RelativePositionEmbedding

class tfts.layers.position_layer.RelativePositionEmbedding(*args, **kwargs)

Bases: Layer

Relative position embedding layer that adds relative positional information.

This layer implements relative position embeddings as described in the paper “Self-Attention with Relative Position Representations” (Shaw et al., 2018). It learns embeddings for relative positions between query and key positions.

Parameters:

• max_len (int, optional) – Maximum sequence length. Defaults to 512.

• output_dim (int, optional) – Output embedding dimension. Defaults to 512.

• name (str, optional) – Layer name. Defaults to None.

Input shape:

• Tuple of two tensors:

  • Query tensor of shape (batch_size, query_length, embedding_dim)

  • Value tensor of shape (batch_size, value_length, embedding_dim)

Output shape:

• Tensor of shape (batch_size, query_length, value_length, output_dim)

Inherited-members:

Methods

add_loss(loss)	Can be called inside of the call() method to add a scalar loss.
add_metric(*args, **kwargs)
add_variable(shape, initializer[, dtype, ...])	Add a weight variable to the layer.
add_weight([shape, initializer, dtype, ...])	Add a weight variable to the layer.
build(input_shape)	Build the layer by creating the embedding weights.
build_from_config(config)	Builds the layer's states with the supplied config dict.
call(inputs)	Applies relative position embeddings to the input tensors.
compute_mask(inputs, previous_mask)
compute_output_shape(input_shape)
compute_output_spec(*args, **kwargs)
count_params()	Count the total number of scalars composing the weights.
from_config(config)	Creates an operation from its config.
get_build_config()	Returns a dictionary with the layer's input shape.
get_config()	Get layer configuration.
get_weights()	Return the values of layer.weights as a list of NumPy arrays.
load_own_variables(store)	Loads the state of the layer.
quantize(mode[, type_check])
quantized_build(input_shape, mode)
quantized_call(*args, **kwargs)
rematerialized_call(layer_call, *args, **kwargs)	Enable rematerialization dynamically for layer's call method.
save_own_variables(store)	Saves the state of the layer.
set_weights(weights)	Sets the values of layer.weights from a list of NumPy arrays.
stateless_call(trainable_variables, ...[, ...])	Call the layer without any side effects.
symbolic_call(*args, **kwargs)

Attributes

compute_dtype	The dtype of the computations performed by the layer.
dtype	Alias of layer.variable_dtype.
dtype_policy
input	Retrieves the input tensor(s) of a symbolic operation.
input_dtype	The dtype layer inputs should be converted to.
input_spec
losses	List of scalar losses from add_loss, regularizers and sublayers.
metrics	List of all metrics.
metrics_variables	List of all metric variables.
non_trainable_variables	List of all non-trainable layer state.
non_trainable_weights	List of all non-trainable weight variables of the layer.
output	Retrieves the output tensor(s) of a layer.
path	The path of the layer.
quantization_mode	The quantization mode of this layer, None if not quantized.
supports_masking	Whether this layer supports computing a mask using compute_mask.
trainable	Settable boolean, whether this layer should be trainable or not.
trainable_variables	List of all trainable layer state.
trainable_weights	List of all trainable weight variables of the layer.
variable_dtype	The dtype of the state (weights) of the layer.
variables	List of all layer state, including random seeds.
weights	List of all weight variables of the layer.

build(input_shape: Tuple[Tuple[int | None, ...], Tuple[int | None, ...]]) → None

Build the layer by creating the embedding weights.

Parameters:

input_shape – Shape of the input tensors

call(inputs: Tuple[Tensor, Tensor]) → Tensor

Applies relative position embeddings to the input tensors.

Parameters:

inputs – Tuple of (query_tensor, value_tensor) where: - query_tensor: Shape (batch_size, query_length, embedding_dim) - value_tensor: Shape (batch_size, value_length, embedding_dim)

Returns:

Tensor of shape (batch_size, query_length, value_length, output_dim) containing relative position embeddings.

get_config() → Dict[str, int | str]

Get layer configuration.

Returns:

Dictionary containing layer configuration.