What is the difference between a deep learning library that is built for research and for production?

(Bassel) #1

i am sorry this question is not really related to pytorch but i thought about it after i knew about pytorch and caffe2.
I always see things like "a deep learning library for research but not so much for production deployment"
What does that mean?
What should i choose?
Can i built a an advanced model (ex: to compete in image-net) using anyone of them?

(Hugh Perkins) #2

For one thing, as far as I know Facebook uses Caffe for production, and torch/pytorch for research, though I could be wrong on this point?