class Neuratron::Optimizers::SGD

• Neuratron::Optimizers::SGD
• Neuratron::Optimizer
• Reference
• Object

Defined in:

neuratron/optimizers/SGD.cr

Constructors

• .new(learning_rate : Float64 = 0.1, momentum : Float64 = 0.0, decay : Float64 = 0.0)

Instance Method Summary

• #call(x_input, layers, expected, loss) : Array(LA::GMat)

  SGD have too many responsabilities and may work only for Dense layers

Constructor Detail

Instance Method Detail