| /external/tensorflow/tensorflow/stream_executor/cuda/ |
| D | cuda_blas.cc | 2604 int64 stride_c, int batch_count) { in DoBlasInternalImpl() argument
2622 stride_b, &beta, GpuMemoryMutable(c), CUDA_R_16F, ldc, stride_c, in DoBlasInternalImpl()
2638 reinterpret_cast<__half *>(GpuMemoryMutable(c) + batch * stride_c); in DoBlasInternalImpl()
2657 float beta, DeviceMemory<float> *c, int ldc, int64 stride_c, in DoBlasInternalImpl() argument
2668 GpuMemory(b), ldb, stride_b, &beta, GpuMemoryMutable(c), ldc, stride_c, in DoBlasInternalImpl()
2676 double beta, DeviceMemory<double> *c, int ldc, int64 stride_c, in DoBlasInternalImpl() argument
2682 GpuMemoryMutable(c), ldc, stride_c, batch_count); in DoBlasInternalImpl()
2691 int64 stride_c, int batch_count) { in DoBlasInternalImpl() argument
2699 GpuComplex(GpuMemoryMutable(c)), ldc, stride_c, batch_count); in DoBlasInternalImpl()
2708 int64 stride_c, int batch_count) { in DoBlasInternalImpl() argument
[all …]
|
| /external/tensorflow/tensorflow/core/kernels/ |
| D | conv_ops_using_gemm.cc | 447 const int64 stride_c = GetTensorDim(strides_, data_format_, 'C'); in Conv2DUsingGemmOp() local
449 context, stride_n == 1 && stride_c == 1, in Conv2DUsingGemmOp()
|
| D | depthwise_conv_op.cc | 285 const int64 stride_c = GetTensorDim(strides_, data_format_, 'C'); in DepthwiseConv2dNativeOp() local
292 context, (stride_n == 1 && stride_c == 1), in DepthwiseConv2dNativeOp()
|
| D | depthwise_conv_grad_op.cc | 566 const int64 stride_c = GetTensorDim(strides_, data_format_, 'C'); in DepthwiseConv2dNativeBackpropInputOp() local
573 context, (stride_n == 1 && stride_c == 1), in DepthwiseConv2dNativeBackpropInputOp()
1069 const int64 stride_c = GetTensorDim(strides_, data_format_, 'C'); in DepthwiseConv2dNativeBackpropFilterOp() local
1076 context, (stride_n == 1 && stride_c == 1), in DepthwiseConv2dNativeBackpropFilterOp()
|
| D | conv_ops_fused_image_transform.cc | 638 const int64 stride_c = GetTensorDim(strides_, FORMAT_NHWC, 'C'); in FusedResizeConv2DUsingGemmOp() local
640 context, stride_n == 1 && stride_c == 1, in FusedResizeConv2DUsingGemmOp()
|
| D | conv_grad_input_ops.h | 367 int stride_c = GetTensorDim(strides_, data_format_, 'C');
371 context, (stride_n == 1 && stride_c == 1),
|
| D | pooling_ops_3d.cc | 673 const int32 stride_c = GetTensorDim(stride_, data_format_, 'C'); in MaxPooling3dGradGradOp() local
674 OP_REQUIRES(context, ksize_c == 1 && stride_c == 1, in MaxPooling3dGradGradOp()
|
| D | conv_grad_filter_ops.cc | 271 int stride_c = GetTensorDim(strides_, data_format_, 'C'); in Conv2DBackpropFilterOp() local
275 context, (stride_n == 1 && stride_c == 1), in Conv2DBackpropFilterOp()
|
| D | conv_ops.cc | 372 const int64 stride_c = GetTensorDim(strides, data_format, 'C'); in InitConv2DParameters() local
376 stride_n == 1 && stride_c == 1, in InitConv2DParameters()
|
| /external/tensorflow/tensorflow/core/kernels/mkl/ |
| D | mkl_conv_ops.h | 570 int stride_c = GetTensorDim(strides_, data_format_, 'C'); in MklConvBackpropCommonOp() local
572 context, (stride_n == 1 && stride_c == 1), in MklConvBackpropCommonOp()
|
