Module light_labyrinth.hyperparams.weights_init
The light_labyrinth.hyperparams.weights_init
module includes WeightsInit
class
that allows to initialize model's weights before training.
Expand source code
"""
The `light_labyrinth.hyperparams.weights_init` module includes `WeightsInit` class
that allows to initialize model's weights before training.
"""
from ._weights_initialization import LightLabyrinthWeightsInit
__all__ = ["LightLabyrinthWeightsInit"]
Classes
class LightLabyrinthWeightsInit (value, names=None, *, module=None, qualname=None, type=None, start=1)
-
This class allows to initialize Light Labyrinth model's weights either with zeros or randomly.
Examples
>>> from light_labyrinth.hyperparams.weights_init import LightLabyrinthWeightsInit >>> from light_labyrinth.dim2 import LightLabyrinthClassifier >>> model = LightLabyrinthClassifier(3, 3, ... weights_init=LightLabyrinthWeightsInit.Random)
Expand source code
class LightLabyrinthWeightsInit(Enum): """ This class allows to initialize Light Labyrinth model's weights either with zeros or randomly. Examples -------- >>> from light_labyrinth.hyperparams.weights_init import LightLabyrinthWeightsInit >>> from light_labyrinth.dim2 import LightLabyrinthClassifier >>> model = LightLabyrinthClassifier(3, 3, ... weights_init=LightLabyrinthWeightsInit.Random) """ Default = 0 """ """ Random = 1 """ Initializes model's weights randomly -- initial weights are drawn from the uniform distribution [-1,1]. """ Zeros = 2 """ Initializes model's weights with zeros. """
Ancestors
- enum.Enum
Class variables
var Default
var Random
-
Initializes model's weights randomly – initial weights are drawn from the uniform distribution [-1,1].
var Zeros
-
Initializes model's weights with zeros.