1/ I just noticed that the latest @TensorFlow release is still on Numpy 1.19, which is increasingly out of step with the rest of the ecosystem. This raises a few questions for me
Show this thread
-
For specific technical concerns (like Numpy dependency version) I'd encourage you to reach out directly to tensorflow-users@google.com.
In general, the existence of other frameworks besides TF should not come at a surprise -- to use a programming language analogy, the existence of Go does not mean that C++ is going out of fashion. They're just different languages.
-
-
You should generally expect several frameworks to coexist going forward, corresponding to different user profiles and different tradeoffs and philosophies. The stable state of a mature ML ecosystem is a small number of specialized frameworks, rather than a single universal choice
-
Just like for programming languages, really. Anyway, happy to discuss more if you're interested
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.