Skip to content

GitLab

Switch branch/tag
  1. 30 Mar, 2023 1 commit
  3. 19 Nov, 2021 1 commit
  5. 15 Nov, 2021 1 commit
  7. 10 Sep, 2021 2 commits
  9. 23 Jun, 2021 4 commits
    • Reed Wanderman-Milne's avatar
      Improve error message when certain flags are not specified. · 8b47c484
      Reed Wanderman-Milne authored 
      In nlp/train.py and vision/beta/train.py, certain flags are marked as required. Additionally, in certain functions, error messages are improved if a necessary flag is not specified, which is a fallback in case a file calling define_flags() does not mark the necessary flags are required. Previously if any of these flags were not specified, it would crash with a cryptic error message, making it hard to tell what went wrong.

In a subsequent change, I will mark flags as required in more files which call define_flags().

PiperOrigin-RevId: 381066985
      8b47c484
    • Reed Wanderman-Milne's avatar
      Improve error message when certain flags are not specified. · 0a9026e4
      Reed Wanderman-Milne authored 
      In nlp/train.py and vision/beta/train.py, certain flags are marked as required. Additionally, in certain functions, error messages are improved if a necessary flag is not specified, which is a fallback in case a file calling define_flags() does not mark the necessary flags are required. Previously if any of these flags were not specified, it would crash with a cryptic error message, making it hard to tell what went wrong.

In a subsequent change, I will mark flags as required in more files which call define_flags().

PiperOrigin-RevId: 381066985
      0a9026e4
    • Reed Wanderman-Milne's avatar
      Return default strategy from get_distribution_strategy when given "off". · 9f9d07e9
      Reed Wanderman-Milne authored 
      Before, it returned None. But almost every use of get_distribution_strategy() assumes an actual strategy is returned and crashes when None is returned. Returning the default strategy fixes these issues and is equivalent to using no strategy, as the default strategy is always in effect when no other strategy is used.

PiperOrigin-RevId: 380951055
      9f9d07e9
    • Reed Wanderman-Milne's avatar
      Return default strategy from get_distribution_strategy when given "off". · 6dc4ae73
      Reed Wanderman-Milne authored 
      Before, it returned None. But almost every use of get_distribution_strategy() assumes an actual strategy is returned and crashes when None is returned. Returning the default strategy fixes these issues and is equivalent to using no strategy, as the default strategy is always in effect when no other strategy is used.

PiperOrigin-RevId: 380951055
      6dc4ae73
  11. 05 Apr, 2021 2 commits
  13. 10 Mar, 2021 2 commits
  15. 08 Mar, 2021 2 commits
  17. 25 Feb, 2021 2 commits
  19. 03 Feb, 2021 2 commits
  21. 05 Jan, 2021 2 commits
  23. 29 Dec, 2020 2 commits
  25. 17 Nov, 2020 2 commits
  27. 30 Oct, 2020 2 commits
  29. 26 Oct, 2020 2 commits
  31. 01 Oct, 2020 2 commits
    • Rick Chao's avatar
      PSv2: Replace existing `tf.distribute.experimental.ParameterServerStrategy`... · 096803bd
      Rick Chao authored 
      PSv2: Replace existing `tf.distribute.experimental.ParameterServerStrategy` usage with `tf.compat.v1.distribute.experimental.ParameterServerStrategy` to prepare for the upcoming TF2 ParameterServerStrategy API release.

The practically only difference for API endpoint switch is the monitoring from V2 to V1, for those who were using `tf.distribute.experimental.ParameterServerStrategy`. It's not supported in V2 and should be tracked as V1 anyway.

PiperOrigin-RevId: 334847114
      096803bd
    • Rick Chao's avatar
      PSv2: Replace existing `tf.distribute.experimental.ParameterServerStrategy`... · 4680f2fa
      Rick Chao authored 
      PSv2: Replace existing `tf.distribute.experimental.ParameterServerStrategy` usage with `tf.compat.v1.distribute.experimental.ParameterServerStrategy` to prepare for the upcoming TF2 ParameterServerStrategy API release.

The practically only difference for API endpoint switch is the monitoring from V2 to V1, for those who were using `tf.distribute.experimental.ParameterServerStrategy`. It's not supported in V2 and should be tracked as V1 anyway.

PiperOrigin-RevId: 334847114
      4680f2fa
  33. 17 Sep, 2020 2 commits
  35. 13 Sep, 2020 2 commits
  37. 02 Sep, 2020 2 commits
  39. 27 Aug, 2020 2 commits
  41. 14 Aug, 2020 1 commit