Struct MultiHeadAttentionLayer 

pub struct MultiHeadAttentionLayer {
    pub mha: MultiHeadAttention,
    /* private fields */
}

Multi-head attention layer wrapping MultiHeadAttention.

Self-attention: Q=K=V=input. Input/output: [seq_len, d_model].

Fields

mha: MultiHeadAttention

Implementations

impl MultiHeadAttentionLayer

pub fn w_q_node(&self) -> Option<NodeId>

pub fn new(d_model: usize, num_heads: usize, _seed: u64) -> Self

pub fn register_params(&mut self, graph: &mut Graph)

pub fn forward( &self, graph: &mut Graph, input: NodeId, ) -> Result<NodeId, ModelError>

pub fn forward_inference(&self, input: &Tensor) -> Result<Tensor, ModelError>

Trait Implementations

impl Clone for MultiHeadAttentionLayer

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for MultiHeadAttentionLayer

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

Formats the value using the given formatter.