Namespace marian::rnn

The namespace rnn.

Detailed Description

Declare class Dense and all the available functions for creating recurrent neural network (RNN) network.

Classes

• Struct AttentionNodeOp

• Struct GRUFastNodeOp

• Struct InputFactory

• Struct LSTMCellNodeOp

• Struct LSTMOutputNodeOp

• Struct State

• Class AttentionFactory

• Class Cell

• Class CellFactory

• Class CellInput

• Class FastLSTM

• Class GlobalAttention

• Class GRU

• Class GRUNematus

• Class MultiCellInput

• Template Class Multiplicative

• Class ReLU

• Class RNNFactory

• Class SlowLSTM

• Class SRU

• Class SSRU

• Class Stackable

• Class StackedCell

• Class StackedCellFactory

• Class States

• Class Tanh

• Class TestLSTM

Enums

• Enum dir

Functions

• Function marian::rnn::attOps

• Function marian::rnn::gruOps

• Function marian::rnn::lstmOpsC

• Function marian::rnn::lstmOpsO

Typedefs

• Typedef marian::rnn::Attention

• Typedef marian::rnn::attention

• Typedef marian::rnn::cell

• Typedef marian::rnn::LSTM

• Typedef marian::rnn::MGRU

• Typedef marian::rnn::MLSTM

• Typedef marian::rnn::rnn

• Typedef marian::rnn::StackableFactory

• Typedef marian::rnn::stacked_cell