10 ms·
even for predictions, you need the model (what was learned in the training phase). deep learning models are of multiple gigs in size. so, in browser wouldn't be
by make3 10y ago
even for predictions, you need the model (what was learned in the training phase). deep learning models are of multiple gigs in size. so, in browser wouldn't be practical, except for toy stuff.
sending the input data to the server, doing the computations there and getting the answers back will be the only practical way to go for remotely serious applications for a while still
- Scea91 10y agoAnd you would also have to make the model available to the client, which is not reasonable in many commercial applications.