Current status?

Thanks for sharing this work =).

Just curious - what is the current status of the project? Are you interesting in updating it to say, Caffe2, or trying it with TF again?

Any chance you're OK sharing the original data-set used for training? (I sent an email to the email in the README).