Used to initialize a parameter in the neural network.
Creates a parameter initialiser that sets the initial value of each element in a parameter tensor to a constant value.
Creates a parameter initialiser that sets the initial value of each element in a parameter tensor to a different sample from a Gaussian distribution.
Creates a parameter initialiser that uses the method of Glorot and Bengio (2010).
Creates a parameter initialiser that uses the method of Glorot and Bengio (2010).
Creates a parameter initialiser that uses the method of He et al. (2015).
Creates a parameter initialiser that uses the method of He et al. (2015).
Creates a parameter initialiser that sets the initial value of each element in a parameter tensor to a different sample from a uniform distribution.
Encapsulates information about network parameters.
This module contains methods for initialising the parameters of neural networks.
Several of the methods implemented in this module rely on fan_in and fan_out values. These are calculated differently depending on the rank of the parameter.
For rank-2 tensors,
fan_in = shape[0], fan_out = shape[1]
For rank-4 tensors,
fan_in = shape[1] * shape[2] * shape[3], fan_out = shape[0] * shape[2] * shape[3]