Module is an abstract class which defines fundamental methods necessary for a training a neural network. Modules are serializable.

Modules contain two states variables: output and gradInput.

Subsections

  1. [output] forward(input)
  2. [gradInput] backward(input, gradOutput)
  3. zeroGradParameters()
  4. updateParameters(learningRate)
  5. share(mlp,s1,s2,...,sn)
  6. clone(mlp,...)
  7. State variables