Module

Modules are the core components that make up deep learning models. Layers, Activation Functions, and containers are all modules. all modules share the common traits that they take a batch of data and output a batch of transformed data, and that they can be chained together.

Layers

Layers are the most common type of module and what most think of as deep learning components. All layers have learnable parameters (weights and biases).

Activation Functions

Activations are mathematical functions applied after layers and determin final neuron output. Activation functions allow models to learn complex non-linear representations, and prevent collapse during training.

Containers

containers are orchestrator modules. They take other modules and define how data flows through the model.

Modules

Name	Module Type
Dense	Layer
Conv2d	Layer
Flatten	Layer
Max Pool2d	Layer
Adaptive Max Pool2d	Layer
Sigmoid	Activation Function
Tanh	Activation Function
ReLU	Activation Function
Leaky ReLU	Activation Function
Softmax	Activation Function
Sequential	container