| /external/tensorflow/tensorflow/compiler/tf2xla/kernels/ |
| const_op.cc | 55 shape.dim_sizes()));
63 shape.dim_sizes()));
71 shape.dim_sizes()));
82 shape.dim_sizes()));
93 shape.dim_sizes()));
101 shape.dim_sizes()));
109 shape.dim_sizes()));
|
| select_op.cc | 68 const auto dim_sizes = then_shape.dim_sizes(); variable
69 absl::Span<const int64> bdims = dim_sizes;
|
| broadcast_to_op.cc | 35 auto output = BroadcastTo(context->Input(0), output_shape.dim_sizes());
|
| aggregate_ops.cc | 51 ctx, sum_shape.dim_sizes() == operand_shape.dim_sizes(),
|
| clip_by_value_op.cc | 48 min = xla::Broadcast(min, shape.dim_sizes());
52 max = xla::Broadcast(max, shape.dim_sizes());
|
| cwise_ops.cc | 81 Computation(ctx, lhs_handle, lhs_shape.dim_sizes(), rhs_handle,
82 rhs_shape.dim_sizes(), bcast, extend_dimension);
|
| matrix_set_diag_op.cc | 69 indicator = xla::Broadcast(indicator, batch_shape.dim_sizes());
78 diag = xla::Add(diag, xla::Broadcast(zero, input_shape.dim_sizes()),
|
| index_ops_cpu.cc | 95 &b, xla::LiteralUtil::CreateR1<int64>(input_shape.dim_sizes())));
100 &b, xla::LiteralUtil::CreateR1<int64>(output_shape.dim_sizes())));
107 xla::S64, output_shape.dim_sizes());
|
| relu_op.cc | 77 xla::Broadcast(XlaHelpers::Zero(b, input_type(0)), shape.dim_sizes());
93 xla::Broadcast(XlaHelpers::Zero(b, input_type(0)), shape.dim_sizes());
95 XlaHelpers::IntegerLiteral(b, input_type(0), 6), shape.dim_sizes());
|
| matrix_band_part_op.cc | 82 indicator = xla::Broadcast(indicator, batch_shape.dim_sizes());
86 indicator, input, xla::Broadcast(zero_input, input_shape.dim_sizes()));
|
| tensor_list_ops.cc | 94 list_shape.dim_sizes());
181 ctx->SetOutput(0, xla::ConstantR1<int64>(b, shape.dim_sizes()));
185 for (int64 s : shape.dim_sizes()) {
227 auto slice_shape = shape.dim_sizes();
283 absl::Span<const int64>(tensor_shape.dim_sizes()).subspan(1)),
368 auto update = xla::Reshape(value, slice_shape.dim_sizes());
411 auto update = xla::Reshape(value, slice_shape.dim_sizes());
453 auto slice_shape = shape.dim_sizes();
|
| diag_op.cc | 88 auto dims = input_shape.dim_sizes();
127 auto dims = input_shape.dim_sizes();
172 auto dims = input_shape.dim_sizes();
198 auto dims = input_shape.dim_sizes();
|
| pack_op.cc | 79 reshaped_inputs[i] = xla::Reshape(values[i], child_shape.dim_sizes());
|
| shape_op.cc | 136 auto existing_dims = input_shape.dim_sizes();
170 auto existing_dims = input_shape.dim_sizes();
229 ctx->SetOutput(0, xla::Broadcast(zero, input_shape.dim_sizes()));
243 ctx->SetOutput(0, xla::Broadcast(one, input_shape.dim_sizes()));
|
| split_op.cc | 180 auto dim_sizes = input_shape.dim_sizes(); variable
181 std::vector<int64> limits(dim_sizes.begin(), dim_sizes.end());
|
| unpack_op.cc | 79 auto result = xla::Reshape(slice, output_shape.dim_sizes());
|
| tensor_array_ops.cc | 166 value = xla::Broadcast(zero, ta_shape.dim_sizes());
221 auto update = xla::Reshape(value, slice_shape.dim_sizes());
225 written = DynamicAddSlice(b, ta, update, slice_shape.dim_sizes(),
270 auto slice_shape = ta_shape.dim_sizes();
