1. 04 12月, 2018 1 次提交
    • Z
      test=develop · deb04809
      ZongwuYang 提交于
      Fix the bug that profiler cannot trace the nccl allreduce operator
      deb04809
  2. 26 11月, 2018 1 次提交
  3. 08 11月, 2018 1 次提交
  4. 13 8月, 2018 1 次提交
  5. 10 8月, 2018 1 次提交
  6. 31 7月, 2018 1 次提交
  7. 30 7月, 2018 3 次提交
  8. 23 7月, 2018 1 次提交
  9. 14 6月, 2018 1 次提交
    • X
      Remove cuptiFinalize. · d2afd210
      Xin Pan 提交于
      In cupti samples, only cuptiFlush is used.
      I can't find any places calling cuptiFinalize and
      this API can error out as not_implemented in some
      cuda installation.
      d2afd210
  10. 08 6月, 2018 2 次提交
  11. 22 5月, 2018 1 次提交
    • X
      multi-thread handlerequest · b4dd4c04
      Xin Pan 提交于
          Experiment on vgg flower, 2 trainers, 1ps.
          more trainer could have more speedup.
      
          After:
          Pass = 0, Iters = 327, Speed = (7.52) img/s
          Before:
          Pass = 0, Iters = 385, Speed = (6.77) img/s
      b4dd4c04
  12. 10 4月, 2018 1 次提交
  13. 14 3月, 2018 1 次提交
  14. 08 3月, 2018 2 次提交
  15. 06 3月, 2018 2 次提交
  16. 02 3月, 2018 1 次提交
  17. 01 3月, 2018 2 次提交
  18. 28 2月, 2018 1 次提交
  19. 26 2月, 2018 2 次提交