pub struct MemoryRequirements {
pub parameter_memory: u64,
pub activation_memory_per_token: usize,
pub kv_cache_memory_per_token: usize,
pub typed_sequence_state: Option<TypedSequenceStateMemory>,
pub overhead_memory: u64,
}Expand description
Memory requirements for model execution. This describes logical model requirements, not measured device allocation or an admission guarantee.
Fields§
§parameter_memory: u64Model parameter memory in bytes
activation_memory_per_token: usizeMinimum activation memory per token
kv_cache_memory_per_token: usizeLegacy KV cache bytes per token per layer. Ignored and omitted from serialization when typed_sequence_state supplies complete-model bytes.
typed_sequence_state: Option<TypedSequenceStateMemory>Exact logical sequence state from a typed model plan. Old wire values without this field retain the legacy per-layer interpretation.
overhead_memory: u64Additional overhead memory
Implementations§
Source§impl MemoryRequirements
impl MemoryRequirements
Trait Implementations§
Source§impl Clone for MemoryRequirements
impl Clone for MemoryRequirements
Source§fn clone(&self) -> MemoryRequirements
fn clone(&self) -> MemoryRequirements
Returns a duplicate of the value. Read more
1.0.0 (const: unstable) · Source§fn clone_from(&mut self, source: &Self)
fn clone_from(&mut self, source: &Self)
Performs copy-assignment from
source. Read moreSource§impl Debug for MemoryRequirements
impl Debug for MemoryRequirements
Source§impl<'de> Deserialize<'de> for MemoryRequirements
impl<'de> Deserialize<'de> for MemoryRequirements
Source§fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>where
__D: Deserializer<'de>,
fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>where
__D: Deserializer<'de>,
Deserialize this value from the given Serde deserializer. Read more
Auto Trait Implementations§
impl Freeze for MemoryRequirements
impl RefUnwindSafe for MemoryRequirements
impl Send for MemoryRequirements
impl Sync for MemoryRequirements
impl Unpin for MemoryRequirements
impl UnsafeUnpin for MemoryRequirements
impl UnwindSafe for MemoryRequirements
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