Skip to main content

CheckpointConfig

Struct CheckpointConfig 

Source
pub struct CheckpointConfig {
    pub strategy: CheckpointStrategy,
    pub checkpoint_attention: bool,
    pub checkpoint_ffn: bool,
    pub min_checkpoint_interval: usize,
}
Expand description

Gradient checkpointing configuration

Fields§

§strategy: CheckpointStrategy

Checkpointing strategy

§checkpoint_attention: bool

Whether to checkpoint attention

§checkpoint_ffn: bool

Whether to checkpoint feed-forward

§min_checkpoint_interval: usize

Minimum layers between checkpoints

Implementations§

Source§

impl CheckpointConfig

Source

pub fn uniform(interval: usize) -> Self

Create a uniform checkpointing configuration

§Arguments
  • interval - Checkpoint every N layers (e.g., 2 means checkpoint layers 0, 2, 4, …)
Source

pub fn selective(layers: Vec<usize>) -> Self

Create a selective checkpointing configuration

§Arguments
  • layers - Specific layer indices to checkpoint
Source

pub fn dynamic(num_layers: usize, memory_fraction: f64) -> Result<Self>

Create a dynamic checkpointing configuration

§Arguments
  • num_layers - Total number of layers in the model
  • memory_fraction - Target memory usage as fraction of full storage (0.0 - 1.0)
Source

pub fn none() -> Self

Disable checkpointing

Source

pub fn with_checkpoint_attention(self, checkpoint: bool) -> Self

Set whether to checkpoint attention sublayers

Source

pub fn with_checkpoint_ffn(self, checkpoint: bool) -> Self

Set whether to checkpoint feed-forward sublayers

Source

pub fn with_min_interval(self, interval: usize) -> Self

Set minimum interval between checkpoints

Source

pub fn should_checkpoint(&self, layer_idx: usize) -> bool

Check if a specific layer should be checkpointed

Source

pub fn memory_savings(&self, num_layers: usize) -> f64

Calculate expected memory savings

Returns the fraction of activation memory saved (0.0 - 1.0)

Source

pub fn compute_overhead(&self, num_layers: usize) -> f64

Calculate expected compute overhead

Returns the multiplicative factor for compute (1.0 = no overhead)

Source

pub fn validate(&self) -> Result<()>

Validate configuration

Source

pub fn summary(&self) -> String

Get human-readable summary

Trait Implementations§

Source§

impl Clone for CheckpointConfig

Source§

fn clone(&self) -> CheckpointConfig

Returns a duplicate of the value. Read more
1.0.0 · Source§

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source. Read more
Source§

impl Debug for CheckpointConfig

Source§

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter. Read more
Source§

impl Default for CheckpointConfig

Source§

fn default() -> Self

Returns the “default value” for a type. Read more
Source§

impl<'de> Deserialize<'de> for CheckpointConfig

Source§

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 CheckpointConfig

Source§

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>
where __S: Serializer,

Serialize this value into the given Serde serializer. Read more

Auto Trait Implementations§

Blanket Implementations§

Source§

impl<T> Any for T
where T: 'static + ?Sized,

Source§

fn type_id(&self) -> TypeId

Gets the TypeId of self. Read more
Source§

impl<T> Borrow<T> for T
where T: ?Sized,

Source§

fn borrow(&self) -> &T

Immutably borrows from an owned value. Read more
Source§

impl<T> BorrowMut<T> for T
where T: ?Sized,

Source§

fn borrow_mut(&mut self) -> &mut T

Mutably borrows from an owned value. Read more
Source§

impl<T> CloneToUninit for T
where T: Clone,

Source§

unsafe fn clone_to_uninit(&self, dest: *mut u8)

🔬This is a nightly-only experimental API. (clone_to_uninit)
Performs copy-assignment from self to dest. Read more
Source§

impl<T> From<T> for T

Source§

fn from(t: T) -> T

Returns the argument unchanged.

Source§

impl<T, U> Into<U> for T
where U: From<T>,

Source§

fn into(self) -> U

Calls U::from(self).

That is, this conversion is whatever the implementation of From<T> for U chooses to do.

Source§

impl<T> ToOwned for T
where T: Clone,

Source§

type Owned = T

The resulting type after obtaining ownership.
Source§

fn to_owned(&self) -> T

Creates owned data from borrowed data, usually by cloning. Read more
Source§

fn clone_into(&self, target: &mut T)

Uses borrowed data to replace owned data, usually by cloning. Read more
Source§

impl<T, U> TryFrom<U> for T
where U: Into<T>,

Source§

type Error = Infallible

The type returned in the event of a conversion error.
Source§

fn try_from(value: U) -> Result<T, <T as TryFrom<U>>::Error>

Performs the conversion.
Source§

impl<T, U> TryInto<U> for T
where U: TryFrom<T>,

Source§

type Error = <U as TryFrom<T>>::Error

The type returned in the event of a conversion error.
Source§

fn try_into(self) -> Result<U, <U as TryFrom<T>>::Error>

Performs the conversion.
Source§

impl<T> DeserializeOwned for T
where T: for<'de> Deserialize<'de>,