1. 18 2月, 2022 3 次提交
    • Z
      bug fix (#39630) · bbf31a4e
      zhaoyingli 提交于
      bbf31a4e
    • Z
      [AMP] support GPU BF16 amp for dygraph (#39029) · 7d6d3848
      zhangbo9674 提交于
      * support dtype param for auto_cast
      
      * add amp_dtype for tracer
      
      * add unsupported bf16 list
      
      * support bf16 amp for O2
      
      * refine python interface for bfloat16
      
      * refine code
      
      * refine code
      
      * refine unittest
      
      * refine code
      
      * refine code
      
      * add bf16 o1
      
      * refine code by comment
      
      * add gradient accumulator
      
      * add recompute
      7d6d3848
    • B
      Fix sharding group (#39668) · bc3ca678
      Baibaifan 提交于
      * fix_sharding_group
      
      * fix_sharding_group
      bc3ca678
  2. 17 2月, 2022 1 次提交
  3. 16 2月, 2022 1 次提交
    • Z
      sync/geo test ok & fix heter_worker program ok (#39511) · b2986bab
      ziyoujiyi 提交于
      * delete gloo connect retry
      
      * the_one_ps dirs reconstruct
      
      * .
      
      * .
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * refactor ps optimize
      
      * refactor ps optimize
      
      * refactor ps optimize
      
      * .
      
      * .
      
      * .
      
      * .
      
      * .
      
      * .
      
      * refactor theoneps
      
      * the_one_ps
      
      * add ps pass unittest
      
      * add ps pass unittest
      
      * ps unitest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * add cpu_async_ps_mode test
      
      * add cpu_async_ps_mode test
      
      * add cpu_async_ps_mode test
      
      * ps unittest ready
      
      * ps unittest ready
      
      * solve dist_pass init conflict
      
      * solve import CommContext error
      
      * unittest ok
      
      * implement AllocateFrom
      
      * solve setup.py.in conflict
      
      * solve conflict
      
      * solve conflict
      
      * solve conflict
      
      * .
      
      * .
      
      * cpu-async-ps minimize test ok & gpu minimize test ok
      
      * add heter 2stage unittest
      
      * add heter 2stage unittest
      
      * add heter 2stage unittest
      
      * sync/geo test ok & fix heter_worker program ok
      
      * .
      Co-authored-by: Nzkh2016 <zhangkaihuo@baidu.com>
      b2986bab
  4. 14 2月, 2022 2 次提交
    • H
      Add Inplace addto pass and unittest. (#39433) · 52af0a60
      hlygit66666 提交于
      * add fuse_relu_depthwise_conv_pass unittest
      
      * fix atol and rtol
      
      * fix according to review
      
      * Update test_dist_fuse_relu_depthwise_conv_pass.py
      
      * add inplace_addto pass and unittest
      52af0a60
    • Z
      统一ps:heter ps 二阶段单测通过 (#39468) · 765a2ada
      ziyoujiyi 提交于
      * delete gloo connect retry
      
      * the_one_ps dirs reconstruct
      
      * .
      
      * .
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * refactor ps optimize
      
      * refactor ps optimize
      
      * refactor ps optimize
      
      * .
      
      * .
      
      * .
      
      * .
      
      * .
      
      * .
      
      * refactor theoneps
      
      * the_one_ps
      
      * add ps pass unittest
      
      * add ps pass unittest
      
      * ps unitest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * add cpu_async_ps_mode test
      
      * add cpu_async_ps_mode test
      
      * add cpu_async_ps_mode test
      
      * ps unittest ready
      
      * ps unittest ready
      
      * solve dist_pass init conflict
      
      * solve import CommContext error
      
      * unittest ok
      
      * implement AllocateFrom
      
      * solve setup.py.in conflict
      
      * solve conflict
      
      * solve conflict
      
      * solve conflict
      
      * .
      
      * .
      
      * cpu-async-ps minimize test ok & gpu minimize test ok
      
      * add heter 2stage unittest
      
      * add heter 2stage unittest
      
      * add heter 2stage unittest
      Co-authored-by: Nzkh2016 <zhangkaihuo@baidu.com>
      765a2ada
  5. 11 2月, 2022 1 次提交
    • Z
      统一 ps 开发 - python (#39431) · 22c67d14
      ziyoujiyi 提交于
      * delete gloo connect retry
      
      * the_one_ps dirs reconstruct
      
      * .
      
      * .
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * refactor ps optimize
      
      * refactor ps optimize
      
      * refactor ps optimize
      
      * .
      
      * .
      
      * .
      
      * .
      
      * .
      
      * .
      
      * refactor theoneps
      
      * the_one_ps
      
      * add ps pass unittest
      
      * add ps pass unittest
      
      * ps unitest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * add cpu_async_ps_mode test
      
      * add cpu_async_ps_mode test
      
      * add cpu_async_ps_mode test
      
      * ps unittest ready
      
      * ps unittest ready
      
      * solve dist_pass init conflict
      
      * solve import CommContext error
      
      * unittest ok
      
      * implement AllocateFrom
      
      * solve setup.py.in conflict
      
      * solve conflict
      
      * solve conflict
      
      * solve conflict
      
      * .
      
      * .
      
