Skip to content

GitLab

Switch branch/tag
  1. 18 Nov, 2021 1 commit
  3. 17 Nov, 2021 1 commit
  5. 12 Nov, 2021 2 commits
  7. 11 Nov, 2021 1 commit
  9. 04 Nov, 2021 1 commit
  11. 02 Nov, 2021 1 commit
  13. 26 Oct, 2021 1 commit
  15. 24 Sep, 2021 2 commits
  17. 23 Sep, 2021 1 commit
  19. 22 Sep, 2021 1 commit
  21. 17 Sep, 2021 1 commit
  23. 08 Aug, 2021 1 commit
  25. 05 Aug, 2021 2 commits
  27. 24 Jul, 2021 1 commit
  29. 23 Jul, 2021 1 commit
  31. 21 Jul, 2021 1 commit
  33. 20 Jul, 2021 1 commit
  35. 15 Jul, 2021 1 commit
  37. 13 Jul, 2021 1 commit
  39. 09 Jul, 2021 1 commit
    • Reed Wanderman-Milne's avatar
      With float16, always use LossScaleOptimizer. · 21286f77
      Reed Wanderman-Milne authored 
      Before, it was too easy to accidentally forget to set runtime.loss_scale, which had to always be done if mixed precision is used, otherwise the model would converge to worse accuracy. Now, all that needs to be done to use mixed precision is to set runtime.mixed_precision_dtype=float16.

PiperOrigin-RevId: 383767033
      21286f77
  41. 24 Jun, 2021 1 commit
  43. 23 Jun, 2021 1 commit
    • 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
  45. 22 Jun, 2021 1 commit
  47. 20 Jun, 2021 1 commit
  49. 16 Jun, 2021 1 commit
  51. 11 Jun, 2021 1 commit
  53. 01 Jun, 2021 1 commit
  55. 28 May, 2021 1 commit
  57. 17 May, 2021 1 commit
  59. 14 May, 2021 1 commit
  61. 13 May, 2021 1 commit
  63. 06 May, 2021 1 commit
  65. 16 Apr, 2021 1 commit
  67. 13 Apr, 2021 2 commits
  69. 12 Apr, 2021 1 commit
    • Reed Wanderman-Milne's avatar
      Use nonexperimental mixed precision API for official models. · 0d8f9807
      Reed Wanderman-Milne authored 
      For all modified calls to set_mixed_precision_policy(), the loss_scale argument was removed, as it cannot be passed if the nonexperimental API is used. For all such callers, the loss_scale is later used to explicitly create a LossScaleOptimizer, so removing the argument has no impact.

Switching to the non-experimental LossScaleOptimizer has no effect, as it has near identical behavior and all isinstance checks within the official models check for the non-experimental version.

PiperOrigin-RevId: 368101975
      0d8f9807
  71. 05 Apr, 2021 1 commit
    • Reed Wanderman-Milne's avatar
      Use nonexperimental LSO API in base_task.py. · cc12499b
      Reed Wanderman-Milne authored 
      This shouldn't break any official models, since I changed all LossScaleOptimizer isinstance checks to use the nonexperimental version (the experimental LSO subclasses the nonexperimental LSO, so changing isinstance checks in this way is always safe).

PiperOrigin-RevId: 366891847
      cc12499b