Struct GgufConfigInputs 

pub struct GgufConfigInputs<'a> {
    pub meta: &'a dyn GgufMeta,
    pub token_embd_vocab: Option<usize>,
    pub has_output_weight: bool,
}

2026-09-26: Inputs to config_from_gguf: the metadata reader plus two facts from the tensor section.

Fields

meta: &'a dyn GgufMeta

token_embd_vocab: Option<usize>

2026-09-26: Rows of token_embd.weight: the vocab size when {arch}.vocab_size is absent, and checked against it when present. None if the loader could not read the tensor shape.

has_output_weight: bool

2026-09-26: An output.weight tensor means an untied LM head; without one the head ties to the input embeddings.