swift for tensorflow vs tensorflow lite?
Do numpy and pandas count as “non-data science modules”?
I highly doubt Python will be deprecated any time soon…fwiw
I bet the S4TF team is absolutely thrashing themselves to be ready for Chris’ first lesson here. The AI world will be watching, in a serious way, so the pressure will be on I’m sure they’ll make a huge amount of progress over the next month. You can follow/watch the Swift for Tensorflow github project to get an idea of how much activity there is - I’ve been following it for a while and it’s been going off.
Question: Are we going to implement backprop from scratch?
they presented working code already at TF dev summit recently
I got the impression numpy isnt part of the building blocks, assuming pytorch tensors should accomplish the same things.
Also from the fast.ai Embracing Swift for Deep Learning post:
“The combination of Python, PyTorch, and fastai is working really well for us, and for our community. We have many ongoing projects using fastai for PyTorch, including a forthcoming new book, many new software features, and the majority of the content in the upcoming courses. This stack will remain the main focus of our teaching and development.”
where do we get help if we have issues with fastai installation?
Maybe Jeremy will present it a bit later, but in case he doesn’t : would it be possible to have a rough week-by-week schedule of the course ?
Yes - there was a LOT of activity on the repo in the few days before the TF Dev Summit
You just had it
We have nothing more than the initial introduction.
I think these 2 libraries are specifically built for machine learning and data processing
Part of the presentation is cut off due to the placement of Jeremy’s video on the top right. Is there a way around it in the future?
I told him not to put anything there, but did he listen?..
Thanks, we will keep this in mind for future weeks.
Jeremy on RL:
You lost me at sqweegel