInfo<scalar_t, index_t> output, in grid_sampler_3d_kernel()
314 TensorInfo<const scalar_t, index_t> grad_output, in grid_sampler_2d_backward_kernel()
317 …TensorInfo<scalar_t, index_t> grad_input, // initialized to zeros (or unused if input_requires_gr… in grid_sampler_2d_backward_kernel()
520 TensorInfo<const scalar_t, index_t> grad_output, in grid_sampler_3d_backward_kernel()
523 …TensorInfo<scalar_t, index_t> grad_input, // initialized to zeros (or unused if input_requires_gr… in grid_sampler_3d_backward_kernel()
H A DReplicationPadding.cu42 PackedTensorAccessor64<const scalar_t, 3> input, in replication_pad_forward_kernel1d()
43 PackedTensorAccessor64<scalar_t, 3> output, in replication_pad_forward_kernel1d()
66 PackedTensorAccessor64<scalar_t, 3> gradInput, in replication_pad_backward_kernel()
67 PackedTensorAccessor64<const scalar_t, 3> gradOutput, in replication_pad_backward_kernel()
90 PackedTensorAccessor64<const scalar_t, 4> input, in replication_pad_forward_kernel2d()
91 PackedTensorAccessor64<scalar_t, 4> output, in replication_pad_forward_kernel2d()
119 PackedTensorAccessor64<scalar_t, 4> gradInput, in replication_pad_backward_kernel()
120 PackedTensorAccessor64<const scalar_t, 4> gradOutput, in replication_pad_backward_kernel()
148 PackedTensorAccessor64<const scalar_t, 5> input, in replication_pad_forward_kernel3d()
149 PackedTensorAccessor64<scalar_t, 5> output, in replication_pad_forward_kernel3d()
[all …]
H A DAveragePool3d.cu37 PackedTensorAccessor64<const scalar_t, 4> input, in avg_pool3d_cuda_update_output()
38 PackedTensorAccessor64<scalar_t, 4> output, in avg_pool3d_cuda_update_output()
106 PackedTensorAccessor64<const scalar_t, 4> input, in avg_pool3d_cuda_update_output()
107 PackedTensorAccessor64<scalar_t, 4> output, in avg_pool3d_cuda_update_output()
172 PackedTensorAccessor64<const scalar_t, 4> gradOutput, in avg_pool3d_single_backward_out_frame_stride1()
173 PackedTensorAccessor64<scalar_t, 4> gradInput, in avg_pool3d_single_backward_out_frame_stride1()
217 PackedTensorAccessor64<const scalar_t, 4> gradOutput, in avg_pool3d_cuda_update_grad_input_atomic()
218 PackedTensorAccessor64<scalar_t, 4> gradInput, in avg_pool3d_cuda_update_grad_input_atomic()
276 PackedTensorAccessor64<const scalar_t, 4> gradOutput, in avg_pool3d_cuda_update_grad_input()
277 PackedTensorAccessor64<scalar_t, 4> gradInput, in avg_pool3d_cuda_update_grad_input()
H A DDepthwiseConv3d.cu29 const PackedTensorAccessor32<const scalar_t, 5> input, in conv_depthwise3d_cuda_kernel()
30 PackedTensorAccessor32<scalar_t, 5> output, in conv_depthwise3d_cuda_kernel()
102 const PackedTensorAccessor32<const scalar_t, 5> grad_output, in conv_depthwise3d_cuda_backward_input_kernel()
103 PackedTensorAccessor32<scalar_t, 5> grad_input, in conv_depthwise3d_cuda_backward_input_kernel()
183 const PackedTensorAccessor32<const scalar_t, 5> grad_output, in conv_depthwise3d_cuda_backward_weight_kernel()
185 PackedTensorAccessor32<scalar_t, 5> grad_kernel, in conv_depthwise3d_cuda_backward_weight_kernel()
H A DFractionalMaxPool2d.cu48 PackedTensorAccessor<scalar_t, 4> output, in fractional_max_pool2d_out_cuda_frame()
50 PackedTensorAccessor<const scalar_t, 4> input, in fractional_max_pool2d_out_cuda_frame()
