pub struct DenseFfnWeightsQ2 {
pub gate_proj: PackedQ2Weight,
pub up_proj: PackedQ2Weight,
pub down_proj: PackedQ2Weight,
}Expand description
2026-09-25: Packed Q2_0 dense MLP weights (block_q2_0), kept 2-bit on the device. Once
installed with set_q2_weights, decode runs q2_0_gemv_vec per projection and prefill
runs the Q2_0 MMQ GEMM or a BF16 dequant (forward_prefill_inner).
Fields§
§gate_proj: PackedQ2Weight§up_proj: PackedQ2Weight§down_proj: PackedQ2WeightAuto Trait Implementations§
impl Freeze for DenseFfnWeightsQ2
impl RefUnwindSafe for DenseFfnWeightsQ2
impl Send for DenseFfnWeightsQ2
impl Sync for DenseFfnWeightsQ2
impl Unpin for DenseFfnWeightsQ2
impl UnwindSafe for DenseFfnWeightsQ2
Blanket Implementations§
Source§impl<T> BorrowMut<T> for Twhere
T: ?Sized,
impl<T> BorrowMut<T> for Twhere
T: ?Sized,
Source§fn borrow_mut(&mut self) -> &mut T
fn borrow_mut(&mut self) -> &mut T
Mutably borrows from an owned value. Read more