Dan Petro

Add our screenshots to the TF datasets public repository so that it's easy to use. https://www.youtube.com/watch?v=-nTe44WT0ZI https://www.tensorflow.org/datasets/add_dataset

Add loss as an evaluation metric

3

Return and print the average loss against the evaluation set.

So TF2.0 has an automatic hyperparameter search functionality that is just SO cool: https://www.tensorflow.org/tensorboard/r2/hyperparameter_tuning_with_hparams Maybe we should implement something like this in our code... or perhaps think about porting up...

Dufflebag currently whitelists all /etc/shadow files. But in practice, a lot of these are blank and don't contain any interesting info. It would be nice to do a light parsing...