Module forward 

2026-09-25: Per-layer decode forwards written against GpuBackend and the QuantWeights trait, so a driver runs a layer without knowing the weight format.

Owner: model-layers. Invariants: none beyond the types.

The only caller is the model-arch example metal_qwen35_inference. The served decode path does not use this module; it lives in model-engine (model/trait_impl/decode_a.rs).

Re-exports

pub use quant_weights::QuantWeights;

Modules

quant_weights

2026-09-25: The QuantWeights trait: the matvec operations a per-layer forward needs from a quantised weight, so super::qwen3_5 never names a concrete weight type.

qwen3_5

2026-09-25: Single-token decode forward for one Qwen3.5 decoder layer, full attention or GDN, through &dyn GpuBackend and a QuantWeights impl.