Commit history

1. Commits on Jun 26, 2018

   • Moving StatusOr from XLA to stream_executor.

     Show description for 8025ac3
     

     Michael Case

     committed
     8025ac3View commit details

     Copy full SHA for 8025ac3

     View code at this pointBrowse repository at this point
2. Commits on May 31, 2018

   • Handle FilterLayout::kOutputYXInput in FilterDescriptor::ToShortString.

     Show description for 18c67a4
     

     bixia1

     authored and

     tensorflower-gardener

     committed
     18c67a4View commit details

     Copy full SHA for 18c67a4

     View code at this pointBrowse repository at this point
3. Commits on May 30, 2018

   • Fix GPU build on windows

     Show description for 9489825
     

     smit-hinsu

     authored and

     tensorflower-gardener

     committed
     9489825View commit details

     Copy full SHA for 9489825

     View code at this pointBrowse repository at this point
4. Commits on May 25, 2018

   • Merge changes from github.

     Show description for b59833c
     

     yifeif

     authored and

     tensorflower-gardener

     committed
     b59833cView commit details

     Copy full SHA for b59833c

     View code at this pointBrowse repository at this point
5. Commits on May 23, 2018

   • Add convolution with NHWC layout to stream executor.

     Show description for 25ad31d
     tensorflower-gardener

     committed
     25ad31dView commit details

     Copy full SHA for 25ad31d

     View code at this pointBrowse repository at this point
6. Commits on May 22, 2018

   • Introduce an option to allocate CUDA unified memory

     Show description for b113981
     

     smit-hinsu

     authored and

     tensorflower-gardener

     committed
     b113981View commit details

     Copy full SHA for b113981

     View code at this pointBrowse repository at this point
7. Commits on May 20, 2018

   • Rollforward of CL 197167501, without enabling CUDNN_FFT_TILING_FORWARD because that breaks XLA tests.

     Show description for c76b815
     tensorflower-gardener

     committed
     c76b815View commit details

     Copy full SHA for c76b815

     View code at this pointBrowse repository at this point
8. Commits on May 18, 2018

   • Automated g4 rollback of changelist 197118212

     Show description for 4fce5d6
     tensorflower-gardener

     committed
     4fce5d6View commit details

     Copy full SHA for 4fce5d6

     View code at this pointBrowse repository at this point

   • Dropping support for CUDA < 8.

     Show description for 2934484
     tensorflower-gardener

     committed
     2934484View commit details

     Copy full SHA for 2934484

     View code at this pointBrowse repository at this point

   • Dropping support for cuDNN < 6.

     Show description for 4b052cc
     tensorflower-gardener

     committed
     4b052ccView commit details

     Copy full SHA for 4b052cc

     View code at this pointBrowse repository at this point
9. Commits on May 17, 2018

   • [TF:STREAM_EXECUTOR] move declaration of multi_platform_manager to header

     Show description for 2b03e28
     

     nickdesaulniers

     authored and

     tensorflower-gardener

     committed
     2b03e28View commit details

     Copy full SHA for 2b03e28

     View code at this pointBrowse repository at this point
10. Commits on May 15, 2018

    • Small polishing changes in stream executor, no functional changes.

      Show description for 14d8615
      tensorflower-gardener

      committed
      14d8615View commit details

      Copy full SHA for 14d8615

      View code at this pointBrowse repository at this point
11. Commits on May 14, 2018

    • avoid having stream_executor depend on tensorflow/core

      Show description for 157c347
      tensorflower-gardener

      committed
      157c347View commit details

      Copy full SHA for 157c347

      View code at this pointBrowse repository at this point
12. Commits on May 11, 2018

    • Remove 'using' of dnn types in CudnnSupport implementation file.

      Show description for 256c1d1
      tensorflower-gardener

      committed
      256c1d1View commit details

      Copy full SHA for 256c1d1

      View code at this pointBrowse repository at this point
13. Commits on May 9, 2018

    • Use parenthesis based construction instead of brace initialization

      Show description for 69bc455
      

      smit-hinsu

      authored and

      tensorflower-gardener

      committed
      69bc455View commit details

      Copy full SHA for 69bc455

      View code at this pointBrowse repository at this point

    • [XLA] Make XLA's memory allocator return an owning smart pointer.

      Show description for 7baa9ff
      

      Justin Lebar

      authored and

      tensorflower-gardener

      committed
      7baa9ffView commit details

      Copy full SHA for 7baa9ff

      View code at this pointBrowse repository at this point

    • Unifying argument documentation style in CudnnSupport.

      Show description for 4a42d16
      tensorflower-gardener

      committed
      4a42d16View commit details

      Copy full SHA for 4a42d16

      View code at this pointBrowse repository at this point
14. Commits on May 8, 2018

    • Better wrapping of stream executor's cuDNN API calls. Replacing mutex locking and setting the cuDNN stream followed by calling wrap::cudnn... with an RAII CudnnHandle object that handles the former…

      Show description for 8fcf647
      tensorflower-gardener

      committed
      8fcf647View commit details

      Copy full SHA for 8fcf647

      View code at this pointBrowse repository at this point
15. Commits on May 7, 2018

    • Remove unused threadpool from stream executor.

