1. 14 7月, 2021 1 次提交
  2. 13 7月, 2021 1 次提交
    • S
      [model weights] zero_to_fp32 multiple improvements (#1181) · 2a921069
      Stas Bekman 提交于
      * add live zero checkpoint to fp32 consolidation version
      
      * some more docs
      
      * zero2 model states uses a different filename
      
      * fix
      
      * make debug mode cli configurable
      
      * copy the script only on node 0 process 0
      
      * validate that we have the right number of files
      
      * revamp _get_zero_param_shapes, instrument with easier debug
      
      * correct assertion
      
      * rename API; add even simpler API
      
      * style
      
      * docs improve
      
      * update the docs
      
      * revert the unpartitioned_params detection and report as it's most likely persistent params
      Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com>
      2a921069
  3. 08 7月, 2021 1 次提交
  4. 02 7月, 2021 1 次提交
  5. 24 6月, 2021 1 次提交
  6. 17 6月, 2021 2 次提交
  7. 15 6月, 2021 1 次提交
  8. 09 6月, 2021 3 次提交
  9. 28 5月, 2021 1 次提交
  10. 25 5月, 2021 5 次提交
  11. 24 5月, 2021 1 次提交
  12. 20 5月, 2021 2 次提交
  13. 14 5月, 2021 1 次提交
    • O
      [docs] unused parameter handling (#1060) · 63c5070e
      Olatunji Ruwase 提交于
      * Fix docstring
      
      * Make screenshots clickable for easier viewing
      
      * Navigation menu in alphabetical order; More clicable screenshots
      
      * Rename 1Cycle doc
      
      * Tweak naming
      
      * Remove no longer used flag
      
      * ZeRO3 Offload release
      
      * Single GPU results
      
      * Rearrange figures
      
      * Single GPU text
      
      * tweak intro
      
      * zero3-offload section
      
      * Add asynchronous i/o docs
      63c5070e
  14. 13 5月, 2021 2 次提交
  15. 01 5月, 2021 1 次提交
  16. 28 4月, 2021 1 次提交
  17. 27 4月, 2021 1 次提交
  18. 25 4月, 2021 1 次提交
    • H
      Add find_unused_parameters option to DeepSpeedEngine (#945) · d0b61f18
      hamlet 提交于
      * Add find_unused_parameters option
      
      As unused parameters in modules may not be expected sometimes, 
      add an explicit error msg when it occurred and an option to avoid the error: https://github.com/microsoft/DeepSpeed/issues/707
      
      * Add find_unused_parameters option
      
      As unused parameters in modules may not be expected sometimes, 
      add an explicit error msg when it occurred and an option to avoid the error: https://github.com/microsoft/DeepSpeed/issues/707
      
      * Fix syntax error
      
      * Fix yapf error
      
      * Fix yapf error
      
      * Fix yapf error
      
      * Fix yapf error
      
      * Move stage2 find_unused_parameters to config file
      
      * Add stage2 find_unused_parameters
      
      * Add stage2 find_unused_parameters
      
      * Add stage2_find_unused_parameters option
      
      * Change error msg to reflect zero_optimization config change
      
      * Fix yapf error
      
      * Fix yapf errors
      
      * Change find_unused_parameters option name
      
      * Change find_unused_parameters option name
      
      * Change find_unused_parameters option name
      
      * Change find_unused_parameters option name
      
      * Change find_unused_parameters option name
      
      * Add UnusedParametersModel for test option find_unused_parameters
      
      * Add unit test for stage2 find_unused_parameters
      
      * Add cpu-adam compatible check
      
      * Remove dups import
      
      * Trim spaces
      
      * Fix yapf errors
      
      * Trim spaces
      
      * Add False Positive test check
      
      * Fix find_unused_parameters test
      
      * Trim spaces
      
      * Fix yapf error
      d0b61f18
  19. 23 4月, 2021 2 次提交
    • O
      Asynchronous I/O docs (#1000) · bff4bc72
      Olatunji Ruwase 提交于
      * Fix docstring
      
      * Make screenshots clickable for easier viewing
      
      * Navigation menu in alphabetical order; More clicable screenshots
      
      * Rename 1Cycle doc
      
      * Tweak naming
      
      * Remove no longer used flag
      
      * ZeRO3 Offload release
      
      * Single GPU results
      
      * Rearrange figures
      
      * Single GPU text
      
      * tweak intro
      
      * zero3-offload section
      
      * Add asynchronous i/o docs
      bff4bc72
    • S
      [doc] add missing pin_memory entry (#999) · ecf2e1bc
      Stas Bekman 提交于
      - `offload_param` was missing `pin_memory` 
      - also moved the entry in `offload_optimizer` to have it in the same place.
      ecf2e1bc
  20. 21 4月, 2021 2 次提交
  21. 20 4月, 2021 1 次提交
  22. 19 4月, 2021 1 次提交
  23. 20 4月, 2021 1 次提交
  24. 19 4月, 2021 4 次提交
  25. 15 4月, 2021 2 次提交