1. 07 10月, 2019 1 次提交
  2. 15 8月, 2019 1 次提交
  3. 23 4月, 2019 1 次提交
  4. 20 3月, 2019 1 次提交
  5. 06 3月, 2019 1 次提交
  6. 05 3月, 2019 1 次提交
  7. 12 12月, 2018 1 次提交
  8. 27 11月, 2018 2 次提交
    • J
      - Fix to test_conv2d_transpose_mkldnn for GPU · 1540df51
      Jacek Czaja 提交于
      test=develop
      1540df51
    • J
      - conv2d transpose MKL-DNN · fb24690a
      Jacek Czaja 提交于
      test=develop
      
      - Added new header for MKLDNN reuse functionality
      
      - Extended conv2d_transpose GetExpectedKernelType for MKL-DNN supporrt
      
      - Buildable conv transpose mkldnn and conv mkldnn using conv template
      
      - Conv2d transpose roughlt implemented and buildable
      
      - Added modifications conv2d transpose MKLDNN unit tests
      
      - Fix to UT of conv2d transpose mkldnn op
      
      - Wrong type of MKLDNN primitive was chosen for conv2d transpose
      
      - HAcks for conv2d transpose
      
      - UT enalbed
      
      - Replaced copying loop with memcpy
      
      - Draft of passing lambda into AcquireMemory
      
      - Made reorder (IOHW->OIHW) to be called only once
      fb24690a
  9. 19 9月, 2018 1 次提交
  10. 27 6月, 2018 1 次提交
  11. 13 6月, 2018 1 次提交
    • Y
      update · c58ba827
      yi.wu 提交于
      c58ba827
  12. 23 5月, 2018 1 次提交
  13. 22 5月, 2018 1 次提交
  14. 08 5月, 2018 1 次提交
    • Y
      Clean OpProtoAndCheckerMaker · 0e78cb69
      Yu Yang 提交于
      Do not use ctor
      
      * Reduce line of codes.
      * We can use virtual function for Maker now.
      * The implementation does not care what maker holds, it is easier to
      refactor later.
      0e78cb69
  15. 19 4月, 2018 1 次提交
  16. 17 4月, 2018 1 次提交
  17. 11 4月, 2018 1 次提交
  18. 12 2月, 2018 1 次提交
  19. 10 2月, 2018 2 次提交
  20. 23 1月, 2018 1 次提交
  21. 17 1月, 2018 3 次提交
  22. 15 1月, 2018 2 次提交
  23. 14 1月, 2018 1 次提交
    • D
      "cudnn operators change to cudnn kernel" (#6660) · 5ad1aef0
      dzhwinter 提交于
      * "unified operators"
      
      * "add CUDNN register"
      
      * "add use cudnn attribute"
      
      * "add attribute"
      
      * "test conv tranpose op"
      
      * "remove duplicated attr"
      
      * "fix op test"
      
      * "add attribute to set cudnn"
      
      * "add more log"
      
      * "need layout op register support"
      
      * "add more log"
      
      * "change GetExpectedKernelType "
      
      * "fix Get attr in conv_op"
      
      * "fix CI"
      
      * "fix tests"
      
      * "removed kernel priority fallback"
      
      * "fix CI"
      
      * "fix stack pointer bug"
      
      * "refine buggy interface"
      
      * "add const cast to save life"
      
      * "fix get_output_with_grad"
      
      * "fix op test with dataformat"
      
      * ""fix pooling
      
      * "fix pooling test"
      
      * "fix CI"
      
      * "fix with_gpu error"
      
      * "add transform needed functional check"
      
      * "fix unpack list error"
      
      * "comment out parallel.do temporary"
      
      * "fix CI"
      
      * "fix compile doc error"
      
      * "make threshold larger"
      5ad1aef0
  24. 26 12月, 2017 1 次提交
  25. 20 12月, 2017 1 次提交
  26. 12 12月, 2017 1 次提交
    • Q
      Refine device context (#6433) · 61ec0b95
      QI JUN 提交于
      There are mainly following fixes:
      
      - take `DeviceContext` as the template parameter of math functors and OpKernel instead of `Place`
      - remove `eigen_device` interface in base class  `DeviceContext`
      - remove `GetEigenDevice` interface in `ExecutionContext` and base class `DeviceContext`
      - remove unused `platform::EigenDeviceConverter`
      - rename `REGISTER_OP_GPU_KERNEL` to `REGISTER_OP_CUDA_KERNEL`
      - rename `USE_GPU_ONLY_OP` to `USE_CUDA_ONLY_OP`
      61ec0b95
  27. 08 12月, 2017 1 次提交
  28. 05 12月, 2017 1 次提交
  29. 28 11月, 2017 1 次提交
  30. 27 11月, 2017 1 次提交
  31. 17 11月, 2017 2 次提交
  32. 15 11月, 2017 1 次提交
  33. 06 11月, 2017 2 次提交