Module Optimise.Learning_Rate

Strategies for learning rate update

type typ =
  1. | Adagrad of float
  2. | Const of float
  3. | Decay of float * float
  4. | Exp_decay of float * float
  5. | RMSprop of float * float
  6. | Adam of float * float * float
  7. | Schedule of float array

Representation of learning rate update strategies. Possible values include:

  • Adam (alpha, beta1, beta2), see ref for parameter meaning
val run : typ -> int -> Algodiff.t -> Algodiff.t array -> Algodiff.t

Execute the computations defined in module typ.

val default : typ -> typ

Create module typ with default values.

val update_ch : typ -> Algodiff.t -> Algodiff.t array -> Algodiff.t array

Update the cache of gradients.

val to_string : typ -> string

Convert the module typ to its string representation.