1. 11 2月, 2022 1 次提交
  2. 13 1月, 2022 1 次提交
    • J
      Added mul BF16/FP32 FWD/BWD oneDNN kernel (#38552) · fc6eed5b
      jakpiase 提交于
      * base changes for mul reimplementation
      
      * empty commit
      
      * tmp save
      
      * full implementation of mul bf16/fp32 fwd bwd
      
      * CI fix
      
      * CI rerun
      
      * changed unity build cmake to avoid gpu issues
      
      * removed mul mkldnn from unity build
      
      * added skipping tests if not cpu_bf16
      
      * CI fix
      
      * CI fix
      
      * CI fix
      fc6eed5b
  3. 09 7月, 2019 1 次提交
  4. 14 5月, 2019 1 次提交
    • K
      support fc_op double grad (#17317) · 60be66e2
      Kaipeng Deng 提交于
      * add double grad for mul_op. test=develop
      
      * fix format. test=develop
      
      * fix format. test=develop
      
      * fix format. test=develop
      
      * refine code. test=develop
      
      * remove setzero. test=develop
      
      * fix dx/dy init bug. test=develop
      
      * fix format. test=develop
      60be66e2
  5. 22 8月, 2018 1 次提交
  6. 04 5月, 2018 1 次提交
  7. 03 5月, 2018 1 次提交
  8. 15 3月, 2018 1 次提交
    • K
      Add fp16 mul op support and bind paddle fp16 to numpy fp16 (#9017) · e26f1123
      Kexin Zhao 提交于
      * add fp16 mul op support
      
      * small fix
      
      * fix bug
      
      * small fix
      
      * fix PADDLE_WITH_CUDA compiling issue
      
      * reorg code
      
      * test for pybind
      
      * treate as float16 as uint16_t in pybind
      
      * bind np.float16 to paddle float16
      
      * small fix
      
      * clean code
      
      * remove redundancy
      
      * fix mul_op test
      
      * address comments
      
      * small fix
      
      * add is_float16_supported func
      e26f1123
  9. 12 2月, 2018 1 次提交
  10. 10 2月, 2018 2 次提交
  11. 26 12月, 2017 1 次提交
  12. 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
  13. 09 11月, 2017 1 次提交
  14. 08 11月, 2017 1 次提交
  15. 24 10月, 2017 1 次提交
  16. 20 10月, 2017 1 次提交
  17. 28 9月, 2017 1 次提交
  18. 19 9月, 2017 1 次提交
  19. 07 9月, 2017 2 次提交
  20. 06 9月, 2017 1 次提交
  21. 05 9月, 2017 1 次提交
  22. 04 9月, 2017 1 次提交
  23. 19 8月, 2017 1 次提交
  24. 18 8月, 2017 1 次提交
  25. 14 8月, 2017 2 次提交
  26. 11 8月, 2017 1 次提交
  27. 10 8月, 2017 1 次提交
  28. 09 8月, 2017 1 次提交
  29. 08 8月, 2017 1 次提交
  30. 07 8月, 2017 2 次提交
  31. 05 8月, 2017 1 次提交
  32. 03 8月, 2017 2 次提交
  33. 02 8月, 2017 2 次提交
  34. 01 8月, 2017 1 次提交
    • Q
      use operator context and infer context (#3024) · 61ebacbc
      Qiao Longfei 提交于
      * use operator context
      
      * optimize code
      
      * update net infershape
      
      * update InferShape
      
      * disable override InferShape(scope) in OperatorBase
      
      * change InferShapeImpl to InferShape
      
      * add template to OperatorContext Input/Output
      
      * merge Input InputVar, Output OutputVar
      
      * change Inputs to MultiInput
      
      * fix conflict
      
      * fix MultiInput bugs and add unit test
      
      * rename KernelContext to ExecutionContext
      
      * clean code
      
      * change InferShape to protected
      
      * fix template bug
      
      * refine code
      
      * use InputVar instead of Input<Variable>
      
      * typo
      
      * optimize code
      61ebacbc