pub enum TrtPrecision {
Fp32,
Fp16,
Bf16,
Int8,
Fp8,
Best,
}Expand description
Serializable mirror of atomr_accel_tensorrt::builder::Precision
so configs can be parsed without pulling the upstream crate.
Variants§
Fp32
FP32 with TF32 matmul (the default).
Fp16
FP16 + TF32.
Bf16
BF16 + TF32.
Int8
INT8 + TF32. Requires PTQ calibration at build time.
Fp8
FP8 (Hopper+) + FP16 + TF32.
Best
Let the builder pick the fastest tactic (FP16 | BF16 | INT8 | FP8 | TF32 all enabled).
Trait Implementations§
Source§impl Clone for TrtPrecision
impl Clone for TrtPrecision
Source§fn clone(&self) -> TrtPrecision
fn clone(&self) -> TrtPrecision
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 TrtPrecision
impl Debug for TrtPrecision
Source§impl Default for TrtPrecision
impl Default for TrtPrecision
Source§fn default() -> TrtPrecision
fn default() -> TrtPrecision
Returns the “default value” for a type. Read more
Source§impl<'de> Deserialize<'de> for TrtPrecision
impl<'de> Deserialize<'de> for TrtPrecision
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
Source§impl Serialize for TrtPrecision
impl Serialize for TrtPrecision
impl Copy for TrtPrecision
Auto Trait Implementations§
impl Freeze for TrtPrecision
impl RefUnwindSafe for TrtPrecision
impl Send for TrtPrecision
impl Sync for TrtPrecision
impl Unpin for TrtPrecision
impl UnsafeUnpin for TrtPrecision
impl UnwindSafe for TrtPrecision
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