Struct HfTranslatingLoader 

pub struct HfTranslatingLoader<L: WeightLoader> { /* private fields */ }

Adapter that lets a HF-safetensors-backed WeightLoader satisfy requests phrased in GGUF-style names (blk.N.attn_q.weight etc.).

Builders like [Qwen35Weights::from_loader] address tensors using the GGUF / llama.cpp convention; the underlying safetensors file stores them under HF / PyTorch names (model.layers.N.self_attn.q_proj.weight). This wrapper:

1. Tries the requested key verbatim (in case it’s already-HF or the file was named GGUF-style).
2. Tries gguf_to_hf_name to translate the GGUF key → HF key.
3. Returns the underlying loader’s error otherwise.

Implementations

impl<L: WeightLoader> HfTranslatingLoader<L>

pub fn new(inner: L) -> Self

pub fn into_inner(self) -> L

pub fn inner(&self) -> &L

pub fn inner_mut(&mut self) -> &mut L

Trait Implementations

impl<L: WeightLoader> WeightLoader for HfTranslatingLoader<L>

fn format_id(&self) -> &'static str

Format id (safetensors, gguf, or a custom registration).

fn len(&self) -> usize

Number of distinct weights in the file.

fn take(&mut self, key: &str) -> Result<(Vec<f32>, Vec<usize>)>

Take the named tensor as (f32_data, shape). Removes from the loader so callers can detect “weights I never used.”

fn take_transposed(&mut self, key: &str) -> Result<(Vec<f32>, Vec<usize>)>

Same as take but transposed (last two dims swapped). Most safetensors weights are stored row-major-of-PyTorch convention, which RLX’s IR consumes column-major; this helper is the convention-bridge.

fn take_packed(&mut self, key: &str) -> Result<Option<PackedWeightTensor>>

Take packed K-quant bytes when supported; default returns None.

fn tensor_bytes_borrowed(&self, key: &str) -> Option<&[u8]>

Borrow packed bytes without marking taken (GGUF mmap path).

fn remaining_keys(&self) -> Vec<String>

Names that haven’t been taken yet — useful for “did the model use every weight?” hygiene checks.

fn is_empty(&self) -> bool

fn arch_hint(&self) -> Option<&str>

Architecture name from the underlying file (general.architecture for GGUF, None for safetensors). Drain-style consumers use this to pick an arch-specific reverse name mapping when the canonical HF name depends on the model family (e.g. Gemma 2’s 4 norms per layer don’t share the Llama 2-norm reverse alias).