pub enum Scale {
PerTensor,
PerToken,
PerChannel,
Group(u32),
Block(u32, u32),
}Expand description
2026-09-28: How the scales of an 8-bit float tensor are shared.
Variants§
PerTensor
2026-09-28: One scale for the whole tensor.
PerToken
2026-09-28: One scale per row (activations).
PerChannel
2026-09-28: One scale per output channel (weights).
Group(u32)
2026-09-28: One scale per n consecutive values along K.
Block(u32, u32)
2026-09-28: One scale per [rows, cols] block (weights).
Trait Implementations§
Source§impl Ord for Scale
impl Ord for Scale
Source§impl PartialOrd for Scale
impl PartialOrd for Scale
impl Copy for Scale
impl Eq for Scale
impl StructuralPartialEq for Scale
Auto Trait Implementations§
impl Freeze for Scale
impl RefUnwindSafe for Scale
impl Send for Scale
impl Sync for Scale
impl Unpin for Scale
impl UnwindSafe for Scale
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
Source§impl<T> CloneToUninit for Twhere
T: Clone,
impl<T> CloneToUninit for Twhere
T: Clone,
§impl<Q, K> Comparable<K> for Q
impl<Q, K> Comparable<K> for Q
§impl<Q, K> Equivalent<K> for Q
impl<Q, K> Equivalent<K> for Q
§fn equivalent(&self, key: &K) -> bool
fn equivalent(&self, key: &K) -> bool
Checks if this value is equivalent to the given key. Read more
§impl<Q, K> Equivalent<K> for Q
impl<Q, K> Equivalent<K> for Q
§fn equivalent(&self, key: &K) -> bool
fn equivalent(&self, key: &K) -> bool
Compare self to
key and return true if they are equal.