405 auto slice_dims = value_shape.dim_sizes();
409 auto value_ends = value_shape.dim_sizes();
461 auto ta_dims = ta_shape.dim_sizes();
533 const xla::XlaOp reshape = xla::Reshape(value, ta_shape.dim_sizes());
|
| reshape_op.cc | 95 ctx->SetOutput(0, xla::Reshape(ctx->Input(0), shape.dim_sizes()));
|
| sparse_to_dense_op.cc | 74 auto buffer = Broadcast(default_value, output_shape.dim_sizes());
|
| /external/tensorflow/tensorflow/python/ops/ragged/ |
| ragged_tensor_shape_test.py | 97 dict(dim_sizes=[], rank=0, expected_dim_sizes=[]),
98 dict(dim_sizes=[], rank=3, expected_dim_sizes=[1, 1, 1]),
99 dict(dim_sizes=[3], rank=1, expected_dim_sizes=[3]),
100 dict(dim_sizes=[3], rank=3, expected_dim_sizes=[1, 1, 3]),
101 dict(dim_sizes=[2, 3], rank=3, expected_dim_sizes=[1, 2, 3]),
102 dict(dim_sizes=[3, [3, 2, 4]], rank=2, expected_dim_sizes=[3, [3, 2, 4]]),
104 dim_sizes=[3, [3, 2, 4]],
108 dim_sizes=[3, [3, 2, 4], 2, 3],
112 def testBroadcastToRank(self, dim_sizes, rank, expected_dim_sizes):
113 shape = RaggedTensorDynamicShape.from_dim_sizes(dim_sizes)
[all...] |
| /external/tensorflow/tensorflow/core/util/ |
| tensor_format.h | 457 return GetTensorDim(gtl::ArraySlice<int64>(tensor_shape.dim_sizes()),
466 return GetFilterDim(gtl::ArraySlice<int64>(tensor_shape.dim_sizes()),
512 gtl::InlinedVector<int64, 6> dim_sizes(dims);
513 dim_sizes[GetTensorBatchDimIndex(dims, format)] = N;
521 dim_sizes[GetTensorInnerWidthDimIndex(dims, format)] = 4;
524 dim_sizes[GetTensorSpatialDimIndex(dims, format, dim)] = dim_size;
532 dim_sizes[GetTensorInnerFeatureDimIndex(dims, format)] = 4;
534 dim_sizes[feature_index] = C;
535 return TensorShape(dim_sizes);
546 gtl::InlinedVector<int64, 6> dim_sizes(dims)
[all...] |
| /external/tensorflow/tensorflow/compiler/xla/service/cpu/ |
| shape_partition_test.cc | 169 std::vector<int64> dim_sizes(num_outer_dims_to_partition);
175 dim_sizes[i] = shape.dimensions(dimension);
176 total_dim_size *= dim_sizes[i];
178 const int64 dim_partition_count = 1 + Rand() % dim_sizes[i];
202 EXPECT_EQ(expected_index, dim_sizes[i]);
|
| /external/tensorflow/tensorflow/core/framework/ |
| tensor_shape.cc | 136 TensorShapeBase<Shape>::TensorShapeBase(gtl::ArraySlice<int64> dim_sizes) {
139 InitDims(dim_sizes);
159 void TensorShapeBase<Shape>::InitDims(gtl::ArraySlice<int64> dim_sizes) {
168 for (auto s : dim_sizes) {
178 switch (dim_sizes.size()) {
181 const int64 size = dim_sizes[0];
188 const int64 size0 = dim_sizes[0];
189 const int64 size1 = dim_sizes[1];
197 const int64 size0 = dim_sizes[0];
198 const int64 size1 = dim_sizes[1]
[all...] |
| /external/tensorflow/tensorflow/compiler/tf2xla/ |
| xla_resource.cc | 132 xla::Broadcast(XlaHelpers::Zero(builder, type_), shape_.dim_sizes());
140 ta_shape.dim_sizes());
149 ta_shape.dim_sizes()),
173 xla::Broadcast(XlaHelpers::Zero(builder, type_), ta_shape.dim_sizes());
|
| xla_helpers.cc | 103 xla::Broadcast(on_value, output_shape.dim_sizes()),
104 xla::Broadcast(off_value, output_shape.dim_sizes()));
|
