pub enum Precision {
F32,
F16,
}Expand description
How much of the graph runs in FP16.
Variants§
F32
FP32 values, weights and GEMMs. Matches Laya to about 1e-4 in the logits.
F16
FP16 weights and GEMM inputs with FP32 accumulation, the residual stream, attention’s qkv and the heads in FP32.
Trait Implementations§
impl Copy for Precision
impl Eq for Precision
impl StructuralPartialEq for Precision
Auto Trait Implementations§
impl Freeze for Precision
impl RefUnwindSafe for Precision
impl Send for Precision
impl Sync for Precision
impl Unpin for Precision
impl UnsafeUnpin for Precision
impl UnwindSafe for Precision
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
impl<ST, DT> CastableFrom<ST, Initialized, Initialized> for DT
impl<ST, DT> CastableFrom<ST, Uninit, Uninit> for DT
Source§impl<T> CloneToUninit for Twhere
T: Clone,
impl<T> CloneToUninit for Twhere
T: Clone,
Source§impl<Q, K> Equivalent<K> for Q
impl<Q, K> Equivalent<K> for Q
Source§impl<Q, K> Equivalent<K> for Q
impl<Q, K> Equivalent<K> for Q
Source§fn equivalent(&self, key: &K) -> bool
fn equivalent(&self, key: &K) -> bool
Compare self to
key and return true if they are equal.