1. 18 11月, 2019 10 次提交
    • J
      Better TensorRT support (#20858) · 330b173c
      Jeng Bai-Cheng 提交于
      * Fix TensorRT detection bug
      
      1. Add new search path for TensorRT at tensorrt.cmake
      2. Add better debug message
      3. Fix the bug of detection of TensorRT version
      
      In NVIDIA official docker image, TensorRT headers are located at
      `/usr/include/x86_64-linux-gnu` and TensorRT libraries are located
      at `/usr/lib/x86_64-linux-gnu`, so using `-DTENSORRT_ROOT` will
      fail to detect TensorRT.
      
      There is no debug/warning message to tell developer that TensorRT
      is failed to be detected.
      
      In later version of TensorRT (e.g. v6), `NV_TENSORRT_MAJOR` is
      defined at `NvInferVersion.h` instead of `NvInfer.h`, so add
      compatibility fix.
      
      * Fix TensorRT variables in CMake
      
      1. Replace `${TENSORRT_ROOT}/include` with `${TENSORRT_INCLUDE_DIR}`
      2. Replace `${TENSORRT_ROOT}/lib` with `${TENSORRT_LIBRARY}`
      
      Manually type path may locate incorrect path of TensorRT. Use the
      paths detected by system instead.
      
      * Fix TensorRT library path
      
      1. Add new variable - `${TENSORRT_LIBRARY_DIR}`
      2. Fix TensorRT library path
      
      inference_lib.cmake and setup.py.in need the path of TensorRT library
      instead of the file of TensorRT library, so add new variable to fix it.
      
      * Add more general search rule for TensoRT
      
      Let system detect architecture instead of manually assign it, so
      replace `x86_64-linux-gnu` with `${CMAKE_LIBRARY_ARCHITECTURE}`.
      
      * Add more general search rule for TensorRT
      
      Remove duplicate search rules for TensorRT libraries. Use
      `${TENSORRT_LIBRARY_DIR}` to get full path of libnvinfer.so
      
      test=develop
      330b173c
    • L
      fix sporadically hang issue on windows(#21201) · d8b6cf2b
      liuwei1031 提交于
      cudaStreamSynchronize randomly hang when used in multi-thread environment, replace it with cudaStreamQuery API on windows
      d8b6cf2b
    • D
      3fe63d67
    • Z
      modified error message and API doc for channel_last supported Op (#21002) · 9cbe7bcc
      Zhang Ting 提交于
      * modified error message for conv and conv_transpose, test=develop
      
      * modified doc of conv and conv_transpose op, test=develop
      
      * modified the expression for error message, test=develop
      
      * modified error message for group_norm op, test=develop
      
      * modified detail of Attr(data_format) or Attr(data_layout)
      
      * add ValueError in API doc for maxout op, test=develop
      9cbe7bcc
    • L
      Control flow API: switch_case (#21103) · 92475282
      liym27 提交于
      * add API switch_case. test=develop
      
      add Nest
      
      * modify code according to reviews:
      1.Attr(branch_index) support 'uint8' and 'int64' besides 'int32'.
      2.remove useless code.
      test=develop
      
      * replace fluid.layers.data with fluid.data and polish API document. test=develop
      92475282
    • Z
      TRT int8: refine trt int8 for dynamic range set (#21112) · 65f70525
      Zhaolong Xing 提交于
      * refine trt int8 for dynamic range set
      test=develop
      
      * refine trt int8
      test=develop
      65f70525
    • G
    • W
      Fix INF bug of softmax_cross_entropy_op (#21165) · 3c98ec90
      WangXi 提交于
      3c98ec90
    • Z
      fix dygraph trace bug, test=develop (#21193) · 0f30d3a2
      Zeng Jinle 提交于
      0f30d3a2
    • C
      Add CI check for error message writing specification (#21107) · 7269ffe3
      Chen Weihang 提交于
      * add ci check for error message specification, test=develop, test=document_fix
      
      * replace spec url & refine failed message, test=develop, test=document_fix
      7269ffe3
  2. 16 11月, 2019 1 次提交
  3. 15 11月, 2019 5 次提交
  4. 14 11月, 2019 12 次提交
  5. 13 11月, 2019 4 次提交
  6. 12 11月, 2019 8 次提交