Skip to content

Releases: neptune-ai/neptune-tensorflow-keras

2.2.2

14 Mar 08:48
a54d412

Choose a tag to compare

What's Changed

Full Changelog: 2.2.1...2.2.2

2.2.1

29 Jan 17:21
0f6be9e

Choose a tag to compare

What's Changed

New Contributors

Full Changelog: 2.2.0...2.2.1

2.2.0

24 Jul 09:04
efdc4e4

Choose a tag to compare

Fixes

  • Support logging learning_rate in Multi-GPU training set-up (#57)

2.1.1

31 Mar 14:45
3b3dec4

Choose a tag to compare

  • Update Callback docstrings (#53)

2.1.0

13 Mar 08:41
66c576f

Choose a tag to compare

Changes

  • Removed neptune and neptune-client from base requirements.
  • Added checking neptune or neptune-client installation during runtime.

2.0.1

06 Mar 15:27
e0d5b5a

Choose a tag to compare

Updated requirements, so that the integration now uses and installs neptune, instead of neptune-client

2.0.0

06 Feb 10:22
042aac1

Choose a tag to compare

We made a bunch of changes that made this integration consistent with the recent update of the neptune-client library. To use this version of the integration, you will need neptune-client to the most recent version.

1.2.1

24 Jan 11:22
12c8c1e

Choose a tag to compare

Remove misleading information from the docstring. You need to install pydot to save the model diagram.

1.2.0

07 Nov 13:42
e976607

Choose a tag to compare

Changes

  • Moved neptune_tensorflow_keras package to src dir (#33)
  • Poetry as a package builder (#40)

Fixes

  • Fixed NeptuneCallback import error - now possible to directly import with from neptune_tensorflow_keras import NeptuneCallback (#35)
  • Drop usage of deprecated File.extension init attribute #38 (#38)

1.1.0

04 Oct 10:42
a181b9a

Choose a tag to compare

We fixed a bug that crashed the integration with an error message ValueError: This model has not yet been built. Build the model first by calling build() or by calling the model on a batch of data., when the input layer of a neural network didn't have the input_shape parameter defined or the model was not built.