Struct LearnedPositionalEncoding 

pub struct LearnedPositionalEncoding<F: Float + Debug + NumAssign> { /* private fields */ }

Learned Positional Encoding

Uses trainable embeddings for each position, similar to BERT. More flexible than sinusoidal but requires training data and doesn’t extrapolate to longer sequences.

Implementations

impl<F: Float + Debug + NumAssign> LearnedPositionalEncoding<F>

pub fn new<R: Rng>(d_model: usize, max_len: usize, rng: &mut R) -> Self

Create a new learned positional encoding with random initialization

Arguments

• d_model - Model dimension
• max_len - Maximum sequence length
• rng - Random number generator

pub fn from_embeddings(embeddings: Array2<F>) -> Self

Create from existing embeddings

pub fn embeddings_mut(&mut self) -> &mut Array2<F>

Get mutable reference to embeddings for training

pub fn embeddings(&self) -> &Array2<F>

Get reference to embeddings

Trait Implementations

impl<F: Clone + Float + Debug + NumAssign> Clone for LearnedPositionalEncoding<F>

fn clone(&self) -> LearnedPositionalEncoding<F>

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl<F: Debug + Float + Debug + NumAssign> Debug for LearnedPositionalEncoding<F>

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<F: Float + Debug + NumAssign> PositionalEncoding<F> for LearnedPositionalEncoding<F>

fn encode(&self, seq_len: usize) -> Array2<F>

Encode positions for a sequence of given length

fn apply(&self, input: &Array3<F>) -> Result<Array3<F>>

Apply positional encoding to an input tensor

fn clone_box(&self) -> Box<dyn PositionalEncoding<F> + Send + Sync>

where F: Send + Sync + 'static,

Clone the positional encoding into a boxed trait object

fn d_model(&self) -> usize

Get the model dimension

fn max_len(&self) -> usize

Get the maximum sequence length supported

fn forward(&self, input: &Array<F, IxDyn>) -> Result<Array<F, IxDyn>>

Forward pass with dynamic array

fn update(&mut self, _learning_rate: F) -> Result<()>

Update trainable parameters (for learned encodings)