Recurrent Cells

Recurrent neural network cells for sequential data processing and temporal modeling. Includes vanilla RNN, gated recurrent units (GRU), minimal gated units (MGU), long short-term memory (LSTM), and unbalanced LSTM variants. Each cell maintains internal state across time steps for memory-dependent computations.

RNNCell	Base class for all recurrent neural network (RNN) cell implementations.
ValinaRNNCell	Vanilla Recurrent Neural Network (RNN) cell implementation.
GRUCell	Gated Recurrent Unit (GRU) cell implementation.
MGUCell	Minimal Gated Unit (MGU) cell implementation.
LSTMCell	Long Short-Term Memory (LSTM) cell implementation.
URLSTMCell	LSTM with UR gating mechanism.