GitLab

* from six.moves import xrange for Python 3

* from __future__ import print_function for Python 3

Also old style exception --> new style exception

* Define basestring for Python 3

Also remove unused import.

* from six import string_types for Python 3

* from __future__ import print_function

* print() --> print('')

Thanks!

Please use the forum for Swivel discussions instead of contacting the authors directly... thank you!

* Refactor to use distributed TF API; add driver script.

* Add explicit GPU flag so multiple cards can be used.

* Update docs to mention distributed.sh.

* Don't make use of HParams yet, since they're not available in TF1.0

This speeds up the training accordingly.

There is no need in building Tensorflow from scratch, fastprep successfully links with the Python wrapper from pip and system protobuf v3. Tested this on Ubuntu.

- TF1.0 has breaking changes for tf.concat
- Replace deprecated summary api
- Replace to be deprecated initialize_all_variables

Add a tool to convert Glove co-occurrence matrices into Swivel shards. (Roundabout way to address issue #127). (#151)