      * cpu-async-ps minimize test ok & gpu minimize test ok
      Co-authored-by: Nzkh2016 <zhangkaihuo@baidu.com>
      22c67d14
  6. 09 2月, 2022 1 次提交
  7. 08 2月, 2022 3 次提交
    • H
      Add FuseOptimizerPass and test_dist_fuse_adam_pass unittest. (#39208) · ccdcfa2d
      hlygit66666 提交于
      * add fuse_relu_depthwise_conv_pass unittest
      
      * fix atol and rtol
      
      * fix according to review
      
      * Add FuseOptimizerPass and fuse_adam_pass unittest
      
      * add sgd and momentum unittest
      
      * add fuse_optimizer_pass
      
      * close amp
      
      * close amp
      
      * update
      
      * fix run on two cards
      
      * Update test_dist_fuse_adam_pass.py
      
      * Update test_dist_fuse_momentum_pass.py
      
      * Update test_dist_fuse_sgd_pass.py
      
      * Create test_dist_fuse_sgd_pass.py
      
      * Create test_dist_fuse_sgd_pass.py
      
      * Create test_dist_fuse_sgd_pass.py
      
      * Update test_dist_fuse_adam_pass.py
      
      * Update test_dist_fuse_momentum_pass.py
      
      * Update test_dist_fuse_sgd_pass.py
      ccdcfa2d
    • Z
      ps optimize refactor (#38982) · 196dbfc2
      ziyoujiyi 提交于
      * delete gloo connect retry
      
      * the_one_ps dirs reconstruct
      
      * .
      
      * .
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * refactor ps optimize
      
      * refactor ps optimize
      
      * refactor ps optimize
      
      * .
      
      * .
      
      * .
      
      * .
      
      * .
      
      * .
      
      * refactor theoneps
      
      * the_one_ps
      
      * add ps pass unittest
      
      * add ps pass unittest
      
      * ps unitest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * ps unittest frame
      
      * add cpu_async_ps_mode test
      
      * add cpu_async_ps_mode test
      
      * add cpu_async_ps_mode test
      
      * ps unittest ready
      
      * ps unittest ready
      
      * solve dist_pass init conflict
      
      * solve import CommContext error
      
      * unittest ok
      
      * implement AllocateFrom
      
      * solve setup.py.in conflict
      
      * solve conflict
      
      * solve conflict
      
      * solve conflict
      
      * .
      
      * .
      Co-authored-by: Nzkh2016 <zhangkaihuo@baidu.com>
      196dbfc2
    • B
      optimize sharding stage3 (#39334) · 23d559dd
      Baibaifan 提交于
      23d559dd
  8. 30 1月, 2022 1 次提交
  9. 29 1月, 2022 2 次提交
  10. 28 1月, 2022 2 次提交
    • F
      [PSLIB] Add Metrics Module, Support User-defined Add Metric (#38789) · 2e6be886
      Fan Zhang 提交于
      * [PSLIB] Add Metrics Module, Support User-defined Add Metric
      
      * [PSLIB] Modify According to CI
      
      * [PSLIB] Modify According to CI
      
      * [PSLIB] Modify According to CI
      
      * [PSLIB] Modify According to CI Coverage
      
      * [PSLIB] Modify According to CI
      
      * [PSLIB] Modify According to CI
      
      * [PSLIB] Modify According to CI
      
      * [PSLIB] Modify According to CI
      
      * [PSLIB] Modify According to CI
      
      * [PSLIB] Modify According to CI Coverage
      
      * [PSLIB] Modify According to CI Coverage
      
      * [PSLIB] Modify According to CI Coverage
      
      * modify role_maker
      
      * update CMakeLists.txt
      2e6be886
    • B
      fix_stage2_minimize (#39285) · 90f44c6f
      Baibaifan 提交于
      90f44c6f
  11. 27 1月, 2022 2 次提交
  12. 26 1月, 2022 1 次提交
  13. 25 1月, 2022 5 次提交
  14. 24 1月, 2022 2 次提交
  15. 21 1月, 2022 1 次提交
    • Y
      [Auto Parallel] Use the new completion algorithm (#39086) · e5cda6fa
      Yulong Ao 提交于
      * Add the backward support for QR
      
      * Remove unnecessary comments
      
      * [Auto Parallel] Improve the dist op interface and compatible computation
      
      * Remove unnecessary modification
      
      * Recover some modifications
      
      * Add lost files
      
      * Fix a minor bug
      
      * Fix the bug of the planner
      
      * Fix the format problem
      
      * [Auto Parallel] Update the completion algorithm
      
      * Fix the bug of auto_searcher unittest
      e5cda6fa
  16. 20 1月, 2022 1 次提交
  17. 18 1月, 2022 2 次提交
  18. 17 1月, 2022 2 次提交
  19. 14 1月, 2022 1 次提交
  20. 13 1月, 2022 1 次提交
  21. 12 1月, 2022 2 次提交
    • Z
      the_one_ps dirs reconstruct (#38804) · 50609214
      ziyoujiyi 提交于
      * delete gloo connect retry
      
      * the_one_ps dirs reconstruct
      
      * .
      
      * .
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * create the_one_ps dirs
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      
      * the one ps dirs modify
      50609214
    • J
      [Dist Pass] Amp Pass (#38764) · cc24427e
      JZ-LIANG 提交于
      * auto parallel sharding base
      
      * chmod
      
      * add unitest
      
      * set unitest cmake dist label
      
      * revise code according to rewiew
      
      * chmod
      
      * bugfix for grad_clip and param broadcast
      
      * chmod
      
      * update unitest
      
      * chmod
      
      * add clip
      
      * chmod
      
      * add amp pass
      
      * chmod
      
      * add unitest
      
      * remove grad update
      
      * fixed bug
      
      * fixed bug
      
      * fixed typose
      
      * fixed typoes
      cc24427e
  22. 11 1月, 2022 1 次提交
  23. 06 1月, 2022 2 次提交