Keras now works with Theano, Tensorflow, CNTK, and MXNet! Sharing models and weights is easier if all frameworks unify around same API
https://twitter.com/fchollet/status/869622717535862784 …
-
-
Replying to @MehdiShibahara
But we don't have a shared weights format, right?
@fchollet1 reply 0 retweets 0 likes -
Replying to @jeremyphoward @MehdiShibahara
We do though. Keras models can be saved and loaded across backends.
2 replies 3 retweets 18 likes -
Replying to @fchollet @MehdiShibahara
Wow that's great! This is a keras 2 thing? It handles dimension ordering differences automatically?
2 replies 0 retweets 7 likes
Replying to @jeremyphoward @MehdiShibahara
Yes, check it out!
7:14 PM - 30 May 2017
0 replies
0 retweets
8 likes
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.
Keras CNTK backend: