pub struct LinearAttentionState {
pub conv1d_state: DevicePtr,
pub gdn_state: DevicePtr,
}Expand description
Per-layer SSM/conv state for a linear-attention layer. Persists across tokens. Caller owns alloc + zero-init.
Fields§
§conv1d_state: DevicePtrFP32 [QKV_TOTAL_LIN, conv_kernel_size].
gdn_state: DevicePtrFP32 [batch=1, num_v_heads_lin, k_head_dim_lin, v_head_dim_lin].
Auto Trait Implementations§
impl Freeze for LinearAttentionState
impl RefUnwindSafe for LinearAttentionState
impl Send for LinearAttentionState
impl Sync for LinearAttentionState
impl Unpin for LinearAttentionState
impl UnwindSafe for LinearAttentionState
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