Module recurrent 

Recurrent layer implementations

Re-exports

pub use bidirectional::Bidirectional;

pub use gru::GRU;

pub use lstm::LSTM;

pub use rnn::RNN;

Modules

bidirectional

Bidirectional wrapper for recurrent layers

gru

Gated Recurrent Unit (GRU) implementation

lstm

Long Short-Term Memory (LSTM) implementation

rnn

Basic Recurrent Neural Network (RNN) implementation

Type Aliases

GruForwardOutput

Type alias for GRU forward output (new_h, (reset_gate, update_gate, new_gate))

GruGateCache

Type alias for GRU gate cache (reset, update, new gates)

GruStateCache

Type alias for GRU state cache

LstmGateCache

Type alias for LSTM gate cache (input, forget, output, cell gates)

LstmStateCache

Type alias for LSTM state cache (hidden, cell)

LstmStepOutput

Type alias for LSTM step output (new_h, new_c, (input_gate, forget_gate, cell_gate, output_gate))

RnnStateCache

Type alias for RNN state cache