| /external/tensorflow/tensorflow/contrib/model_pruning/ |
| README.md | 138 The convolution layer tensors are always pruned used block dimensions of [1,1].
|
| /external/tensorflow/tensorflow/contrib/quantize/python/ |
| quantize_graph_test.py | 215 """Add a basic convolution layer to the default graph."""
|
| /external/tensorflow/tensorflow/contrib/slim/python/slim/nets/ |
| inception_v3.py | 79 min_depth: Minimum depth value (number of channels) for all convolution ops.
83 for all convolution ops. The value must be greater than zero. Typical
542 min_depth: Minimum depth value (number of channels) for all convolution ops.
546 for all convolution ops. The value must be greater than zero. Typical
|
| /external/tensorflow/tensorflow/core/kernels/ |
| eigen_cuboid_convolution.h | 27 * \brief Applies a 3D convolution over a multichannel input voxel block.
|
| mkl_conv_grad_bias_ops.cc | 17 // layout and primitives, use MKL dnn primitives to compute convolution backward
|
| eigen_spatial_convolutions.h | [all...] |
| /external/tensorflow/tensorflow/core/kernels/neon/ |
| neon_depthwise_conv_op.cc | 65 // For 2D convolution, there should be 4 dimensions.
|
| /external/tensorflow/tensorflow/examples/tutorials/mnist/ |
| mnist_deep.py | 104 """conv2d returns a 2d convolution layer with full stride."""
|
| /external/tensorflow/tensorflow/compiler/tf2xla/kernels/ |
| pooling_ops.cc | 428 // * a convolution with a kernel consisting entirely of 1s, where the
437 // For an explanation of backpropagation for convolution, see the comments
460 // The input gradients are computed by a convolution of the output
|
| /external/tensorflow/tensorflow/python/tools/ |
| optimize_for_inference_lib.py | 208 scaling into the convolution weights. This function identifies the typical
369 """Merges preceding resize and mirror pad ops into a specialized convolution.
371 There's a common pattern of enlarging the input to a convolution using a
|
| /external/tensorflow/tensorflow/tools/graph_transforms/ |
| README.md | 363 This transform flattens atrous convolution, corresponding to a sequence of
366 to run graphs having atrous convolution on platforms that do not yet natively
380 multiplies the convolution's (or matrix multiplication's) weights with the Mul
426 Quantized operations like convolution or matrix multiplies take their inputs as
517 convolution's im2col patch generation. This transform looks out for that
519 the resizing and padding with the convolution.
[all...] |
| /external/tensorflow/tensorflow/docs_src/performance/xla/ |
| operation_semantics.md | 419 ## Conv (convolution)
429 ## ConvWithGeneralPadding (convolution)
434 Computes a convolution of the kind used in neural networks. Here, a convolution
455 for which convolution is carried out.
478 dimension before doing the convolution. `padding[0]` specifies the padding for
486 doing the convolution.
492 array. The holes are filled with a no-op value, which for convolution means
495 Dilation of the rhs is also called atrous convolution. For more details, see
497 convolution. For more details, see @{tf.nn.conv2d_transpose}
[all...] |
| /external/tensorflow/tensorflow/compiler/xla/service/ |
| shape_inference.cc | [all...] |
| /external/tensorflow/tensorflow/python/keras/_impl/keras/layers/ |
| core.py | 148 early convolution layers) then regular dropout will not regularize the
185 early convolution layers) then regular dropout will not regularize the
239 early convolution layers) then regular dropout will not regularize the
|
| /external/tensorflow/tensorflow/stream_executor/ |
| stream_executor_pimpl.h | 352 // Get the list of supported algorithms for the forward convolution opeartion.
356 // Get the list of supported algorithms for the backward convolution on data.
361 // Get the list of supported algorithms for the backward convolution on the
[all...] |
| /external/dng_sdk/source/ |
| dng_opcodes.h | 264 /// \brief Class to represent a filter opcode, such as a convolution.
|
| /external/eigen/unsupported/Eigen/CXX11/src/Tensor/ |
| TensorConvolution.h | 18 * \brief Tensor convolution class.
599 // Compute the convolution
659 // Convolution
725 // Convolution
[all...] |
| /external/mesa3d/docs/relnotes/ |
| 3.3 | 266 image convolution. This will (hopefully) be done for Mesa 3.5/3.6.
|
| /external/opencv/cv/src/ |
| cvpyramids.cpp | [all...] |
| /external/skia/src/gpu/effects/ |
| GrMatrixConvolutionEffect.cpp | 243 // Static function to create a 2D convolution
|
| /external/skqp/src/gpu/effects/ |
| GrMatrixConvolutionEffect.cpp | 243 // Static function to create a 2D convolution
|
| /external/tensorflow/tensorflow/compiler/tests/ |
| conv3d_test.py | 48 # Make a convolution op with the current settings, just to easily get
|
| /external/tensorflow/tensorflow/compiler/xla/client/ |
| computation_builder.cc | 579 InvalidArgument("Convolution arguments must have same number of "
588 "Convolution expects argument arrays with >= 3 dimensions. "
609 InvalidArgument("Convolution %s[%d] is out of bounds: %lld",
649 NoteError(InternalError("failed to verify convolution"));
[all...] |
| /external/tensorflow/tensorflow/compiler/xla/service/gpu/ |
| ir_emission_utils.cc | 138 // conv_result is the actual result of the convolution, and scratch_memory is
|
| ir_emitter.h | 77 Status HandleConvolution(HloInstruction* convolution) override;
|
