aboutsummaryrefslogtreecommitdiffhomepage
path: root/tensorflow/core/grappler/optimizers/layout_optimizer_test.cc
Commit message (Expand)AuthorAge
* Add support for dilation. This is previously missed and would result inGravatar Yao Zhang2018-06-06
* Fix kernel creation bug, due to constant folding always use CPU.Gravatar A. Unique TensorFlower2018-04-27
* Use the host implementation of vec permute op if the input on the host. NoteGravatar Yao Zhang2018-04-18
* Improve layout optimizer tests -- Evaluate nodes before and after optimizatio...Gravatar A. Unique TensorFlower2018-04-13
* Place data format op on CPU:0.Gravatar Yao Zhang2018-04-10
* Place data format op on host if input tensor is in host memory.Gravatar Yao Zhang2018-04-06
* Avoid live lock in a loop.Gravatar Yao Zhang2018-03-13
* Fix StridedSlice.Gravatar Yao Zhang2018-01-23
* Use suffix instead of prefix for nodes added by layout optimizer so thatGravatar Yao Zhang2018-01-08
* Merged commit includes the following changes:Gravatar A. Unique TensorFlower2017-12-21
* Support IdentityN op.Gravatar Yao Zhang2017-12-19
* Support more binary ops.Gravatar Yao Zhang2017-12-19
* Support merge and switch nodes.Gravatar Yao Zhang2017-12-15
* Avoid unnecessary layout transpose to input of ShapeN.Gravatar Yao Zhang2017-12-13
* Generate a unique name for const used by transpose node, so that we don't use...Gravatar Yao Zhang2017-12-13
* Support Shape and ShapeN ops.Gravatar Yao Zhang2017-12-13
* Simplify the naming of added transpose node. Refactor dimension check and the...Gravatar Yao Zhang2017-12-13
* Fix the handling of unknown rank. Previous code would wrongly treat a tensor ...Gravatar Yao Zhang2017-12-11
* Use DataFormatVecPermute instead Gather, which is very slow.Gravatar Yao Zhang2017-12-11
* Support non-const input sizes for Conv2DBackpropInput.Gravatar Yao Zhang2017-12-08
* Share constant creation code.Gravatar Yao Zhang2017-12-08
* If a graph is already optimized by layout optimizer, no need to and do not op...Gravatar Yao Zhang2017-12-06
* Support non-const axis for split and concat.Gravatar Yao Zhang2017-12-05
* Support a vector and a 4D tensor as inputs to a binary op.Gravatar Yao Zhang2017-12-05
* Share code of SplitProcessor and ConcatProcessor.Gravatar Yao Zhang2017-12-01
* Remove SliceProcessorConcatOffset, which is not robust as it modifies nodesGravatar Yao Zhang2017-12-01
* Automated g4 rollback of changelist 177505909Gravatar Yao Zhang2017-11-30
* Automated g4 rollback of changelist 177499365Gravatar Gunhan Gulsoy2017-11-30
* Support binary operations with a scalar and a 4d tensor as input; refactorGravatar Yao Zhang2017-11-30
* Simplify const node creation.Gravatar Yao Zhang2017-11-29
* Support op split.Gravatar Yao Zhang2017-11-27
* Do not convert layout for FusedBatchNormGrad if is_training is false (freezeGravatar Yao Zhang2017-11-22
* Only convert the layout if the node is placed on GPU.Gravatar Yao Zhang2017-11-22
* Preserve fetch nodes.Gravatar Yao Zhang2017-11-21
* Pass VirtualCluster to tf_optimizer to enable cross-optimization of models.Gravatar Yao Zhang2017-11-21
* Fix a bug where op name (instead of node name) should be used.Gravatar Yao Zhang2017-11-10
* Support pad node.Gravatar Yao Zhang2017-11-04
* Produce valid node names for transpose.Gravatar Jingyue Wu2017-08-19
* Add statically inferred output shapes to the graph. Update aGravatar Yao Zhang2017-06-08
* When converting the layout of Conv2DBackpropInput, we need to permute one ofGravatar Yao Zhang2017-06-05
* Add an option to not convert layout if GEMM is used internally in Conv2D, Con...Gravatar Yao Zhang2017-05-04
generated by cgit v1.2.3 (git 2.39.1) at 2024-06-05 22:52:02 +0000