# Module: tf.contrib.layers

<br />

|---|
| [![](https://www.tensorflow.org/images/GitHub-Mark-32px.png) View source on GitHub](https://github.com/tensorflow/tensorflow/blob/v1.15.0/tensorflow/contrib/layers/__init__.py) |

Ops for building neural network layers, regularizers, summaries, etc.

## Modules

[`feature_column`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/feature_column) module: This API defines FeatureColumn abstraction.

[`summaries`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/summaries) module: Utility functions for summary creation.

## Classes

[`class GDN`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/GDN): Generalized divisive normalization layer.

[`class RevBlock`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/RevBlock): Block of reversible layers. See rev_block.

## Functions

[`apply_regularization(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/apply_regularization): Returns the summed penalty by applying `regularizer` to the `weights_list`.

[`avg_pool2d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/avg_pool2d): Adds a 2D average pooling op.

[`avg_pool3d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/avg_pool3d): Adds a 3D average pooling op.

[`batch_norm(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/batch_norm): Adds a Batch Normalization layer from <http://arxiv.org/abs/1502.03167>

[`bias_add(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/bias_add): Adds a bias to the inputs.

[`bow_encoder(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/bow_encoder): Maps a sequence of symbols to a vector per example by averaging embeddings.

[`bucketized_column(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/bucketized_column): Creates a _BucketizedColumn for discretizing dense input.

[`check_feature_columns(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/check_feature_columns): Checks the validity of the set of FeatureColumns.

[`conv1d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv1d): Adds an N-D convolution followed by an optional batch_norm layer.

[`conv2d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv2d): Adds an N-D convolution followed by an optional batch_norm layer.

[`conv2d_in_plane(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv2d_in_plane): Performs the same in-plane convolution to each channel independently.

[`conv2d_transpose(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv2d_transpose): Adds a convolution2d_transpose with an optional batch normalization layer.

[`conv3d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv3d): Adds an N-D convolution followed by an optional batch_norm layer.

[`conv3d_transpose(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv3d_transpose): Adds a convolution3d_transpose with an optional batch normalization layer.

[`convolution(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/convolution): Adds an N-D convolution followed by an optional batch_norm layer.

[`convolution1d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv1d): Adds an N-D convolution followed by an optional batch_norm layer.

[`convolution2d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv2d): Adds an N-D convolution followed by an optional batch_norm layer.

[`convolution2d_in_plane(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv2d_in_plane): Performs the same in-plane convolution to each channel independently.

[`convolution2d_transpose(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv2d_transpose): Adds a convolution2d_transpose with an optional batch normalization layer.

[`convolution3d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv3d): Adds an N-D convolution followed by an optional batch_norm layer.

[`convolution3d_transpose(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/conv3d_transpose): Adds a convolution3d_transpose with an optional batch normalization layer.

[`create_feature_spec_for_parsing(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/create_feature_spec_for_parsing): Helper that prepares features config from input feature_columns.

[`crossed_column(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/crossed_column): Creates a _CrossedColumn for performing feature crosses.

[`dense_to_sparse(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/dense_to_sparse): Converts a dense tensor into a sparse tensor.

[`dropout(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/dropout): Returns a dropout op applied to the input.

[`elu(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/elu): partial(func, \*args, \*\*keywords) - new function with partial application

[`embed_sequence(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/embed_sequence): Maps a sequence of symbols to a sequence of embeddings.

[`embedding_column(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/embedding_column): Creates an `_EmbeddingColumn` for feeding sparse data into a DNN.

[`embedding_lookup_unique(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/embedding_lookup_unique): Version of embedding_lookup that avoids duplicate lookups.

[`flatten(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/flatten): Flattens the input while maintaining the batch_size.

[`fully_connected(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/fully_connected): Adds a fully connected layer.

[`gdn(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/gdn): Functional interface for GDN layer.

[`group_norm(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/group_norm): Functional interface for the group normalization layer.

[`images_to_sequence(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/images_to_sequence): Convert a batch of images into a batch of sequences.

[`infer_real_valued_columns(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/infer_real_valued_columns)

[`input_from_feature_columns(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/input_from_feature_columns): A tf.contrib.layers style input layer builder based on FeatureColumns.

[`instance_norm(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/instance_norm): Functional interface for the instance normalization layer.

[`joint_weighted_sum_from_feature_columns(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/joint_weighted_sum_from_feature_columns): A restricted linear prediction builder based on FeatureColumns.

[`l1_l2_regularizer(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/l1_l2_regularizer): Returns a function that can be used to apply L1 L2 regularizations.

[`l1_regularizer(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/l1_regularizer): Returns a function that can be used to apply L1 regularization to weights.

[`l2_regularizer(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/l2_regularizer): Returns a function that can be used to apply L2 regularization to weights.

[`layer_norm(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/layer_norm): Adds a Layer Normalization layer.

[`legacy_fully_connected(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/legacy_fully_connected): Adds the parameters for a fully connected layer and returns the output.

[`legacy_linear(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/legacy_linear): partial(func, \*args, \*\*keywords) - new function with partial application

[`legacy_relu(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/legacy_relu): partial(func, \*args, \*\*keywords) - new function with partial application

[`linear(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/linear): partial(func, \*args, \*\*keywords) - new function with partial application

[`make_place_holder_tensors_for_base_features(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/make_place_holder_tensors_for_base_features): Returns placeholder tensors for inference.

