brainpy.layers
module#
Basic ANN Layer Class#
|
Base class for a layer of artificial neural network. |
Convolutional Layers#
|
One-dimensional convolution. |
|
Two-dimensional convolution. |
|
Three-dimensional convolution. |
alias of |
|
alias of |
|
alias of |
|
|
One dimensional transposed convolution (aka. |
|
Two dimensional transposed convolution (aka. |
|
Three dimensional transposed convolution (aka. |
Dropout Layers#
|
A layer that stochastically ignores a subset of inputs each training step. |
Function Layers#
|
Applies an activation function to the inputs |
|
Flattens a contiguous range of dims into 2D or 1D. |
|
Dense Connection Layers#
|
A linear transformation applied over the last dimension of the input. |
alias of |
|
|
A placeholder identity operator that is argument-insensitive. |
Normalization Layers#
|
1-D batch normalization [1]_. |
|
2-D batch normalization [1]_. |
|
3-D batch normalization [1]_. |
alias of |
|
alias of |
|
alias of |
|
|
Layer normalization (https://arxiv.org/abs/1607.06450). |
|
Group normalization layer. |
|
Instance normalization layer. |
NVAR Layers#
|
Nonlinear vector auto-regression (NVAR) node. |
Pooling Layers#
|
Pools the input by taking the maximum over a window. |
|
Pools the input by taking the minimum over a window. |
|
Pools the input by taking the average over a window. |
|
Applies a 1D average pooling over an input signal composed of several input |
|
Applies a 2D average pooling over an input signal composed of several input |
|
Applies a 3D average pooling over an input signal composed of several input |
|
Applies a 1D max pooling over an input signal composed of several input |
|
Applies a 1D max pooling over an input signal composed of several input |
|
Applies a 1D max pooling over an input signal composed of several input |
|
Adaptive one-dimensional average down-sampling. |
|
Adaptive two-dimensional average down-sampling. |
|
Adaptive three-dimensional average down-sampling. |
|
Adaptive one-dimensional maximum down-sampling. |
|
Adaptive two-dimensional maximum down-sampling. |
|
Adaptive three-dimensional maximum down-sampling. |
Reservoir Layers#
|
Reservoir node, a pool of leaky-integrator neurons with random recurrent connections [1]_. |
Artificial Recurrent Layers#
|
Basic fully-connected RNN core. |
|
Gated Recurrent Unit. |
|
Long short-term memory (LSTM) RNN core. |
|
1-D convolutional LSTM. |
|
2-D convolutional LSTM. |
|
3-D convolutional LSTM. |