      Show description for e0be7b7
      tensorflower-gardener

      committed
      e0be7b7View commit details

      Copy full SHA for e0be7b7

      View code at this pointBrowse repository at this point
16. Commits on May 3, 2018

    • Replaced calls to tensorflow::StringPiece::ToString with std::string conversions.

      Show description for 223be4a
      tensorflower-gardener

      committed
      223be4aView commit details

      Copy full SHA for 223be4a

      View code at this pointBrowse repository at this point
17. Commits on May 2, 2018

    • Replaced calls to tensorflow::StringPiece::ToString with std::string conversions.

      Show description for 5e9e696
      

      hawkinsp

      authored and

      yifeif

      committed
      5e9e696View commit details

      Copy full SHA for 5e9e696

      View code at this pointBrowse repository at this point
18. Commits on Apr 30, 2018

    • Adding a depthwise convolution kernel op (with label 'cudnn_grouped_convolution') which forwards to cuDNN grouped convolutions.

      Show description for a5a51ad
      tensorflower-gardener

      committed
      a5a51adView commit details

      Copy full SHA for a5a51ad

      View code at this pointBrowse repository at this point
19. Commits on Apr 26, 2018

    • Cudnn RNN v2 kernels with autotune capability

      Show description for 270a6e9
      

      protoget

      authored and

      tensorflower-gardener

      committed
      270a6e9View commit details

      Copy full SHA for 270a6e9

      View code at this pointBrowse repository at this point
20. Commits on Apr 25, 2018

    • Stop using gpu:: as an alias for stream_executor::.

      Show description for 09a39ad
      

      Justin Lebar

      authored and

      tensorflower-gardener

      committed
      09a39adView commit details

      Copy full SHA for 09a39ad

      View code at this pointBrowse repository at this point
21. Commits on Apr 24, 2018

    • Add variants of DoBlasGemmWithAlgorithm with alpha being on device.

      Show description for 9f38ab7
      tensorflower-gardener

      committed
      9f38ab7View commit details

      Copy full SHA for 9f38ab7

      View code at this pointBrowse repository at this point

    • Merge changes from github.

      Show description for 22f3a97
      

      yifeif

      authored and

      tensorflower-gardener

      committed
      22f3a97View commit details

      Copy full SHA for 22f3a97

      View code at this pointBrowse repository at this point
22. Commits on Apr 22, 2018

    • [StreamExecutor] [XLA] Delete copy/pasted implementations of MakeUnique.

      Show description for ea0c8a7
      

      Justin Lebar

      authored and

      tensorflower-gardener

      committed
      ea0c8a7View commit details

      Copy full SHA for ea0c8a7

      View code at this pointBrowse repository at this point
23. Commits on Apr 20, 2018

    • Temporarily set cudnn Rnn math precision to fp32.

      Show description for 5fbb1fe
      

      protoget

      authored and

      tensorflower-gardener

      committed
      5fbb1feView commit details

      Copy full SHA for 5fbb1fe

      View code at this pointBrowse repository at this point
24. Commits on Apr 17, 2018

    • [StreamExecutor] Rename ::perftools::gputools -> ::stream_executor, part 1.

      Show description for 4764bf2
      

      Justin Lebar

      authored and

      tensorflower-gardener

      committed
      4764bf2View commit details

      Copy full SHA for 4764bf2

      View code at this pointBrowse repository at this point
25. Commits on Apr 13, 2018

    • Add more logging for failure cases in CUDATimer

      Show description for e42ebc5
      

      protoget

      authored and

      tensorflower-gardener

      committed
      e42ebc5View commit details

      Copy full SHA for e42ebc5

      View code at this pointBrowse repository at this point

    • Add debugging checks for setting cuda stream, so it will check fail if the

      Show description for cd7ba43
      

      aaroey

      authored and

      tensorflower-gardener

      committed
      cd7ba43View commit details

      Copy full SHA for cd7ba43

      View code at this pointBrowse repository at this point

    • Print error msg in CUDATimer.Init() when CreateEvent() is not ok().

      Show description for 457e8b3
      

      protoget

      authored and

      tensorflower-gardener

      committed
      457e8b3View commit details

      Copy full SHA for 457e8b3

      View code at this pointBrowse repository at this point
26. Commits on Apr 11, 2018

    • Internal change.

      Show description for c5d59c6
      

      annarev

      authored and

      tensorflower-gardener

      committed
      c5d59c6View commit details

      Copy full SHA for c5d59c6

      View code at this pointBrowse repository at this point

    • [StreamExecutor] Merge StreamExecutor's and XLA's StatusOr classes.

      Show description for e7b1ab0
      

      Justin Lebar

      authored and

      tensorflower-gardener

      committed
      e7b1ab0View commit details

      Copy full SHA for e7b1ab0

      View code at this pointBrowse repository at this point
27. Commits on Apr 7, 2018

    • Revamp a few ScratchAllocator classes in cudnn_rnn_ops

      Show description for 22cb902
      

      protoget

      authored and

      tensorflower-gardener

      committed
      22cb902View commit details

      Copy full SHA for 22cb902

      View code at this pointBrowse repository at this point