Keras usage is most dominant in the industry (both large companies and startups) and in the overall data science community. In the research community, Keras has a much smaller market share
Show this thread
TensorFlow is the platform of choice for deep learning in the research community. These are deep learning framework mentions on arXiv over the past 3 monthspic.twitter.com/v6ZEi63hzP
-
-
-
-
Unfortunately I had to remove forks from the score, as they appear to be artificially inflated for TF. TF has 60k forks, growing at 10k/month, which doesn't make any sense. Possibly someone misusing the GitHub API...pic.twitter.com/eUSdXdtRo4
Show this thread
End of conversation
New conversation -
-
-
What do you use to search over Arxiv here?
-
Just the Google search index. We don't crawl and parse the papers, but that would be more accurate (and a lot more work intensive to set up)
- Show replies
New conversation -
-
-
How easy to plot # of papers mentioning use of linear/logistic regressions as well as RF models?
-
Most machine learning can be done in excel.
End of conversation
New conversation -
-
-
Kind of surprised mxnet and PyTorch appear so infrequently. Though relative to TF, which has simply been around longer, I suppose it makes sense.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
interesting to see that Theano is going down (or, is down) in popularity/activity. for a while there that was the pkg of choice. I guess now it's more of an academic tool with little to no use in Industry. again, very interesting.
-
MILA has deprecated support for theano (http://deeplearning.net/software/theano/NEWS.html …). It was my first deep learning framework that I learnt and I was a little sad when I heard about it but I think a consolidation of frameworks might be a good thing for the scene.
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.