Every story and comment from the Hacker News API, in a layout borrowed from 2001.
A comment on Cognitive Toolkit beta for deep learning advances.
New ops in tensorflow seem to be oriented towards forward-mode AD rather than reverse-mode (for which one needs a pull back op on a dual).
There are no comments on this comment yet.
New ops in tensorflow seem to be oriented towards forward-mode AD rather than reverse-mode (for which one needs a pull back op on a dual).