1. 22 12月, 2021 1 次提交
  2. 20 12月, 2021 1 次提交
  3. 14 12月, 2021 3 次提交
  4. 07 12月, 2021 1 次提交
  5. 30 11月, 2021 3 次提交
  6. 29 11月, 2021 1 次提交
  7. 24 11月, 2021 1 次提交
  8. 22 11月, 2021 1 次提交
  9. 17 11月, 2021 2 次提交
  10. 16 11月, 2021 2 次提交
  11. 11 11月, 2021 1 次提交
    • J
      Added softplus + activation oneDNN fuse pass (#36657) · a346c4dc
      jakpiase 提交于
      * added softplus + activation fuse plass
      
      * minor change
      
      * implemented reviewer suggestion
      
      * minor fix
      
      * minor fix
      
      * added scale_out parameter
      
      * minor fix
      
      * fix for iScan CI
      
      * conditionally disabled logs
      
      * refactored pass builder
      a346c4dc
  12. 10 11月, 2021 1 次提交
  13. 05 11月, 2021 2 次提交
    • J
      Added caching of scales for bias in conv2d int8 (#36980) · 3705b12c
      Jacek Czaja 提交于
      * - Cached bias scales
      
      * - Fix
      
      * - fixes after review
      
      * - second round of fixes after internal review
      3705b12c
    • J
      Disable pool&conv_transpose&quantize caching (#36695) · db6c00c4
      Jacek Czaja 提交于
      * - WIP
      
      - compilation fix
      
      - fix
      
      - fixes
      
      - fix
      
      - fix
      
      - fix again
      
      - fix
      
      - another fix
      
      - another compilation fix
      
      - fix
      
      - fix
      
      - fix
      
      - lint
      
      * - pool2d partially stripped from cache
      
      - pool2d partially stripped of caching
      
      * - compilation fix
      
      * - compilation fix
      
      * - Fix to UT of caching
      
      * - Enabling test_conv3d_mkldnn
      
      * - conv_transpose stripped of cache
      
      * - compilation fix
      
      * - fix
      
      * - fix
      
      * - compilation fix
      
      * - fix
      
      * Reverted disabling caching of conv2d
      
      * - compilation fix
      
      * - ut reverted
      db6c00c4
  14. 02 11月, 2021 2 次提交
  15. 27 10月, 2021 1 次提交
  16. 23 10月, 2021 1 次提交
  17. 21 10月, 2021 1 次提交
    • J
      Added matmul_v2+transpose+reshape fuse pass (#36481) · 856cb9c5
      jakpiase 提交于
      * added base changes for matmul_v2+trans+resh fuse pass
      
      * added full matmul_v2+transpose+reshape pass
      
      * removed a file added by mistake
      
      * added reviewers suggestions
      
      * Changed ops type in checking capatibility version
      
      * Deteled one statement
      856cb9c5
  18. 18 10月, 2021 1 次提交
    • J
      Added softplus FP32 FWD OneDNN kernel (#36382) · bdac9ff6
      jakpiase 提交于
      * added softplus
      
      * refactored softplus op
      
      * deleted unnecessary file
      
      * added missing file
      
      * added formatting
      
      * disabled tests if GPU is used
      
      * added reviewer suggestion
      
      * unified softplus kernel
      bdac9ff6
  19. 14 10月, 2021 1 次提交
  20. 13 10月, 2021 1 次提交
  21. 11 10月, 2021 1 次提交
  22. 08 10月, 2021 2 次提交
  23. 07 10月, 2021 1 次提交
  24. 05 10月, 2021 1 次提交
    • J
      Added concat BF16/FP32 BWD OneDNN kernel (#35889) · dc4d5719
      jakpiase 提交于
      * tmp
      
      * added concat BF16/FP32 BWD oneDNN kernel
      
      * minor change
      
      * minor change
      
      * fix for CI
      
      * added formatting
      
      * Reverted deleting static keyword
      
      * added reviewers suggestions
      
      * reverted deleting concat bf16 test file
      
      * fixed concat tests
      dc4d5719
  25. 27 9月, 2021 1 次提交
  26. 24 9月, 2021 1 次提交
  27. 21 9月, 2021 1 次提交
    • A
      Reuse OneDNN handler for SGD and SUM for SelectedRows input tensors. (#35510) · 799f3861
      Adam Osewski 提交于
      * Create stateful OneDNNAXPYHandler object.
      
      This makes it possible to call it multiple times without recreating the
      oneDNN primitives every time.
      
      * Prepare SGDOpKernel to reuse its implementation from OneDNN kernel.
      
      * OneDNN SGD kernel.
      
      * Update call to use new OneDNNAXPYHandler object api.
      
      * Setup seed in proper place.
      
      * Enable OneDNN kernel only for single case.
      
      * For dense param and sparse grad.
      
      * Small refactor.
      
      * Enable oneDNN by op attr or by cmd line flag.
      
      * Use int64_t type for number of elements.
      
      * Support dense param and grad from OneDNN kernel.
      
      * Enable SGD OneDNN kernel when use MP BF16 optimizer.
      
      * Force non-copyable/movable OneDNNAXPYHandler.
      
      * Reuse OneDNNAXPYHandler for spare tensors in SUM op.
      
      * Fix SFINAE rules.
      
      * Remove recording event inside AXPY.
      
      * Get rid of internal primitive caching.
      
      * Stop use PP cache mechanims to store mem and primitive obj.
      * Handler obj store and reuse needed desc & prim
      
      * Do not derive from MKLDNNHandlerT
      799f3861
  28. 18 9月, 2021 1 次提交
    • J
      [oneDNN] Disable caching of Reorder operation (#35664) · e4c2a854
      Jacek Czaja 提交于
      * - REorder disabling caching
      
      * - compilation fix
      
      * - another compilation fix
      
      * - another compilation fix
      
      * - compilation fix
      
      * - Fix
      
      * - yet another compilation fix
      
      * - suppresingly another compilation fix
      
      * - lint
      
      * - fix after review
      
      * - fix
      e4c2a854
  29. 15 9月, 2021 1 次提交
  30. 13 9月, 2021 1 次提交
  31. 10 9月, 2021 1 次提交