Weather forecasting with AI, Kaggle tips and tricks, dealing with missing data, deep learning with Jesper Dramsch, The Data Scientist Show #040
The Data Scientist Show - Daliana Liu
00:00
Pi Torch Lightning and Distributed Data Centres
In my experience, caros is enough for most of things. And i think we're seeing that development as well with pytorch lightning right now. Getting pytorch models on to a distributed architecture can be a bit difficult. You have to adjust all your devices and and all that stuff. But in caros, you literally just say, use this strategy for for my for my model,. And i think pi torch lightning does it very, very similarly. So am i surprised that people really thought of and it was so easy to do? I'm very happy that i don't have to do that. A unless you have actual, like, distributed data centres where you not only train
Transcript
Play full episode
