RNN cell that is composed by applying a sequence of RNN cells in order.
This means that the output of each RNN is fed to the next one as input, while the states remain separate.
Note that this class does no variable management at all. Variable sharing should be handled based on the RNN cells
the caller provides to this class. The learn API provides a layer version of this class that also does some
management of the variables involved.
RNN cell that is composed by applying a sequence of RNN cells in order.
This means that the output of each RNN is fed to the next one as input, while the states remain separate.
Note that this class does no variable management at all. Variable sharing should be handled based on the RNN cells the caller provides to this class. The learn API provides a layer version of this class that also does some management of the variables involved.