We're writing an academic paper about fastai2. As well as citing the obvious prior work (pytorch, tf, blocks/fuel, keras) I'd like to make sure we don't miss stuff with significant innovations. Any suggestions for stuff we should look at (which launched specific new innovations)?
How about tensorly? It has multiple back ends and allows high level tensor operations and tensorized neural networks @JeanKossaifi
-
-
Thanks for the suggestion! I've often promoted Tensorly, as you know - although fastai doesn't have any functionality inspired by or related to it, so a citation wouldn't quite fit, unfortunately
-
Thanks
@jeremyphoward I hope in future you can incorporate tensorly. Happy to chat more about it
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.