tfa.seq2seq.BahdanauAttention

Implements Bahdanau-style (additive) attention.

Inherits From: AttentionMechanism

Used in the notebooks

Used in the tutorials

This attention has two forms. The first is Bahdanau attention, as described in:

Dzmitry Bahdanau, Kyunghyun Cho, Yoshua Bengio. "Neural Machine Translation by Jointly Learning to Align and Translate." ICLR 2015. https://arxiv.org/abs/1409.0473

The second is the normalized form. This form is inspired by the weight normalization article:

Tim Salimans, Diederik P. Kingma. "Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks." https://arxiv.org/abs/1602.07868

To enable the second form, construct the object with parameter normalize=True.

units The depth of the query mechanism.
memory The memory to query; usually the output of an RNN encoder. This tensor should be shaped [batch_size, max_time, ...].
memory_sequence_length (optional): Sequence lengths for the batch entries in memory. If provided, the memory tensor rows are masked with zeros for values past the respective sequence lengths.
normalize Python boolean. Whether to normalize the energy term.
probability_fn (optional) string, the name of function to convert the attention score to probabilities. The default is softmax which is tf.nn.softmax. Other options is hardmax, which is hardmax() within this module. Any other value will result into validation error. Default to use softmax.
kernel_initializer (optional), the name of the initializer for the attention kernel.
dtype The data type for the query and memory layers of the attention mechanism.
name Name to use when creating ops.
**kwargs Dictionary that contains other common arguments for layer creation.

activity_regularizer Optional regularizer function for the output of this layer.
alignments_size

compute_dtype The dtype of the layer's computations.

This is equivalent to Layer.dtype_policy.compute_dtype. Unless mixed precision is used, this is the same as Layer.dtype, the dtype of the weights.

Layers automatically cast their inputs to the compute dtype, which causes computations and the output to be in the compute dtype as well. This is done by the base Layer class in Layer.call, so you do not have to insert these casts if implementing your own layer.

Layers often perform certain internal computations in higher precision when compute_dtype is float16 or bfloat16 for numeric stability. The output will still typically be float16 or bfloat16 in such cases.

dtype The dtype of the layer weights.

This is equivalent to Layer.dtype_policy.variable_dtype. Unless mixed precision is used, this is the same as Layer.compute_dtype, the dtype of the layer's computations.

dtype_policy The dtype policy associated with this layer.

This is an instance of a tf.keras.mixed_precision.Policy.

dynamic Whether the layer is dynamic (eager-only); set in the constructor.
input Retrieves the input tensor(s) of a layer.

Only applicable if the layer has exactly one input, i.e. if it is connected to one incoming layer.

input_spec InputSpec instance(s) describing the input format for this layer.

When you create a layer subclass, you can set self.input_spec to enable the layer to run input compatibility checks when it is called. Consider a Conv2D layer: it can only be called on a single input tensor of rank 4. As such, you can set, in __init__():

self.input_spec = tf.keras.layers.InputSpec(ndim=4)

Now, if you try to call the layer on an input that isn't rank 4 (for instance, an input of shape (2,), it will raise a nicely-formatted error:

ValueError: Input 0 of layer conv2d is incompatible with the layer:
expected ndim=4, found ndim=1. Full shape received: [2]

Input checks that can be specified via input_spec include:

  • Structure (e.g. a single input, a list of 2 inputs, etc)
  • Shape
  • Rank (ndim)
  • Dtype

For more information, see tf.keras.layers.InputSpec.

losses List of losses added using the add_loss() API.

Variable regularization tensors are created when this property is accessed, so it is eager safe: accessing losses under a tf.GradientTape will propagate gradients back to the corresponding variables.

class MyLayer(tf.keras.layers.Layer):
  def call(self, inputs):
    self.add_loss(tf.abs(tf.reduce_mean(inputs)))
    return inputs
l = MyLayer()
l(np.ones((10, 1)))
l.losses
[1.0]
inputs = tf.keras.Input(shape=(10,))
x = tf.keras.layers.Dense(10)(inputs)
outputs = tf.keras.layers.Dense(1)(x)
model = tf.keras.Model(inputs, outputs)
# Activity regularization.
len(model.losses)
0
model.add_loss(tf.abs(tf.reduce_mean(x)))
len(model.losses)
1
inputs = tf.keras.Input(shape=(10,))
d = tf.keras.layers.Dense(10, kernel_initializer='ones')
x = d(inputs)
outputs = tf.keras.layers.Dense(1)(x)
model = tf.keras.Model(inputs, outputs)
# Weight regularization.
model.add_loss(lambda: tf.reduce_mean(d.kernel))
model.losses
[<tf.Tensor: shape=(), dtype=float32, numpy=1.0>]

memory_initialized Returns True if this attention mechanism has been initialized with a memory.
metrics List of metrics added using the add_metric() API.

