How can DVC be used in a jupyter notebook based system for pipelining each stages? Where assuming all stages for data processing, feature engineering to even model tuning. We use just one single jupyter notebook system?
from the top of my head - system with Python decorators @danfischetti implemented for his team comes to my mind as a solution for this (it's not publicly available yet).
NB would probably be useful to change from tqdm import tqdm
to from tqdm.auto import tqdm
to get nice notebook progress bars
@casperdcl Oh, it will make it detect if it is running from a notebook? Neat!
Yes
Most helpful comment
Yes