Struct ModelSizeConfig 

pub struct ModelSizeConfig {
    pub parameters: String,
    pub hidden_dim: usize,
    pub num_layers: usize,
    pub num_heads: usize,
    pub num_kv_heads: usize,
    pub intermediate_dim: usize,
    pub vocab_size: usize,
    pub max_position_embeddings: usize,
    pub head_dim: usize,
    pub rope_theta: f64,
    pub norm_eps: f64,
}

Configuration for a specific model size within a family.

Fields

parameters: String

Human-readable parameter count (e.g., “0.5B”, “7B”)

hidden_dim: usize

Hidden dimension

num_layers: usize

Number of transformer layers

num_heads: usize

Number of attention heads

num_kv_heads: usize

Number of key-value heads (for GQA)

intermediate_dim: usize

Intermediate (FFN) dimension

vocab_size: usize

Vocabulary size

max_position_embeddings: usize

Maximum position embeddings

head_dim: usize

Per-head dimension (hidden_dim / num_heads)

rope_theta: f64

RoPE theta frequency (0.0 if not using RoPE)

norm_eps: f64

Normalization epsilon

Trait Implementations

impl Clone for ModelSizeConfig

fn clone(&self) -> ModelSizeConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ModelSizeConfig

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

Formats the value using the given formatter.