Enum leaf::solver::LRPolicy
[−]
[src]
pub enum LRPolicy { Fixed, Step, Exp, }
Learning Rate Policy for a Solver
The variables mentioned below are defined in the SolverConfig apart from iter, which is the current iteration of the solver, that is supplied as a parameter for the learning rate calculation.
Variants
Fixed | always return base_lr | |
Step | learning rate decays every | |
Exp | return base_lr * gamma ^ iter |