onnxruntime
TensorRT EP - timing cache
#14767
Merged

Commits
  • add timing cache
    chilo-ms committed 4 years ago
  • enable timing cache for test
    chilo-ms committed 4 years ago
  • Make it only on Linux
    chilo-ms committed 4 years ago
  • undo last commit
    chilo-ms committed 4 years ago
  • Merge branch 'master' into trt_timing_cache
    chilo-ms committed 4 years ago
  • add 'timing_cache_enable' tensorrt provider options
    chilo-ms committed 4 years ago
  • fix bug
    chilo-ms committed 4 years ago
  • fix bug
    chilo-ms committed 4 years ago
  • revert modification
    chilo-ms committed 4 years ago
  • small modification
    chilo-ms committed 4 years ago
  • remove intrumentation code for recording engine build latency
    chilo-ms committed 4 years ago
  • add timing_cache_enable as additional member of internal TensorRT provider options struct
    chilo-ms committed 4 years ago
  • fix warning
    chilo-ms committed 4 years ago
  • enable trt timing cache for model tests
    chilo-ms committed 4 years ago
  • Merge branch 'master' into trt_timing_cache
    chilo-ms committed 4 years ago
  • enable timing cache for model tests
    chilo-ms committed 4 years ago
  • change pool
    chilo-ms committed 4 years ago
  • change back previous pool
    chilo-ms committed 4 years ago
  • change path of trt_timing_cache
    chilo-ms committed 4 years ago
  • refactor code
    chilo-ms committed 4 years ago
  • Add test cases for timing cache
    chilo-ms committed 4 years ago
  • fix bug
    chilo-ms committed 4 years ago
  • fix bug for CI
    chilo-ms committed 4 years ago
  • Merge branch 'master' into trt_timing_cache
    chilo-ms committed 4 years ago
  • fix bug
    chilo-ms committed 4 years ago
  • fix bug
    chilo-ms committed 4 years ago
  • Merge branch 'trt_timing_cache'
    gedoensmax committed 3 years ago
  • timing cache test
    gedoensmax committed 3 years ago
  • append compute capability to cache and add force option
    gedoensmax committed 3 years ago
  • Take timing of first warm up inference
    gedoensmax committed 3 years ago
  • Merge branch 'main' into tensorrt_timing_cache
    gedoensmax committed 3 years ago
  • detailed build log option
    gedoensmax committed 3 years ago
  • format changes and adding force timing cache to provider options
    gedoensmax committed 3 years ago
  • fix pybind OrtTensorRTProviderOptionsV2
    gedoensmax committed 3 years ago
  • reset OrtTensorRTProviderOptions
    gedoensmax committed 3 years ago
Loading