| D | mkl_conv_ops.cc | 450 const int64 stride_c = GetTensorDim(strides_, data_format_, 'C'); in MklConvOp() local
452 context, stride_n == 1 && stride_c == 1, in MklConvOp()
|
| /external/tensorflow/tensorflow/stream_executor/ |
| D | blas.h | 231 int64 stride_c = 0; member
1200 int64 stride_c, int batch_count) = 0;
1205 float beta, DeviceMemory<float> *c, int ldc, int64 stride_c,
1211 double beta, DeviceMemory<double> *c, int ldc, int64 stride_c,
1219 int64 stride_c, int batch_count) = 0;
1226 int64 stride_c, int batch_count) = 0;
2169 DeviceMemory<Eigen::half> *c, int ldc, int64 stride_c, int batch_count); \
2175 int64 stride_c, int batch_count); \
2181 DeviceMemory<double> *c, int ldc, int64 stride_c, int batch_count); \
2188 int64 stride_c, int batch_count); \
[all …]
|
| D | stream.cc | 4220 float beta, DeviceMemory<Eigen::half> *c, int ldc, int64 stride_c, in ThenBlasGemmStridedBatched() argument
4225 PARAM(stride_c), PARAM(batch_count)); in ThenBlasGemmStridedBatched()
4234 c, ldc, stride_c, batch_count); in ThenBlasGemmStridedBatched()
4241 float beta, DeviceMemory<float> *c, int ldc, int64 stride_c, in ThenBlasGemmStridedBatched() argument
4246 PARAM(stride_c), PARAM(batch_count)); in ThenBlasGemmStridedBatched()
4255 c, ldc, stride_c, batch_count); in ThenBlasGemmStridedBatched()
4262 double beta, DeviceMemory<double> *c, int ldc, int64 stride_c, in ThenBlasGemmStridedBatched() argument
4267 PARAM(stride_c), PARAM(batch_count)); in ThenBlasGemmStridedBatched()
4276 c, ldc, stride_c, batch_count); in ThenBlasGemmStridedBatched()
4285 int64 stride_c, int batch_count) { in ThenBlasGemmStridedBatched() argument
[all …]
|
| D | stream.h | 1478 int64 stride_c, int batch_count);
1483 float beta, DeviceMemory<float> *c, int ldc, int64 stride_c,
1489 double beta, DeviceMemory<double> *c, int ldc, int64 stride_c,
1497 int64 stride_c, int batch_count);
1504 int64 stride_c, int batch_count);
|
| /external/tensorflow/tensorflow/stream_executor/rocm/ |
| D | rocm_blas.cc | 2566 int64 stride_c, int batch_count) { in DoBlasGemmStridedBatched() argument
2579 reinterpret_cast<rocblas_half *>(GpuMemoryMutable(c)), ldc, stride_c, in DoBlasGemmStridedBatched()
2587 float beta, DeviceMemory<float> *c, int ldc, int64 stride_c, in DoBlasGemmStridedBatched() argument
2600 stride_c, batch_count); in DoBlasGemmStridedBatched()
2606 double beta, DeviceMemory<double> *c, int ldc, int64 stride_c, in DoBlasGemmStridedBatched() argument
2619 stride_c, batch_count); in DoBlasGemmStridedBatched()
2627 int64 stride_c, int batch_count) { in DoBlasGemmStridedBatched() argument
2633 complex_cast(beta), complex_cast(c), ldc, stride_c, in DoBlasGemmStridedBatched()
2642 int64 stride_c, int batch_count) { in DoBlasGemmStridedBatched() argument
2648 complex_cast(beta), complex_cast(c), ldc, stride_c, in DoBlasGemmStridedBatched()
|
| /external/tensorflow/tensorflow/lite/kernels/internal/optimized/ |
| D | legacy_optimized_ops.h | 2489 int stride_c = n; in Conv() local
2492 stride_a, b, stride_b, 0.0f, c, stride_c); in Conv()
|
| D | optimized_ops.h | 1301 int stride_c = n; in Conv() local
1304 stride_a, b, stride_b, 0.0f, c, stride_c); in Conv()
|