105 PackedTensorAccessor<scalar_t, 4> gradInput, in fractional_max_pool2d_backward_out_cuda_frame()
106 PackedTensorAccessor<const scalar_t, 4> gradOutput, in fractional_max_pool2d_backward_out_cuda_frame()
H A DFractionalMaxPool3d.cu56 PackedTensorAccessor64<const scalar_t, 5> input, in fractional_max_pool3d_out_frame()
57 PackedTensorAccessor64<scalar_t, 5> output, in fractional_max_pool3d_out_frame()
122 PackedTensorAccessor64<scalar_t, 5> gradInput, in fractional_max_pool3d_backward_out_frame()
123 PackedTensorAccessor64<const scalar_t, 5> gradOutput, in fractional_max_pool3d_backward_out_frame()
H A DPowKernel.cu44 using scalar_t = c10::complex<at::Half>; in pow_scalar_tensor_impl() typedef
80 using scalar_t = c10::complex<at::Half>; in pow_chalf_tensor_scalar_impl() typedef
102 using scalar_t = c10::complex<at::Half>; in pow_tensor_tensor_kernel() typedef
183 using scalar_t = c10::complex<at::Half>; in pow_tensor_scalar_kernel() typedef
H A DSorting.cu25 cuda::detail::TensorInfo<const scalar_t, index_t> input, in gatherKthValue()
30 cuda::detail::TensorInfo<scalar_t, index_t> kthValue, in gatherKthValue()
93 cuda::detail::TensorInfo<scalar_t, index_t> values, in gatherMedian()
95 cuda::detail::TensorInfo<const scalar_t, index_t> input, in gatherMedian()
H A DUpSampleLinear1d.cu31 const PackedTensorAccessor64<const scalar_t, 3> idata, in upsample_linear1d_out_frame()
32 PackedTensorAccessor64<scalar_t, 3> odata) { in upsample_linear1d_out_frame()
78 PackedTensorAccessor64<scalar_t, 3> idata, in upsample_linear1d_out_frame_backward()
79 const PackedTensorAccessor64<const scalar_t, 3> odata) { in upsample_linear1d_out_frame_backward()
H A DUpSampleBicubic2d.cu29 const PackedTensorAccessor64<const scalar_t, 4> idata, in upsample_bicubic2d_out_frame()
30 PackedTensorAccessor64<scalar_t, 4> odata) { in upsample_bicubic2d_out_frame()
104 PackedTensorAccessor64<scalar_t, 4> idata, in upsample_bicubic2d_backward_out_frame()
105 const PackedTensorAccessor64<const scalar_t, 4> odata) { in upsample_bicubic2d_backward_out_frame()
H A DDropout.cu48 fused_dropout_kernel_vec(at::cuda::detail::TensorInfo<const scalar_t, IndexType> a, in fused_dropout_kernel_vec()
49 at::cuda::detail::TensorInfo<scalar_t, IndexType> b, in fused_dropout_kernel_vec()
136 fused_dropout_kernel(cuda::detail::TensorInfo<const scalar_t, IndexType> a, in fused_dropout_kernel()
137 cuda::detail::TensorInfo<scalar_t, IndexType> b, in fused_dropout_kernel()
H A DUpSampleTrilinear3d.cu46 const PackedTensorAccessor64<const scalar_t, 5> idata, in upsample_trilinear3d_out_frame()
47 PackedTensorAccessor64<scalar_t, 5> odata) { in upsample_trilinear3d_out_frame()
130 PackedTensorAccessor64<scalar_t, 5> idata, in upsample_trilinear3d_backward_out_frame()
131 const PackedTensorAccessor64<const scalar_t, 5> odata, in upsample_trilinear3d_backward_out_frame()
H A DUpSampleBilinear2d.cu40 const PackedTensorAccessor<const scalar_t, 4> idata, in upsample_bilinear2d_out_frame()
41 PackedTensorAccessor<scalar_t, 4> odata) { in upsample_bilinear2d_out_frame()
477 const PackedTensorAccessor64<const scalar_t, 4> idata, in upsample_gen2d_aa_out_frame()
478 PackedTensorAccessor64<scalar_t, 4> odata, in upsample_gen2d_aa_out_frame()
