1. 12 11月, 2019 1 次提交
    • L
      fix the computation for dx (grad for x) for prelu operation. (#20949) · e249d9a3
      lilong12 提交于
      * set the default value of alpha for prelu to 0.25, test=develop
      
      * add the call to __syncthreads(), test=develop
      
      * fix the implementation of cpu prelu, test=develop
      
      * repair the implementation of element mode prelu, test=develop
      
      * modify test_prelu_op.py, test=develop
      e249d9a3
  2. 05 9月, 2019 1 次提交
  3. 17 7月, 2019 1 次提交
    • Y
      Add cuda implementation for `prelu` backward pass (#18633) · ce1ec332
      Yang Zhang 提交于
      * Add GPU implementation for `prelu` backward pass
      
      test=develop
      
      * Fix logic error in `prelu` GPU backward and simplify a bit
      
      test=develop
      
      * Fix `prelu` backward CUDA implementation
      
      test=develop
      
      CPU version was not used actually, so test passed
      ce1ec332
  4. 03 12月, 2018 1 次提交
  5. 15 8月, 2018 1 次提交
  6. 12 2月, 2018 1 次提交
  7. 10 2月, 2018 2 次提交
  8. 26 12月, 2017 1 次提交
  9. 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
  10. 16 9月, 2017 1 次提交
  11. 22 8月, 2017 1 次提交
  12. 21 8月, 2017 2 次提交
  13. 25 7月, 2017 1 次提交
  14. 17 7月, 2017 2 次提交
  15. 11 7月, 2017 2 次提交
  16. 06 7月, 2017 2 次提交
  17. 05 7月, 2017 1 次提交
  18. 04 7月, 2017 4 次提交
  19. 03 7月, 2017 1 次提交
  20. 28 6月, 2017 2 次提交