/external/tensorflow/tensorflow/compiler/xla/service/gpu/ |
D | gpu_conv_runner.cc | 91 params.config.output_descriptor, &output_buf, scratch_allocator, in RunGpuConvForward() 108 params.config.output_descriptor.feature_map_count()) in RunGpuConvForwardActivation() 109 .set_layout(params.config.output_descriptor.layout()); in RunGpuConvForwardActivation() 134 params.config.fusion->mode, params.config.output_descriptor, &output_buf, in RunGpuConvForwardActivation() 168 params.config.output_descriptor, output_buf, params.config.conv_desc, in RunGpuConvInternalImpl() 180 params.config.output_descriptor, output_buf, params.config.conv_desc, in RunGpuConvInternalImpl() 416 BatchDescriptor& output_descriptor = config.output_descriptor; in GetGpuConvConfig() local 417 output_descriptor = BatchDescriptor(effective_num_dimensions); in GetGpuConvConfig() 418 output_descriptor.set_layout(output_dl) in GetGpuConvConfig() 423 output_descriptor.set_spatial_dim( in GetGpuConvConfig() [all …]
|
D | gpu_conv_runner.h | 72 se::dnn::BatchDescriptor output_descriptor; member
|
D | gpu_conv_algorithm_picker.cc | 145 params.config.output_descriptor, params.output_buf, in GetMIOpenAlgorithms()
|
/external/tensorflow/tensorflow/stream_executor/rocm/ |
D | rocm_dnn.h | 239 const dnn::BatchDescriptor& output_descriptor, 312 const dnn::BatchDescriptor& output_descriptor, 327 const dnn::BatchDescriptor& output_descriptor, 341 const dnn::BatchDescriptor& output_descriptor, 356 const dnn::BatchDescriptor& output_descriptor, 371 const dnn::BatchDescriptor& output_descriptor, 383 const dnn::BatchDescriptor& output_descriptor, in DoConvolveQuantized() argument 396 const dnn::BatchDescriptor& output_descriptor, in DoConvolveQuantized() argument 409 const dnn::BatchDescriptor& output_descriptor, in DoSeparableConvolve() argument 593 const dnn::BatchDescriptor& output_descriptor, [all …]
|
D | rocm_dnn.cc | 1053 miopenTensorDescriptor_t output_descriptor, in Execute() argument 1057 output_descriptor, output_data, fusion_args_); in Execute() 2847 DeviceMemoryBase filter_data, const dnn::BatchDescriptor& output_descriptor, in DoPrepareForConvolution() argument 2904 BatchDescriptor* output_descriptor, DeviceMemoryBase backward_output_data, in MaybeTransformLayout() argument 2906 if (output_descriptor->layout() == dnn::DataLayout::kBatchDepthYX) { in MaybeTransformLayout() 2909 CHECK(output_descriptor->layout() == dnn::DataLayout::kBatchYXDepth); in MaybeTransformLayout() 2914 transformed_output_descriptor.CloneFrom(*output_descriptor); in MaybeTransformLayout() 2917 *output_descriptor, static_cast<miopenDataType_t>(miopen_type)}; in MaybeTransformLayout() 2934 output_descriptor->set_layout(dnn::DataLayout::kBatchDepthYX); in MaybeTransformLayout() 2943 DeviceMemoryBase filter_data, const dnn::BatchDescriptor& output_descriptor, in DoConvolve() argument [all …]
|
/external/tensorflow/tensorflow/stream_executor/cuda/ |
D | cuda_dnn.h | 274 const dnn::BatchDescriptor& output_descriptor, 289 const dnn::BatchDescriptor& output_descriptor, 303 const dnn::BatchDescriptor& output_descriptor, 318 const dnn::BatchDescriptor& output_descriptor, 333 const dnn::BatchDescriptor& output_descriptor, 347 const dnn::BatchDescriptor& output_descriptor, 359 const dnn::BatchDescriptor& output_descriptor, in DoConvolveQuantized() argument 372 const dnn::BatchDescriptor& output_descriptor, in DoConvolveQuantized() argument 385 const dnn::BatchDescriptor& output_descriptor, in DoSeparableConvolve() argument 626 const dnn::BatchDescriptor& output_descriptor, [all …]
|