input = tf.keras.layers.Input(shape=(3,))
d = tf.keras.layers.Dense(2)
output = d(input)
d.add_metric(tf.reduce_max(output), name='max')
d.add_metric(tf.reduce_min(output), name='min')
[m.name for m in d.metrics]
['max', 'min']

name Name of the layer (string), set in the constructor.
name_scope Returns a tf.name_scope instance for this class.
non_trainable_weights List of all non-trainable weights tracked by this layer.

Non-trainable weights are not updated during training. They are expected to be updated manually in call().

output Retrieves the output tensor(s) of a layer.

Only applicable if the layer has exactly one output, i.e. if it is connected to one incoming layer.

state_size

submodules Sequence of all sub-modules.

Submodules are modules which are properties of this module, or found as properties of modules which are properties of this module (and so on).

a = tf.Module()
b = tf.Module()
c = tf.Module()
a.b = b
b.c = c
list(a.submodules) == [b, c]
True
list(b.submodules) == [c]
True
list(c.submodules) == []
True

supports_masking Whether this layer supports computing a mask using compute_mask.
trainable

trainable_weights List of all trainable weights tracked by this layer.

Trainable weights are updated via gradient descent during training.

variable_dtype Alias of Layer.dtype, the dtype of the weights.
weights Returns the list of all layer variables/weights.

Methods

add_loss

Add loss tensor(s), potentially dependent on layer inputs.

Some losses (for instance, activity regularization losses) may be dependent on the inputs passed when calling a layer. Hence, when reusing the same layer on different inputs a and b, some entries in layer.losses may be dependent on a and some on b. This method automatically keeps track of dependencies.

This method can be used inside a subclassed layer or model's call function, in which case losses should be a Tensor or list of Tensors.

Example:

class MyLayer(tf.keras.layers.Layer):
  def call(self, inputs):
    self.add_loss(tf.abs(tf.reduce_mean(inputs)))
    return inputs

The same code works in distributed training: the input to add_loss() is treated like a regularization loss and averaged across replicas by the training loop (both built-in Model.fit() and compliant custom training loops).

The add_loss method can also be called directly on a Functional Model during construction. In this case, any loss Tensors passed to this Model must be symbolic and be able to be traced back to the model's Inputs. These losses become part of the model's topology and are tracked in get_config.

Example:

inputs = tf.keras.Input(shape=(10,))
x = tf.keras.layers.Dense(10)(inputs)
outputs = tf.keras.layers.Dense(1)(x)
model = tf.keras.Model(inputs, outputs)
# Activity regularization.
model.add_loss(tf.abs(tf.reduce_mean(x)))

If this is not the case for your loss (if, for example, your loss references a Variable of one of the model's layers), you can wrap your loss in a zero-argument lambda. These losses are not tracked as part of the model's topology since they can't be serialized.

Example:

inputs = tf.keras.Input(shape=(10,))
d = tf.keras.layers.Dense(10)
x = d(inputs)
outputs = tf.keras.layers.Dense(1)(x)
model = tf.keras.Model(inputs, outputs)
# Weight regularization.
model.add_loss(lambda: tf.reduce_mean(d.kernel))

Args
losses Loss tensor, or list/tuple of tensors. Rather than tensors, losses may also be zero-argument callables which create a loss tensor.
**kwargs Used for backwards compatibility only.

add_metric

Adds metric tensor to the layer.

This method can be used inside the call() method of a subclassed layer or model.

class MyMetricLayer(tf.keras.layers.Layer):
  def __init__(self):
    super(MyMetricLayer, self).__init__(name='my_metric_layer')
    self.mean = tf.keras.metrics.Mean(name='metric_1')

  def call(self, inputs):
    self.add_metric(self.mean(inputs))
    self.add_metric(tf.reduce_sum(inputs), name='metric_2')
    return inputs

This method can also be called directly on a Functional Model during construction. In this case, any tensor passed to this Model must be symbolic and be able to be traced back to the model's Inputs. These metrics become part of the model's topology and are tracked when you save the model via save().

