Glossary term
Glossary term
Foundations
The layer of a neural network that holds the feature vector. That is, the input layer provides examples for training or inference. For example, the input layer in the following neural network consists of two features:
Created for this library
A retail forecasting team designs its input layer to accept categorical and continuous features with separate embedding and dense paths.
An NLP team uses a token embedding input layer that is shared between encoder and decoder to reduce parameter count.
A computer vision team includes per-channel normalization at the input layer of its CNN so it can accept raw image tensors.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License