1. 10 2月, 2018 2 次提交
  2. 26 12月, 2017 1 次提交
  3. 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
  4. 11 10月, 2017 1 次提交
    • M
      Conv Shift Operator (#4591) · a281b383
      Markus Kliegl 提交于
      * conv_shift_op: initial implementation using Eigen
      
      Limitations:
      - both gradient outputs must be specified and are always computed
      - explicit for loops => could be optimized in various ways
        (e.g., different memory layout)
      
      * conv shift - gradient fixes
      
      fix case when not all output gradients desired
      
      * conv shift: minor cleanup
      
      * conv shift - more minor cleanup
      
      * conv shift: clean up & initial GPU implementation
      
      * fix rebase issue
      a281b383
  5. 09 8月, 2017 1 次提交
  6. 01 8月, 2017 1 次提交
  7. 26 7月, 2017 2 次提交
  8. 25 7月, 2017 1 次提交
  9. 17 7月, 2017 2 次提交
  10. 11 7月, 2017 2 次提交
  11. 06 7月, 2017 2 次提交
  12. 05 7月, 2017 1 次提交
  13. 04 7月, 2017 4 次提交
  14. 03 7月, 2017 1 次提交
  15. 28 6月, 2017 2 次提交