synax

This is the public API.

Classes

Attention	Attention.
Bias	Bias (translation).
Conv	Convolution.
Embed	Embedding.
Func	Function application.
Linear	Linear map.
Scale	Elementwise scaling.
DropConnect	DropConnect.
Chain	Sequential composition.
Parallel	Parallel composition.
Residual	Residual map.
Switch	Switch map.
GLU	Gated linear unit.
AutoEncoder	Auto-encoder.
PReLU	Parametric ReLU.
GRU	Gated recurrent unit.
LSTM	Long short term memory.
MGU	Minimal gated unit.
SimpleRNN	Simple recurrent unit.
AntiSymmetricMatrix	Symmetric matrix.
SpecialOrthogonalMatrix	Special orthogonal matrix
SymmetricMatrix	Symmetric matrix.
Constant	Constant.

Functions

MLP(, linear_initializer, bias_initializer, ...)	Multi-layer perceptron.
AlexNet(→ Module)	AlexNet convolutional network.
LeNet(→ Module)	LeNet convolutional network.
layer_norm(→ Callable[[jax.Array], jax.Array])	Layer normalization.
max_pool(→ Callable[[jax.Array], jax.Array])	Max pooling.
mean_pool(→ Callable[[jax.Array], jax.Array])	Mean pooling.
rms_norm(→ Callable[[jax.Array], jax.Array])	Root mean square layer normalization.