Re-write data processing code to use `dask.distributed` for serving data to models

[lzamparo]

This is a bit pie-in-the-sky, but eventually I'd like to not have the burden of micro-managing data processing, instead off-loading scheduling and load-balancing to dask.

This would involve a totally new dataset_reader (or maybe just parallel functionality within?), which would function something like this:

• dask-scheduler from dask.distributed works to distribute loads sets of factor files for training individually
• dask unpacks the files, turning them into macrobatches
• still need to understand how dask & Theano will work together; maybe working on the Scipy Dask tutorial will help.

[lzamparo]

Working on this in collaboration with Dask guy. Doesn't seem like Dask distributed is really ready for prime-time; there's no PBS template that just works, nevermind LSF.