# Module: tf.contrib.rnn
|---| | [![](https://www.tensorflow.org/images/GitHub-Mark-32px.png) View source on GitHub](https://github.com/tensorflow/tensorflow/blob/v1.15.0/tensorflow/contrib/rnn/__init__.py) | RNN Cells and additional RNN operations. ## Classes [`class AttentionCellWrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/AttentionCellWrapper): Basic attention cell wrapper. [`class BasicLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/BasicLSTMCell): DEPRECATED: Please use [`tf.compat.v1.nn.rnn_cell.LSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/LSTMCell) instead. [`class BasicRNNCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/BasicRNNCell): The most basic RNN cell. [`class BidirectionalGridLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/BidirectionalGridLSTMCell): Bidirectional GridLstm cell. [`class CompiledWrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/CompiledWrapper): Wraps step execution in an XLA JIT scope. [`class Conv1DLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/Conv1DLSTMCell): 1D Convolutional LSTM recurrent network cell. [`class Conv2DLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/Conv2DLSTMCell): 2D Convolutional LSTM recurrent network cell. [`class Conv3DLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/Conv3DLSTMCell): 3D Convolutional LSTM recurrent network cell. [`class ConvLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/ConvLSTMCell): Convolutional LSTM recurrent network cell. [`class CoupledInputForgetGateLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/CoupledInputForgetGateLSTMCell): Long short-term memory unit (LSTM) recurrent network cell. [`class DeviceWrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/DeviceWrapper): Operator that ensures an RNNCell runs on a particular device. [`class DropoutWrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/DropoutWrapper): Operator adding dropout to inputs and outputs of the given cell. [`class EmbeddingWrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/EmbeddingWrapper): Operator adding input embedding to the given cell. [`class FusedRNNCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/FusedRNNCell): Abstract object representing a fused RNN cell. [`class FusedRNNCellAdaptor`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/FusedRNNCellAdaptor): This is an adaptor for RNNCell classes to be used with `FusedRNNCell`. [`class GLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/GLSTMCell): Group LSTM cell (G-LSTM). [`class GRUBlockCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/GRUBlockCell): Block GRU cell implementation. [`class GRUBlockCellV2`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/GRUBlockCellV2): Temporary GRUBlockCell impl with a different variable naming scheme. [`class GRUCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/GRUCell): Gated Recurrent Unit cell (cf. [`class GridLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/GridLSTMCell): Grid Long short-term memory unit (LSTM) recurrent network cell. [`class HighwayWrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/HighwayWrapper): RNNCell wrapper that adds highway connection on cell input and output. [`class IndRNNCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/IndRNNCell): Independently Recurrent Neural Network (IndRNN) cell [`class IndyGRUCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/IndyGRUCell): Independently Gated Recurrent Unit cell. [`class IndyLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/IndyLSTMCell): Basic IndyLSTM recurrent network cell. [`class InputProjectionWrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/InputProjectionWrapper): Operator adding an input projection to the given cell. [`class IntersectionRNNCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/IntersectionRNNCell): Intersection Recurrent Neural Network (+RNN) cell. [`class LSTMBlockCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/LSTMBlockCell): Basic LSTM recurrent network cell. [`class LSTMBlockFusedCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/LSTMBlockFusedCell): FusedRNNCell implementation of LSTM. [`class LSTMBlockWrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/LSTMBlockWrapper): This is a helper class that provides housekeeping for LSTM cells. [`class LSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/LSTMCell): Long short-term memory unit (LSTM) recurrent network cell. [`class LSTMStateTuple`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/LSTMStateTuple): Tuple used by LSTM Cells for `state_size`, `zero_state`, and output state. [`class LayerNormBasicLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/LayerNormBasicLSTMCell): LSTM unit with layer normalization and recurrent dropout. [`class LayerRNNCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/LayerRNNCell): Subclass of RNNCells that act like proper `tf.Layer` objects. [`class MultiRNNCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/MultiRNNCell): RNN cell composed sequentially of multiple simple cells. [`class NASCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/NASCell): Neural Architecture Search (NAS) recurrent network cell. [`class OutputProjectionWrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/OutputProjectionWrapper): Operator adding an output projection to the given cell. [`class PhasedLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/PhasedLSTMCell): Phased LSTM recurrent network cell. [`class RNNCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/RNNCell): Abstract object representing an RNN cell. [`class ResidualWrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/rnn_cell/ResidualWrapper): RNNCell wrapper that ensures cell inputs are added to the outputs. [`class SRUCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/SRUCell): SRU, Simple Recurrent Unit. [`class TimeFreqLSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/TimeFreqLSTMCell): Time-Frequency Long short-term memory unit (LSTM) recurrent network cell. [`class TimeReversedFusedRNN`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/TimeReversedFusedRNN): This is an adaptor to time-reverse a FusedRNNCell. [`class UGRNNCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/UGRNNCell): Update Gate Recurrent Neural Network (UGRNN) cell. ## Functions [`best_effort_input_batch_size(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/best_effort_input_batch_size): Get static input batch size if available, with fallback to the dynamic one. [`stack_bidirectional_dynamic_rnn(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/stack_bidirectional_dynamic_rnn): Creates a dynamic bidirectional recurrent neural network. [`stack_bidirectional_rnn(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/stack_bidirectional_rnn): Creates a bidirectional recurrent neural network. [`static_bidirectional_rnn(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/static_bidirectional_rnn): Creates a bidirectional recurrent neural network. (deprecated) [`static_rnn(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/static_rnn): Creates a recurrent neural network specified by RNNCell `cell`. (deprecated) [`static_state_saving_rnn(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/nn/static_state_saving_rnn): RNN that accepts a state saver for time-truncated RNN calculation. (deprecated) [`transpose_batch_time(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/rnn/transpose_batch_time): Transposes the batch and time dimensions of a Tensor.