Struct LoRAEngine 

pub struct LoRAEngine {
    pub micro: MicroLoRA,
    pub base: BaseLoRA,
    pub micro_enabled: bool,
    pub base_enabled: bool,
}

Combined LoRA engine managing both tiers

Fields

micro: MicroLoRA

Micro-LoRA for instant adaptation

base: BaseLoRA

Base LoRA for background adaptation

micro_enabled: bool

Whether micro-LoRA is enabled

base_enabled: bool

Whether base LoRA is enabled

Implementations

impl LoRAEngine

pub fn new( hidden_dim: usize, micro_rank: usize, base_rank: usize, num_layers: usize, ) -> Self

Create new LoRA engine

pub fn forward(&self, layer_idx: usize, input: &[f32], output: &mut [f32])

Apply both LoRA tiers

pub fn accumulate_micro(&mut self, signal: &LearningSignal)

Accumulate micro-LoRA gradient

pub fn apply_micro(&mut self, learning_rate: f32)

Apply micro-LoRA updates

Trait Implementations

impl Clone for LoRAEngine

fn clone(&self) -> LoRAEngine

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for LoRAEngine

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

Formats the value using the given formatter.