In general, search trends, GitHub metrics and Arxiv trends all tell complementary sides of a same story. It feels like model ensembling
Google web search interest around deep learning frameworks over time. If you remember Q4 2015 and Q1-2 2016 as confusing, you weren't alone.pic.twitter.com/1f1VQVGr8n
-
-
-
Comparing Keras & TensorFlow monthly mentions on Kaggle forums. Spike in TF = YouTube-8M competition.pic.twitter.com/FTvhUQVYnc
End of conversation
New conversation -
-
-
When you think we should use TensorFlow alone over Keras ?
-
They are complementary. Write raw TensorFlow for maximum flexibility when implementing brand new techniques or doing research.
- Show replies
New conversation -
-
-
What's the y-axis scale?
#makechartsgreatagainThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Where's the Keras one line multi-GPU data parallelism?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I remember Caffe being dominant. Oh, how quickly that changed with TF.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Once we found Tensorflow and Keras we havent looked back..
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
And now I want Keras reimplemented in Scala...
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Interesting. Curious: where would MxNet and DyNet be in this graph?
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.