I am used to do all my programming in emacs and using the jupyter notebook doesn’t seem as efficient when programming large pieces of code. Although I can see that the notebooks can be a good collaborative tool I wonder to what extent it is really used. In lesson 1 most of the code is done in separate .py files instead of notebooks so I wonder which is the real workflow when solving deep learning problems.
Thanks
I guess this is different for everyone, but I tend to prototype in jupyter and then gradually factor out into proper python as I’m happy with certain pieces.
And ever since discovering https://github.com/lambdalisue/jupyter-vim-binding my productiveness in jupyter has gone way up.