niftynet.engine.application_optimiser module¶
To customise optimisers including new optimisation methods, learning rate decay schedule, or customise other optional parameters of the optimiser:
create a newclass.py that has a class NewOptimisor and implement get_instance(). and set config parameter in config file or from command line specify –optimiser newclass.NewOptimisor
-
class
niftynet.engine.application_optimiser.
Adagrad
¶ Bases:
object
Adagrad optimiser with default hyper parameters
-
static
get_instance
(learning_rate)¶ create an instance of the optimiser
-
static
-
class
niftynet.engine.application_optimiser.
Adam
¶ Bases:
object
Adam optimiser with default hyper parameters
-
static
get_instance
(learning_rate)¶ create an instance of the optimiser
-
static