Glossary term
Glossary term
Foundations
The weights and biases that a model learns during training. For example, in a linear regression model, the parameters consist of the bias (b) and all the weights (w1, w2, and so on) in the following formula:
In contrast, hyperparameters are the values that you (or a hyperparameter tuning service) supply to the model. For example, learning rate is a hyperparameter.
Created for this library
A research team reports parameter count when comparing model sizes so stakeholders see the cost dimension.
An ML platform team tracks parameter counts of production models to plan device memory and serving cost.
A startup chooses model parameter counts based on inference budget rather than maximum accuracy because per-call cost dictates the design.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License