rockpool.nn.modules.torch

Modules using Torch as a backend

Modules

rockpool.nn.modules.torch.ahp_lif_torch

Implement a modified version of LIF Module (ahp, after hyperpolarization feedback, is added)), using a Torch backend

rockpool.nn.modules.torch.bool_state

A module implementing a boolean state variable

rockpool.nn.modules.torch.dropout

A module implementing random dropout of neurons and time steps

rockpool.nn.modules.torch.exp_syn_torch

Implement a exponential synapse module, using a Torch backend

rockpool.nn.modules.torch.lif_bitshift_torch

Implement a LIF Module with bit-shift decay, using a Torch backend

rockpool.nn.modules.torch.lif_neuron_torch

Implement a LIF Neuron Module, using a Torch backend

rockpool.nn.modules.torch.lif_torch

Implement a LIF Module, using a Torch backend

rockpool.nn.modules.torch.linear_torch

Implement a linear module, using a Torch backend

rockpool.nn.modules.torch.lowpass

"

rockpool.nn.modules.torch.rate_torch

Rate dynamics module with torch backend

rockpool.nn.modules.torch.torch_module

Provide a base class for build Torch-compatible modules

rockpool.nn.modules.torch.updown_torch

Feedforward layer that converts each analogue input channel to one spiking up and one down channel