Glossary term
Glossary term
Foundations
A full training pass over the entire training set such that each example has been processed once.
An epoch represents N/batch size training iterations, where N is the total number of examples.
For instance, suppose the following:
The dataset consists of 1,000 examples.
The batch size is 50 examples.
Therefore, a single epoch requires 20 iterations:
1 epoch = (N/batch size) = (1,000 / 50) = 20 iterations
See Linear regression: Hyperparameters in Machine Learning Crash Course for more information.
Created for this library
An ML team trains a vision model for 90 epochs based on a published recipe and tunes the learning rate schedule across those epochs.
A speech recognition team trains for 10 epochs on a fresh dataset and uses early stopping to choose the best epoch by validation word error rate.
A churn team trains for one epoch on a very large click-stream dataset because additional epochs no longer improve validation metrics.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License