Lessons learned from building machine learning systems

In this blog post I want to collect some of the lessons that I’ve learned over the years building machine learning systems.

Lessons learned

1. How to set up Amazon Sagemaker for model training
2. How to optimize your Docker: install your dependencies first and your code second
3. How to scale up your code
4. When you make a dataset, make a small statistics file
5. Make sure you can find back your mlflow runs
6. Keep seeking input
7. Use assertions to check tensor sizes
8. Your argparse is not the interface of your application

Subscribe

I blog about how to grow as a machine learning engineer! Liked this article and want to hear more? Join 40+ others and subscribe!