571 PackedTensorAccessor64<scalar_t, 4> idata, in upsample_gen2d_aa_backward_out_frame()
572 const PackedTensorAccessor64<const scalar_t, 4> odata, in upsample_gen2d_aa_backward_out_frame()
H A DEmbeddingBackwardKernel.cu87 acc_type<scalar_t, true> *grad_weight_per_segment, in compute_grad_weight_bags()
132 acc_type<scalar_t, true> *grad_weight_per_segment, in compute_grad_weight()
163 const acc_type<scalar_t, true> *grad_weight_per_segment, in sum_and_scatter()
/aosp_15_r20/external/pytorch/aten/src/ATen/native/
H A DConvolutionMM3d.cpp264 TensorAccessor<const scalar_t, 4> input, in slow_conv3d_update_output_frame()
265 TensorAccessor<scalar_t, 4> output, in slow_conv3d_update_output_frame()
313 TensorAccessor<scalar_t, 4> grad_input, in slow_conv3d_backward_update_grad_input_frame()
314 TensorAccessor<const scalar_t, 4> grad_output, in slow_conv3d_backward_update_grad_input_frame()
466 TensorAccessor<scalar_t, 2> grad_weight, in slow_conv3d_backward_weight_frame()
467 TensorAccessor<const scalar_t, 4> grad_output, in slow_conv3d_backward_weight_frame()
H A DConvolutionMM2d.cpp223 TensorAccessor<const scalar_t, 3> input, in slow_conv2d_update_output_frame()
224 TensorAccessor<scalar_t, 3> output, in slow_conv2d_update_output_frame()
287 TensorAccessor<scalar_t, 3> grad_input, in slow_conv2d_backward_update_grad_input_frame()
288 TensorAccessor<const scalar_t, 3> grad_output, in slow_conv2d_backward_update_grad_input_frame()
436 TensorAccessor<scalar_t, 2> grad_weight, in slow_conv2d_backward_weight_frame()
437 TensorAccessor<const scalar_t, 3> grad_output, in slow_conv2d_backward_weight_frame()
/aosp_15_r20/external/pytorch/aten/src/ATen/native/cpu/
H A DSumKernel.cpp92 using scalar_t = vechold_type<vec_t>; typedef
114 using scalar_t = vechold_type<vec_t>; typedef
137 using scalar_t = vechold_type<vec_t>; typedef
161 using scalar_t = vechold_type<vec_t>; typedef
229 using scalar_t = vechold_type<vec_t>; typedef
251 using scalar_t = vechold_type<vec_t>; typedef
294 const std::array<scalar_t, numel> &values) { in store()
H A DLinearAlgebraKernel.cpp16 using scalar_t = bool; in addr_kernel() typedef
26 scalar_t vec2_val) __ubsan_ignore_undefined__ -> scalar_t { in addr_kernel() argument
59 scalar_t vec2_val) __ubsan_ignore_undefined__ -> scalar_t { in addr_kernel() argument
H A DReduceAllOpsKernel.cpp35 [&](int64_t start, int64_t end, const scalar_t /*ident*/) -> scalar_t { in reduce_all_impl_vec()
121 const std::pair<scalar_t, scalar_t>& ident_v, in reduce_all_impl_two_outputs()
146 const std::pair<scalar_t, scalar_t>& ident_v, in reduce_all_impl_vec_two_outputs()
/aosp_15_r20/external/pytorch/aten/src/ATen/native/transformers/cuda/
H A Dattention.cu100 const PackedTensorAccessor64<scalar_t, 3, RestrictPtrTraits> qkv, in transform_bias_rescale_qkv_kernel()
201 const PackedTensorAccessor64<scalar_t, 1, RestrictPtrTraits> qkv, in transform_bias_rescale_qkv_add_padding_kernel()
1139 using scalar_t = typename Kernel::scalar_t; in _efficient_attention_forward() typedef
/aosp_15_r20/external/pytorch/aten/src/ATen/native/sparse/cuda/
H A DSoftMax.cu98 PackedTensorAccessor<scalar_t, 2> input_values_acc, in cuda_sparse_coo_softmax_kernel()
159 PackedTensorAccessor<scalar_t, 2> values_accessor, in cuda_sparse_coo_softmax_backward_kernel()

123