pub enum KVProfile {
KvRaw,
KvInt8,
KvMixed,
KvPaged,
}Expand description
KV-cache compression profile.
Variants§
KvRaw
Raw FP16/BF16 KV cache.
KvInt8
INT8-quantised KV cache.
KvMixed
Mixed-precision KV cache (attention heads split).
KvPaged
Paged KV cache for variable sequence lengths.
Trait Implementations§
impl Copy for KVProfile
Source§impl<'de> Deserialize<'de> for KVProfile
impl<'de> Deserialize<'de> for KVProfile
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
impl Eq for KVProfile
impl StructuralPartialEq for KVProfile
Auto Trait Implementations§
impl Freeze for KVProfile
impl RefUnwindSafe for KVProfile
impl Send for KVProfile
impl Sync for KVProfile
impl Unpin for KVProfile
impl UnsafeUnpin for KVProfile
impl UnwindSafe for KVProfile
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