Keras tip: if you're training models in unreliable environments like spot instances (or for long stretches of time), make sure to use the BackupAndRestore callback to be able to resume training where you left off, in just a couple of lines of configurationhttps://www.tensorflow.org/api_docs/python/tf/keras/callbacks/experimental/BackupAndRestore …
-
I'm a pytorch user. But seeing u use the right insights (spot instance) makes me think of going back to tensorflow..
Making everything cloud-backed and instantly restorable is a big life improvement in any framework. Also a big life improvement: if you're a Keras user you'll probably use Keras 90% of the time and TF 10% of the time.
-
-
I learned deeplearning through the keras interface (easy, understandable and programmer friendly).. Unfortunately had to move to pytorch, but maybe will be back
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
In your opinion sir does the 10% need to use pure tensorflow code although tensorflow 2 make it easier with keras integration ( tf.keras) or they just want to ?
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.