Struct StateTensorPolicy 

pub struct StateTensorPolicy {
    pub role: StateTensorRole,
    pub shape: Vec<StateTensorDimension>,
    pub dtype: StateTensorDtype,
    pub residency: StateResidencyClass,
    pub presence: StateTensorPresence,
}

Exact semantic role, symbolic shape, and dtype contract for a state tensor.

Fields

role: StateTensorRole

Meaning of this tensor within its owner.

shape: Vec<StateTensorDimension>

Symbolic shape resolved from batch and prefix geometry.

dtype: StateTensorDtype

Accepted dtype family.

residency: StateResidencyClass

Authoritative live-state residency behavior.

presence: StateTensorPresence

Condition under which persisted caches materialize this tensor.

Implementations

impl StateTensorPolicy

pub fn new( role: StateTensorRole, shape: Vec<StateTensorDimension>, dtype: StateTensorDtype, residency: MutableStateResidency, ) -> Result<Self, CachePolicyError>

Constructs and validates a state-tensor policy.

pub fn new_with_residency( role: StateTensorRole, shape: Vec<StateTensorDimension>, dtype: StateTensorDtype, residency: StateResidencyClass, ) -> Result<Self, CachePolicyError>

Constructs state with an explicit pageable or mutable residency class.

pub const fn optional(self) -> Self

Marks this tensor as optional.

pub const fn when_prefix_remainder_nonzero(self, divisor: NonZeroU32) -> Self

Requires the tensor when the prefix has a non-zero remainder.

pub const fn when_prefix_at_least(self, divisor: NonZeroU32) -> Self

Requires the tensor when the prefix contains a complete group.

pub fn is_required_for(&self, prefix_tokens: usize) -> bool

Returns whether the tensor is required for an exact prefix length.

pub fn residency_class(&self) -> StateResidencyClass

Returns the unified residency classification.

pub fn resolved_shape( &self, batch_size: usize, prefix_tokens: usize, ) -> Result<Vec<i32>, CachePolicyError>

Resolves symbolic dimensions for an exact batch and prefix length.

pub fn accepts_dtype_name(&self, dtype: &str) -> bool

Tests a stable serialized dtype name against this policy.

Trait Implementations

impl Clone for StateTensorPolicy

fn clone(&self) -> StateTensorPolicy

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for StateTensorPolicy

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for StateTensorPolicy

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Eq for StateTensorPolicy

impl Hash for StateTensorPolicy

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl PartialEq for StateTensorPolicy

fn eq(&self, other: &StateTensorPolicy) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl Serialize for StateTensorPolicy

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for StateTensorPolicy