The future: one standard deep learning API, train on TPU, export to any platform (server-side, mobile, browser, embedded).
-
-
Show this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
is transfer learning by simply adding a fully connected layer to the softmax outputs a good approach? i haven't encountered that before.pic.twitter.com/3XIHy1VKjc
-
It's not very different from other transfer learning approaches. One thing to keep in mind though is that Dense layers are pretty destructive (in particular you lose spatial info), so their features will only be useful for classes that are related to the original classes.
- Show replies
New conversation -
-
-
Can != should
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.