[`max_pool2d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/max_pool2d): Adds a 2D Max Pooling op.

[`max_pool3d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/max_pool3d): Adds a 3D Max Pooling op.

[`maxout(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/maxout): Adds a maxout op from <https://arxiv.org/abs/1302.4389>

[`multi_class_target(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/multi_class_target): Creates a _TargetColumn for multi class single label classification. (deprecated)

[`one_hot_column(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/one_hot_column): Creates an `_OneHotColumn` for a one-hot or multi-hot repr in a DNN.

[`one_hot_encoding(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/one_hot_encoding): Transform numeric labels into onehot_labels using [`tf.one_hot`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/one_hot).

[`optimize_loss(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/optimize_loss): Given loss and parameters for optimizer, returns a training op.

[`parse_feature_columns_from_examples(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/parse_feature_columns_from_examples): Parses tf.Examples to extract tensors for given feature_columns.

[`parse_feature_columns_from_sequence_examples(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/parse_feature_columns_from_sequence_examples): Parses tf.SequenceExamples to extract tensors for given `FeatureColumn`s.

[`real_valued_column(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/real_valued_column): Creates a `_RealValuedColumn` for dense numeric data.

[`recompute_grad(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/recompute_grad): Decorator that recomputes the function on the backwards pass.

[`regression_target(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/regression_target): Creates a _TargetColumn for linear regression. (deprecated)

[`relu(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/relu): partial(func, \*args, \*\*keywords) - new function with partial application

[`relu6(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/relu6): partial(func, \*args, \*\*keywords) - new function with partial application

[`repeat(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/repeat): Applies the same layer with the same arguments repeatedly.

[`rev_block(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/rev_block): A block of reversible residual layers.

[`safe_embedding_lookup_sparse(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/safe_embedding_lookup_sparse): Lookup embedding results, accounting for invalid IDs and empty features.

[`scale_gradient(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/scale_gradient): _OverloadedFunction encapsulates an overloaded function.

[`scattered_embedding_column(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/scattered_embedding_column): Creates an embedding column of a sparse feature using parameter hashing.

[`separable_conv2d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/separable_conv2d): Adds a depth-separable 2D convolution with optional batch_norm layer.

[`separable_convolution2d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/separable_conv2d): Adds a depth-separable 2D convolution with optional batch_norm layer.

[`sequence_input_from_feature_columns(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/sequence_input_from_feature_columns): Builds inputs for sequence models from `FeatureColumn`s. (experimental)

[`sequence_to_images(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/sequence_to_images): Convert a batch of sequences into a batch of images.

[`shared_embedding_columns(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/shared_embedding_columns): Creates a list of `_EmbeddingColumn` sharing the same embedding.

[`softmax(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/softmax): Performs softmax on Nth dimension of N-dimensional logit tensor.

[`sparse_column_with_hash_bucket(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/sparse_column_with_hash_bucket): Creates a _SparseColumn with hashed bucket configuration.

[`sparse_column_with_integerized_feature(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/sparse_column_with_integerized_feature): Creates an integerized _SparseColumn.

[`sparse_column_with_keys(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/sparse_column_with_keys): Creates a _SparseColumn with keys.

[`sparse_column_with_vocabulary_file(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/sparse_column_with_vocabulary_file): Creates a _SparseColumn with vocabulary file configuration.

[`spatial_softmax(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/spatial_softmax): Computes the spatial softmax of a convolutional feature map.

[`stack(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/stack): Builds a stack of layers by applying layer repeatedly using stack_args.

[`sum_regularizer(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/sum_regularizer): Returns a function that applies the sum of multiple regularizers.

[`summarize_activation(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/summarize_activation): Summarize an activation.

[`summarize_activations(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/summarize_activations): Summarize activations, using `summarize_activation` to summarize.

[`summarize_collection(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/summarize_collection): Summarize a graph collection of tensors, possibly filtered by name.

[`summarize_tensor(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/summarize_tensor): Summarize a tensor using a suitable summary type.

[`summarize_tensors(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/summarize_tensors): Summarize a set of tensors.

[`transform_features(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/transform_features): Returns transformed features based on features columns passed in.

[`unit_norm(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/unit_norm): Normalizes the given input across the specified dimension to unit length.

[`variance_scaling_initializer(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/variance_scaling_initializer): Returns an initializer that generates tensors without scaling variance.

[`weighted_sparse_column(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/weighted_sparse_column): Creates a _SparseColumn by combining sparse_id_column with a weight column.

[`weighted_sum_from_feature_columns(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/weighted_sum_from_feature_columns): A tf.contrib.layers style linear prediction builder based on FeatureColumn.

[`xavier_initializer(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/xavier_initializer): Returns an initializer performing "Xavier" initialization for weights.

[`xavier_initializer_conv2d(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/contrib/layers/xavier_initializer): Returns an initializer performing "Xavier" initialization for weights.

## Other Members

- `OPTIMIZER_CLS_NAMES`
- `OPTIMIZER_SUMMARIES`
- `SPARSE_FEATURE_CROSS_DEFAULT_HASH_KEY = 956888297470`