Looks like a graphical version of Keras, wdyt @fchollet ? #keras #MLpic.twitter.com/rdJ72tTM8a
-
I think ultimately there will be an IDE for deep learning, but "defining models" will be a very small part of it. It should cover everything from data management, data exploration, autoML search, benchmarks and evaluation...
The problem with model def GUIs is that they only cover the easy part of the ML workflow (creating a Keras model) and they don't make it meaningfully easier -- the actions required are the same and the cognitive load on the user is the same. Except now it's visual instead of code
-
-
For a GUI to make sense, it needs to make the underlying workflow significantly easier / simpler (i.e. fewer actions and reduce cognitive load). It also need to focus on the difficult parts of the ML workflow (not model definition), e.g. data management & problem definition
-
I agree - The modeling is one part of the solution, and data management etc. needs to be included. I'm curious how you'd envision an implementation of "problem definition"
@fchollet?
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.