Better support for conv nets with TfModel Interface

What’s Changed

Also includes:

• Possibly changes with the version record attributes.
• Deprecated support for old TF versions

• fixed a few bugs on not normalizing output data when building and tra… (#37) @grantbuster

More features!

What’s Changed

Added more features including squeeze and excitation layer and better integration of the phygnn interface with higher dimensional inputs

• Gb/more sup3r features (#36) @grantbuster

New custom random axis layer

What’s Changed

• Gb/rand axis layer (#35) @grantbuster

Added more data manipulation custom layers for sup3r project

What’s Changed

• Gb/more sup3r layers (#34) @grantbuster

Bug fix on st expansion layer

What’s Changed

Spatiotemporalexpansion layer was not able to handle new coarse dimensions. Fixed that and tested.

• Gb/st expand bug (#32) @grantbuster

Added layer to flatten tensor dimensions

What’s Changed

• Gb/unpack st (#31) @grantbuster

More features for sup3r project

What’s Changed

• Gb/sup3r features (#30) @grantbuster

Architecture Refactor

What’s Changed

Major refactor of phygnn architecture to break out a custom network base class and isolate a module for custom phygnn layers. The reorganization will likely cause breaking changes for old saved phygnn models.

• Gb/refactor (#29) @grantbuster

Bug fixes and better integration and testing with tf1.15

What’s Changed

Better integration and testing across all TF versions from 1.15 to 2.6.

• Gb/fix tf save (#26) @grantbuster
• fixed tf save model bug in tf2.6, still works in 2.4 (#25) @grantbuster

Refactor for convolutional layers and multi dimensional features

What’s Changed

Significant refactor to allow for convolutional layers in the phygnn model along with multi dimensional features or labels (tested up to 5D features). Includes BREAKING CHANGES with the phygnn model save/load functionality. The model Layers handler object is now saved directly to the pkl file instead of trying to save and load model weights.

• Gb/conv rnn layers (#24) @grantbuster
• Gb/nsrdb gan (#23) @grantbuster