1. 08 11月, 2019 1 次提交
    • L
      fix bug in pool/conv/conv_transpose: UpdatePaddingAndDilation,... · 26a6e27a
      liym27 提交于
      fix bug in pool/conv/conv_transpose: UpdatePaddingAndDilation, _get_padding_with_SAME and conv2dtranspose_forward_naive. (#20997)
      
      * fix bug in pool/conv/conv_transpose:
      1. It should be stride[i] not stride[0] in UpdatePaddingAndDilation;
      2. fix bug of func  _get_padding_with_SAME in test_conv/conv_transpose_op.py;
      3. fix bug of the computation process in function conv2dtranspose_forward_naive.
      test=develop
      
      * change test to make the data of different dimensions different. test=develop
      26a6e27a
  2. 31 10月, 2019 1 次提交
  3. 13 10月, 2019 1 次提交
  4. 07 10月, 2019 1 次提交
  5. 29 9月, 2018 1 次提交
  6. 27 6月, 2018 1 次提交
  7. 22 5月, 2018 1 次提交
  8. 08 5月, 2018 2 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
    • C
      fix MatMul parameter · 187e23a7
      chengduoZH 提交于
      187e23a7
  9. 04 5月, 2018 1 次提交
  10. 03 5月, 2018 1 次提交
  11. 11 4月, 2018 1 次提交
  12. 12 2月, 2018 1 次提交
  13. 10 2月, 2018 2 次提交
  14. 23 1月, 2018 1 次提交
  15. 15 1月, 2018 2 次提交
  16. 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
  17. 20 12月, 2017 1 次提交
  18. 14 12月, 2017 1 次提交
  19. 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
  20. 11 12月, 2017 1 次提交
  21. 08 12月, 2017 1 次提交
  22. 27 11月, 2017 1 次提交
  23. 22 11月, 2017 1 次提交
  24. 21 11月, 2017 1 次提交
  25. 17 11月, 2017 1 次提交
  26. 15 11月, 2017 1 次提交
  27. 10 11月, 2017 1 次提交
  28. 08 11月, 2017 1 次提交
  29. 06 11月, 2017 2 次提交
  30. 05 11月, 2017 1 次提交
  31. 27 10月, 2017 1 次提交
  32. 26 10月, 2017 1 次提交
  33. 24 10月, 2017 1 次提交
  34. 21 10月, 2017 2 次提交
  35. 20 10月, 2017 1 次提交