index
:
tensorflow
master
machine learning framework
about
summary
refs
log
tree
commit
diff
homepage
log msg
author
committer
range
path:
root
/
tensorflow
/
core
/
kernels
/
conv_ops.cc
Commit message (
Expand
)
Author
Age
*
Extract Conv2D dimensions parsing and validation into helper functions.
Eugene Zhulenev
2018-09-26
*
Do not segfault in Conv2d/3d if cuDNN version is too low.
Reed Wanderman-Milne
2018-09-17
*
Generalize TransformFilter method in preparation of NHWC Conv support
Smit Hinsu
2018-09-14
*
Add data format as a parameter in ConvParameters to support NHWC format
Smit Hinsu
2018-07-24
*
Adding a depthwise convolution kernel op (with label 'cudnn_grouped_convoluti...
A. Unique TensorFlower
2018-04-30
*
Use tensorflow::se instead of perftools::gputools for StreamExecutor.
Justin Lebar
2018-04-23
*
Eliminate the guard around Winograd non-fused convolutions with cudnn7.
A. Unique TensorFlower
2018-04-20
*
Adds float64 support for Conv2d, Conv2dBackpropInput, and Conv2dBackpropFilter
Brian Patton
2018-03-22
*
Moved LIBXSMM convolutions to a separate --define flag so that they are disab...
A. Unique TensorFlower
2018-02-22
*
Activates Eigen path for CPU implementation of atrous/dilated convolution (on...
A. Unique TensorFlower
2018-02-16
*
Cleanup: Ran clang-format on all *.{cc,h} files in tensorflow/core/kernels.
A. Unique TensorFlower
2018-01-26
*
Return unimplemented error when trying to use dilated rate > 1 combined with ...
Yangzihao Wang
2017-12-12
*
Add native dilated support for conv2d and its gradients in cudnn v>=6.
Yangzihao Wang
2017-11-29
*
Merge changes from github.
Frank Chen
2017-10-06
*
Merge changes from github.
Shanqing Cai
2017-09-25
*
Fix "wrong enum" comparison.
A. Unique TensorFlower
2017-09-05
*
Exposing launchpad for conv2d backprop, and unify launchpads for conv2d and d...
A. Unique TensorFlower
2017-09-04
*
Allows cudnn convolution algorithms to proceed when only find one type of alg...
Yangzihao Wang
2017-08-01
*
Merge changes from github.
Jonathan Hseu
2017-06-09
*
Set winograd nofused flag to be true by default.
Yangzihao Wang
2017-05-28
*
Don't assert on unimplemented layout
Benoit Steiner
2017-04-17
*
Merge changes from github.
A. Unique TensorFlower
2017-04-04
*
Add a multi-pass mode for autotuning. It tests the autotune multiple times,
Xiaoqiang Zheng
2017-03-21
*
Added data type info to conv autotune parameters.
Yangzihao Wang
2017-03-17
*
Generalize autotuning to 3d ops.
A. Unique TensorFlower
2017-03-15
*
Merge changes from github.
Benoit Steiner
2017-02-08
*
Run convolution as GEMM if the input activations and filters have the same sp...
A. Unique TensorFlower
2017-01-30
*
Merge changes from github.
Shanqing Cai
2017-01-23
*
Added support for TensorFlow backward convolutions to use libxsmm; weight
A. Unique TensorFlower
2016-12-20
*
Changed parameters and file locations to use libxsmm 1.6.1.
A. Unique TensorFlower
2016-12-05
*
Added support for some use of libxsmm to apply convolutions.
A. Unique TensorFlower
2016-11-30
*
Fix a bug in the default cudnn workspace size.
A. Unique TensorFlower
2016-11-28
*
Change Autotune to the global scope. Switch to hash table for the autotune
Xiaoqiang Zheng
2016-10-10
*
Experimental Conv2D implementation specialized for deep convolutions (i.e. la...
A. Unique TensorFlower
2016-08-31
*
Merge changes from github.
A. Unique TensorFlower
2016-08-26
*
Relax static shape information requirements for depthwise and separable conv2d.
A. Unique TensorFlower
2016-08-17
*
Remove unnecessary copies from conv ops and pad ops.
A. Unique TensorFlower
2016-08-05
*
Clip the padding for negative values.
Xiaoqiang Zheng
2016-06-30
*
Improve convolution autotune process. The max batch size VGG model can handle
Xiaoqiang Zheng
2016-06-21
*
Refactor Get2dOutputSizes/Get2dOutputSizesVerbose/Get3dOutputSizes to share a...
A. Unique TensorFlower
2016-06-20
*
Change some kernels to use TF_CALL* macros, so that the instantiations for some
A. Unique TensorFlower
2016-06-06
*
Update copyright for 3p/tf/core.
A. Unique TensorFlower
2016-06-02
*
Added support for convolutions of 16bit floats on CPU
Benoit Steiner
2016-05-31
*
Enable fp16 for convolution operations, gated on CUDA 7.5. (The fp16 tests
A. Unique TensorFlower
2016-05-27
*
Merge changes from github.
Derek Murray
2016-05-23
*
Add the autotune version for the backward passes. It is currently
Xiaoqiang Zheng
2016-05-11
*
Adding autotune to the Cudnn conv algorithm selection.
Xiaoqiang Zheng
2016-05-09
*
Add support for 3d convolutions and pooling. CPU kernels use Eigen, GPU kerne...
A. Unique TensorFlower
2016-05-04
*
Disable conv/matmul on eigen since it takes too long to compile on nvcc.
Vijay Vasudevan
2016-04-28
*
Merge changes from github.
Illia Polosukhin
2016-04-18
[next]