pub struct LayerwiseTraversalPartitionExecutor<A, B, S, P>where
B: SubmissionBackend<Executor = <<B as NeuralBackend>::Tensor as Tensor>::Context>,
S: RuntimeState<B>,
A: LayeredArchitecture<B, S>,
P: LayerwisePolicy<B, A::Unit>,
B::ParallelContext: Sized,{ /* private fields */ }Expand description
Reusable executor for one architecture-owned, fully local traversal inside a neutral partitioned runtime.
The parallel context is an opaque backend mechanism selected from the manifest. Family input, forward context, and traversal decisions remain statically typed by the architecture.
Implementations§
Source§impl<A, B, S, P> LayerwiseTraversalPartitionExecutor<A, B, S, P>where
B: SubmissionBackend<Executor = <<B as NeuralBackend>::Tensor as Tensor>::Context>,
S: RuntimeState<B>,
A: LayeredArchitecture<B, S>,
P: LayerwisePolicy<B, A::Unit>,
B::ParallelContext: Sized,
impl<A, B, S, P> LayerwiseTraversalPartitionExecutor<A, B, S, P>where
B: SubmissionBackend<Executor = <<B as NeuralBackend>::Tensor as Tensor>::Context>,
S: RuntimeState<B>,
A: LayeredArchitecture<B, S>,
P: LayerwisePolicy<B, A::Unit>,
B::ParallelContext: Sized,
Sourcepub const fn new(
runtime: LayerwiseRuntime<A, B, S, P>,
parallel: B::ParallelContext,
) -> Self
pub const fn new( runtime: LayerwiseRuntime<A, B, S, P>, parallel: B::ParallelContext, ) -> Self
Pairs one local layered runtime with its selected opaque parallel context.
Sourcepub const fn runtime(&self) -> &LayerwiseRuntime<A, B, S, P>
pub const fn runtime(&self) -> &LayerwiseRuntime<A, B, S, P>
Borrows the installed layered runtime for residency reporting.
Auto Trait Implementations§
impl<A, B, S, P> Freeze for LayerwiseTraversalPartitionExecutor<A, B, S, P>
impl<A, B, S, P> RefUnwindSafe for LayerwiseTraversalPartitionExecutor<A, B, S, P>where
<B as NeuralBackend>::ParallelContext: Sized + RefUnwindSafe,
LayerwiseRuntime<A, B, S, P>: RefUnwindSafe,
impl<A, B, S, P> Send for LayerwiseTraversalPartitionExecutor<A, B, S, P>
impl<A, B, S, P> Sync for LayerwiseTraversalPartitionExecutor<A, B, S, P>
impl<A, B, S, P> Unpin for LayerwiseTraversalPartitionExecutor<A, B, S, P>
impl<A, B, S, P> UnsafeUnpin for LayerwiseTraversalPartitionExecutor<A, B, S, P>where
<B as NeuralBackend>::ParallelContext: Sized + UnsafeUnpin,
LayerwiseRuntime<A, B, S, P>: UnsafeUnpin,
impl<A, B, S, P> UnwindSafe for LayerwiseTraversalPartitionExecutor<A, B, S, P>where
<B as NeuralBackend>::ParallelContext: Sized + UnwindSafe,
LayerwiseRuntime<A, B, S, P>: UnwindSafe,
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