RNNCell_cdopt

RNNCell_cdopt#

CLASS cdopt.nn.RNNCell_cdopt(input_size, hidden_size, bias=True, nonlinearity='tanh', device=None, dtype=None, manifold_class = euclidean_torch, penalty_param = 0, manifold_args = {})

An Elman RNN cell with tanh or ReLU non-linearity,

h′=tanh⁡(Wihx+bih+Whhh+bhh),

where the weight for hidden states Whh is constrained over the manifold defined by manifold_class.

If nonlinearity is relu, then ReLU is used in place of tanh.

Parameters#

  • input_size – The number of expected features in the input x

  • hidden_size – The number of features in the hidden state h

  • bias – If False, then the layer does not use bias weights b_ih and b_hh. Default: True

  • nonlinearity – The non-linearity to use. Can be either 'tanh' or 'relu'. Default: 'tanh'

  • manifold_class – The manifold class for the weight matrix. Default: cdopt.manifold_torch.euclidean_torch

  • penalty_param – The penalty parameter for the quadratic penalty terms in constraint dissolving function

  • manifold_args - The additional key-word arguments that helps to define the manifold constraints.

Shapes#

  • input: (N,Hin) or (Hin) tensor containing input features where Hin = input_size.

  • hidden: (N,Hout) or (Hout) tensor containing the initial hidden state where Hout=hidden_size. Defaults to zero if not provided.

  • output: (N,Hout) or (Hout) tensor containing the next hidden state.

Attributes#

  • manifold (cdopt manifold class) – the manifold that defines the constraints. The shape of the variables in manifold is set as var_shape.

  • weight_ih (torch.Tensor) – the learnable input-hidden weights, of shape (hidden_size, input_size)

  • weight_hh (torch.Tensor) – the learnable hidden-hidden weights, of shape (hidden_size, hidden_size)

  • bias_ih – the learnable input-hidden bias, of shape (hidden_size)

  • bias_hh – the learnable hidden-hidden bias, of shape (hidden_size)

  • quad_penalty (callable) – the function that returns the quadratic penalty terms of the weights. Its return value equals to ||manifold.C(weight)||2.

Example#

rnn = cdopt.nn.RNNCell_cdopt(10, 20)
input = torch.randn(6, 3, 10)
hx = torch.randn(3, 20)
output = []
for i in range(6):
    hx = rnn(input[i], hx)
    output.append(hx)
    
print(output)
Contents