inputs = tf.keras.Input(shape=(10,))
x = tf.keras.layers.Dense(10)(inputs)
outputs = tf.keras.layers.Dense(1)(x)
model = tf.keras.Model(inputs, outputs)
model.add_metric(math_ops.reduce_sum(x), name='metric_1')
inputs = tf.keras.Input(shape=(10,))
x = tf.keras.layers.Dense(10)(inputs)
outputs = tf.keras.layers.Dense(1)(x)
model = tf.keras.Model(inputs, outputs)
model.add_metric(tf.keras.metrics.Mean()(x), name='metric_1')

Args
value Metric tensor.
name String metric name.
**kwargs Additional keyword arguments for backward compatibility. Accepted values: aggregation - When the value tensor provided is not the result of calling a keras.Metric instance, it will be aggregated by default using a keras.Metric.Mean.

build

View source

Creates the variables of the layer (for subclass implementers).

This is a method that implementers of subclasses of Layer or Model can override if they need a state-creation step in-between layer instantiation and layer call. It is invoked automatically before the first execution of call().

This is typically used to create the weights of Layer subclasses (at the discretion of the subclass implementer).

Args
input_shape Instance of TensorShape, or list of instances of TensorShape if the layer expects a list of inputs (one instance per input).

build_from_config

compute_mask

View source

Computes an output mask tensor.

Args
inputs Tensor or list of tensors.
mask Tensor or list of tensors.

Returns
None or a tensor (or list of tensors, one per output tensor of the layer).

compute_output_shape

Computes the output shape of the layer.

This method will cause the layer's state to be built, if that has not happened before. This requires that the layer will later be used with inputs that match the input shape provided here.

Args
input_shape Shape tuple (tuple of integers) or tf.TensorShape, or structure of shape tuples / tf.TensorShape instances (one per output tensor of the layer). Shape tuples can include None for free dimensions, instead of an integer.

Returns
A tf.TensorShape instance or structure of tf.TensorShape instances.

count_params

Count the total number of scalars composing the weights.

Returns
An integer count.

Raises
ValueError if the layer isn't yet built (in which case its weights aren't yet defined).

deserialize_inner_layer_from_config

View source

Helper method that reconstruct the query and memory from the config.

In the get_config() method, the query and memory layer configs are serialized into dict for persistence, this method perform the reverse action to reconstruct the layer from the config.

Args
config dict, the configs that will be used to reconstruct the object.
custom_objects dict mapping class names (or function names) of custom (non-Keras) objects to class/functions.

Returns
config dict, the config with layer instance created, which is ready to be used as init parameters.

from_config

View source

Creates a layer from its config.

This method is the reverse of get_config, capable of instantiating the same layer from the config dictionary. It does not handle layer connectivity (handled by Network), nor weights (handled by set_weights).

Args
config A Python dictionary, typically the output of get_config.

Returns
A layer instance.

get_build_config

get_config

View source

Returns the config of the layer.

A layer config is a Python dictionary (serializable) containing the configuration of a layer. The same layer can be reinstantiated later (without its trained weights) from this configuration.

The config of a layer does not include connectivity information, nor the layer class name. These are handled by Network (one layer of abstraction above).

Note that get_config() does not guarantee to return a fresh copy of dict every time it is called. The callers should make a copy of the returned dict if they want to modify it.

Returns
Python dictionary.

get_weights

Returns the current weights of the layer, as NumPy arrays.

The weights of a layer represent the state of the layer. This function returns both trainable and non-trainable weight values associated with this layer as a list of NumPy arrays, which can in turn be used to load state into similarly parameterized layers.

For example, a Dense layer returns a list of two values: the kernel matrix and the bias vector. These can be used to set the weights of another Dense layer:

layer_a = tf.keras.layers.Dense(1,
  kernel_initializer=tf.constant_initializer(1.))
a_out = layer_a(tf.convert_to_tensor([[1., 2., 3.]]))
layer_a.get_weights()
[array([[1.],
       [1.],
       [1.]], dtype=float32), array([0.], dtype=float32)]
layer_b = tf.keras.layers.Dense(1,
  kernel_initializer=tf.constant_initializer(2.))
b_out = layer_b(tf.convert_to_tensor([[10., 20., 30.]]))
layer_b.get_weights()
[array([[2.],
       [2.],
       [2.]], dtype=float32), array([0.], dtype=float32)]
layer_b.set_weights(layer_a.get_weights())
layer_b.get_weights()
[array([[1.],
       [1.],
       [1.]], dtype=float32), array([0.], dtype=float32)]

Returns
Weights values as a list of NumPy arrays.

initial_alignments

View source

Creates the initial alignment values for the tfa.seq2seq.AttentionWrapper class.

