You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
People have correctly pointed out that every one of these graphs is just tracking overall deep learning adoption, so they aren't hugely informative. Indeed! Here's a market share plot since 2016 that's more interesting.https://twitter.com/fchollet/status/1029477656876613632 …
You can see a long-term decline of niche options (i.e. consolidation), but things are very stable otherwise. PyTorch seems to have peaked about 6 month ago. I expect to see more consolidation in the future.pic.twitter.com/zQyMbFmpOu
PyTorch is just getting started. Pragmatic Facebook did a pretty good job with Apache Hive too, for example. And they have LeCun. And Mikolov. New papers are implemented in PyTorch. Karpathy, http://Fast.ai love it. It is no-nonsense, easy to use and does the job.
It seems that pytorch took something of tensorflow and not from keras. It is obvious but in any case interesting.
Market share = number_of _search_counts / total_search_counts?
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