D | cuda_dnn.cc | 2987 DeviceMemoryBase filter_data, const dnn::BatchDescriptor& output_descriptor, in DoPrepareForConvolution() argument 3000 output_descriptor, in DoPrepareForConvolution() 3001 ToCudnnDataType(element_type, output_descriptor.layout())); in DoPrepareForConvolution() 3043 DeviceMemoryBase filter_data, const dnn::BatchDescriptor& output_descriptor, in DoConvolve() argument 3050 CudnnTensorDescriptor output_nd(output_descriptor, in DoConvolve() 3186 const dnn::BatchDescriptor& output_descriptor, in DoFusedConvolveImpl() argument 3202 output_descriptor, in DoFusedConvolveImpl() 3240 activation_mode, CUDNN_NOT_PROPAGATE_NAN, output_descriptor.value_max()); in DoFusedConvolveImpl() 3765 const dnn::BatchDescriptor& output_descriptor, in DoFusedConvolve() argument 3773 output_descriptor, output_data, in DoFusedConvolve() [all …]
|
/external/tensorflow/tensorflow/stream_executor/ |
D | dnn.h | 1152 const dnn::BatchDescriptor& output_descriptor, in DoFusedConvolve() argument 1170 const dnn::BatchDescriptor& output_descriptor, in DoFusedConvolve() argument 1190 const dnn::BatchDescriptor& output_descriptor, in DoFusedConvolve() argument 1210 const dnn::BatchDescriptor& output_descriptor, in DoFusedConvolve() argument 1247 const BatchDescriptor& output_descriptor, in PrepareForConvolution() argument 1255 input_data, filter_descriptor, filter_data, output_descriptor, in PrepareForConvolution() 1298 DeviceMemoryBase filter_data, const BatchDescriptor& output_descriptor, 1310 const dnn::BatchDescriptor& output_descriptor, in DoConvolve() argument 1319 output_descriptor, *output_data, convolution_descriptor, in DoConvolve() 1335 const dnn::BatchDescriptor& output_descriptor, [all …]
|
D | stream.h | 302 const dnn::BatchDescriptor &output_descriptor, 312 const dnn::BatchDescriptor &output_descriptor, 322 const dnn::BatchDescriptor &output_descriptor, 332 const dnn::BatchDescriptor &output_descriptor, in ConvolveWithAlgorithm() argument 341 filter_descriptor, filter_data, output_descriptor, *output, in ConvolveWithAlgorithm() 347 input_data, filter_descriptor, filter_data, output_descriptor, in ConvolveWithAlgorithm() 363 const dnn::BatchDescriptor &output_descriptor, 377 const dnn::BatchDescriptor &output_descriptor, 392 const dnn::BatchDescriptor &output_descriptor, 406 const dnn::BatchDescriptor &output_descriptor, DeviceMemory<int8> *output, [all …]
|
D | stream.cc | 455 const dnn::BatchDescriptor &output_descriptor, DeviceMemory<double> *output, in FusedConvolveWithAlgorithm() argument 463 PARAM(activation_mode), PARAM(output_descriptor), PARAM(output), in FusedConvolveWithAlgorithm() 471 output_descriptor, output, scratch_allocator, algorithm_config, in FusedConvolveWithAlgorithm() 486 const dnn::BatchDescriptor &output_descriptor, DeviceMemory<float> *output, in FusedConvolveWithAlgorithm() argument 494 PARAM(activation_mode), PARAM(output_descriptor), PARAM(output), in FusedConvolveWithAlgorithm() 502 output_descriptor, output, scratch_allocator, algorithm_config, in FusedConvolveWithAlgorithm() 518 const dnn::BatchDescriptor &output_descriptor, in FusedConvolveWithAlgorithm() argument 527 PARAM(output_descriptor), PARAM(output), PARAM(algorithm_config)); in FusedConvolveWithAlgorithm() 534 output_descriptor, output, scratch_allocator, algorithm_config, in FusedConvolveWithAlgorithm() 549 const dnn::BatchDescriptor &output_descriptor, DeviceMemory<int8> *output, in FusedConvolveWithAlgorithm() argument [all …]
|
D | stream_executor_pimpl.cc | 281 DeviceMemoryBase filter_data, const dnn::BatchDescriptor &output_descriptor, in GetMIOpenConvolveAlgorithms() argument 292 filter_descriptor, filter_data, output_descriptor, output_data, in GetMIOpenConvolveAlgorithms()
|
D | stream_executor_pimpl.h | 375 const dnn::BatchDescriptor &output_descriptor,
|