Skip to content

GitLab

Switch branch/tag
  1. 18 Jan, 2022 1 commit
  3. 04 Jan, 2022 1 commit
  5. 29 Dec, 2021 1 commit
    • Xin Wang's avatar
      Added block diagonal feedforward layer. · a7e60974
      Xin Wang authored 
      This layer replaces the weight matrix of the output_dense layer with a block diagonal matrix to save layer parameters and FLOPs. A linear mixing layer can be added optionally to improve layer expressibility.

PiperOrigin-RevId: 418828099
      a7e60974
  7. 21 Dec, 2021 1 commit
  9. 16 Dec, 2021 1 commit
  11. 03 Dec, 2021 1 commit
  13. 30 Nov, 2021 1 commit
  15. 25 Nov, 2021 1 commit
  17. 10 Nov, 2021 2 commits
  19. 04 Nov, 2021 1 commit
  21. 09 Oct, 2021 1 commit
  23. 08 Oct, 2021 1 commit
  25. 07 Oct, 2021 1 commit
  27. 24 Sep, 2021 1 commit
  29. 10 Sep, 2021 2 commits
  31. 25 Aug, 2021 1 commit
  33. 09 Aug, 2021 1 commit
  35. 28 Jul, 2021 2 commits
  37. 13 Jul, 2021 1 commit
  39. 10 Jul, 2021 1 commit
  41. 03 Jul, 2021 1 commit
  43. 06 Jun, 2021 2 commits
  45. 19 May, 2021 1 commit
  47. 18 May, 2021 1 commit
  49. 17 May, 2021 1 commit
  51. 14 May, 2021 1 commit
  53. 13 May, 2021 1 commit
  55. 20 Apr, 2021 1 commit
  57. 17 Apr, 2021 1 commit
    • Chen Chen's avatar
      Catch tf.text NotFoundError: · ffaa4035
      Chen Chen authored 
      tensorflow.python.framework.errors_impl.NotFoundError: /usr/local/lib/python3.6/dist-packages/tensorflow_text/python/metrics/_text_similarity_metric_ops.so: undefined symbol: _ZN10tensorflow6StatusC1ENS_5error4CodeEN4absl12lts_2021032411string_viewEOSt6vectorINS_10StackFrameESaIS7_EE

PiperOrigin-RevId: 368978629
      ffaa4035
  59. 15 Apr, 2021 1 commit
  61. 13 Apr, 2021 2 commits
    • A. Unique TensorFlower's avatar
      Internal change · b1aa44d9
      A. Unique TensorFlower authored 
      PiperOrigin-RevId: 368260712
      b1aa44d9
    • Reed Wanderman-Milne's avatar
      Use nonexperimental mixed precision API. · 4334a892
      Reed Wanderman-Milne authored 
      This replaces symbols in tf.keras.mixed_precision.experimental with the corresponding nonexperimental symbols. In some cases, passing a Policy is replaced with passing a policy name for conciseness.

Additionally, for the Shakespeare model, the loss_scale flag is removed, since supporting it with the nonexperimental API is slightly more verbose and it is recommended users use the default loss scale.

PiperOrigin-RevId: 368123944
      4334a892
  63. 06 Apr, 2021 1 commit
    • Jeremiah Liu's avatar
      Disable temperature scaling during training. · fab47e9e
      Jeremiah Liu authored 
      For the `GaussianProcessClassificationHead`, the temperature scaling needs to be disabled during training to avoid unexpected modification to the learning rate, which harms model quality. (Unfortunately, this seems to require adding `training` to the `call` method).

Also set the default of `gp_cov_ridge_penalty` in `RandomFeatureGaussianProcess` to 1 to be consistent with that in the `GaussianProcessClassificationHead`.

PiperOrigin-RevId: 366917075
      fab47e9e
  65. 05 Apr, 2021 1 commit
    • Jeremiah Liu's avatar
      Adds `call()` method to `GaussianProcessClassificationHead`. · 2ae06c8a
      Jeremiah Liu authored 
      This change allows `GaussianProcessClassificationHead` to output only the predictive logits during training and evaluation (instead of outputting a tuple  `(logits, covmat)`). The goal is to make the layer more compatible with `SentencePredictionTask` and Keras' `model.fit()` API.

PiperOrigin-RevId: 366891298
      2ae06c8a
  67. 01 Apr, 2021 2 commits