This is important for attention mechanisms that use the previous alignment to calculate the alignment at the next time step (e.g. monotonic attention).

The default behavior is to return a tensor of all zeros.

Args
batch_size int32 scalar, the batch_size.
dtype The dtype.

Returns
A dtype tensor shaped [batch_size, alignments_size] (alignments_size is the values' max_time).

initial_state

View source

Creates the initial state values for the tfa.seq2seq.AttentionWrapper class.

This is important for attention mechanisms that use the previous alignment to calculate the alignment at the next time step (e.g. monotonic attention).

The default behavior is to return the same output as initial_alignments.

Args
batch_size int32 scalar, the batch_size.
dtype The dtype.

Returns
A structure of all-zero tensors with shapes as described by state_size.

set_weights

Sets the weights of the layer, from NumPy arrays.

The weights of a layer represent the state of the layer. This function sets the weight values from numpy arrays. The weight values should be passed in the order they are created by the layer. Note that the layer's weights must be instantiated before calling this function, by calling the layer.

For example, a Dense layer returns a list of two values: the kernel matrix and the bias vector. These can be used to set the weights of another Dense layer:

layer_a = tf.keras.layers.Dense(1,
  kernel_initializer=tf.constant_initializer(1.))
a_out = layer_a(tf.convert_to_tensor([[1., 2., 3.]]))
layer_a.get_weights()
[array([[1.],
       [1.],
       [1.]], dtype=float32), array([0.], dtype=float32)]
layer_b = tf.keras.layers.Dense(1,
  kernel_initializer=tf.constant_initializer(2.))
b_out = layer_b(tf.convert_to_tensor([[10., 20., 30.]]))
layer_b.get_weights()
[array([[2.],
       [2.],
       [2.]], dtype=float32), array([0.], dtype=float32)]
layer_b.set_weights(layer_a.get_weights())
layer_b.get_weights()
[array([[1.],
       [1.],
       [1.]], dtype=float32), array([0.], dtype=float32)]

Args
weights a list of NumPy arrays. The number of arrays and their shape must match number of the dimensions of the weights of the layer (i.e. it should match the output of get_weights).

Raises
ValueError If the provided weights list does not match the layer's specifications.

setup_memory

View source

Pre-process the memory before actually query the memory.

This should only be called once at the first invocation of call().

Args
memory The memory to query; usually the output of an RNN encoder. This tensor should be shaped [batch_size, max_time, ...].
memory_sequence_length optional

Sequence lengths for the batch entries in memory. If provided, the memory tensor rows are masked with zeros for values past the respective sequence lengths.

memory_mask (Optional) The boolean tensor with shape [batch_size, max_time]. For any value equal to False, the corresponding value in memory should be ignored.

with_name_scope

Decorator to automatically enter the module name scope.

class MyModule(tf.Module):
  @tf.Module.with_name_scope
  def __call__(self, x):
    if not hasattr(self, 'w'):
      self.w = tf.Variable(tf.random.normal([x.shape[1], 3]))
    return tf.matmul(x, self.w)

Using the above module would produce tf.Variables and tf.Tensors whose names included the module name:

mod = MyModule()
mod(tf.ones([1, 2]))
<tf.Tensor: shape=(1, 3), dtype=float32, numpy=..., dtype=float32)>
mod.w
<tf.Variable 'my_module/Variable:0' shape=(2, 3) dtype=float32,
numpy=..., dtype=float32)>

Args
method The method to wrap.

Returns
The original method wrapped such that it enters the module's name scope.

__call__

View source

Preprocess the inputs before calling base_layer.__call__().

Note that there are situation here, one for setup memory, and one with actual query and state.

  1. When the memory has not been configured, we just pass all the param to base_layer.__call__(), which will then invoke self.call() with proper inputs, which allows this class to setup memory.
  2. When the memory has already been setup, the input should contain query and state, and optionally processed memory. If the processed memory is not included in the input, we will have to append it to the inputs and give it to the base_layer.__call__(). The processed memory is the output of first invocation of self.__call__(). If we don't add it here, then from keras perspective, the graph is disconnected since the output from previous call is never used.

Args
inputs the inputs tensors.
**kwargs dict, other keyeword arguments for the __call__()