How to Use Lightning to Improve Your Model Performance

The rule of thumb is something like 80% training, 20% validation. So the feature store would basically be something that you plug into the training work where what you do is process the data sets ahead of time. And then let's say it's like images, maybe tabular data is better. You process the data set ahead of times and you transform like you normalize time so that it's like an o'clock or whatever people are doing nowadays. It can actually dramatically speed up training, which is really cool.