Anybody have any clever hints on how to speed up training a net from scratch when using lots of images? #deeplearning
I would recommend the use of a downsized version of Xception. In general, separable convolutions are faster and more parameter-efficient
-
-
e.g. something along these lines https://gist.github.com/fchollet/0f8eaa08f09e33e547431023d5955709 …
-
Ok, thank you, I will try that.
End of conversation
New conversation -
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.