The Conference for Machine Learning Innovation

How to write a good Training Script?

Session
Join the ML Revolution!
Register until September 15:
✓ Save up to $323
✓ 3 Day Special
✓ Team discount
Register Now
Join the ML Revolution!
Register until September 15:
✓ Save up to $323
✓ 3 Day Special
✓ Team discount
Register Now
Join the ML Revolution!
Register until August 18:
✓ Pre-conference workshops for free
✓ Save up to €503
✓ 10 % Team Discount
Register Now
Join the ML Revolution!
Register until August 18:
✓ Pre-conference workshops for free
✓ Save up to €503
✓ 10 % Team Discount
Register Now
Join the ML Revolution!
Until the Conference starts:
✓ Group discount
✓ Special discount for freelancers
Register Now
Join the ML Revolution!
Until the Conference starts:
✓ Group discount
✓ Special discount for freelancers
Register Now

While Jupyter Notebooks are a great way to do research and iterate fast, at some point comes the time where you need a robust, reliable and repeatable training process. So it is time to write your ML/DL model down in a script that can be run automatically and for much longer durations on, for example, a training server. In this talk I will present one opinionated way to write a good training script using pytorch. The talk will address a number of issues that I have encountered in practice and noticed beginners struggle with: – What needs to be configurable, what needs to be hardcoded? – How do we configure hyperparameters? – How do we make experiments repeatable? – When do we need "software architecture", when is simple copy-pasting enough? – What do we need to track and log, and how? – How to best structure your script? The goal is to give a a coarse but complete overview on how to do professional model training.

This Session belongs to the Diese Session gehört zum Programm vom BerlinBerlin program. Take me to the program of . Hier geht es zum Programm von Singapore Singapore .

Take me to the full program of Zum vollständigen Programm von Berlin Berlin .

This Session belongs to the Diese Session gehört zum Programm vom BerlinBerlin program. Take me to the program of . Hier geht es zum Programm von Munich Munich .

This Session Diese Session belongs to the gehört zum Programm von BerlinBerlin program. Take me to the current program of . Hier geht es zum aktuellen Programm von Singapore Singapore , Berlin Berlin or oder Munich Munich .

Behind the Tracks