pub struct DistributedTrainer<T: FloatBounds> { /* private fields */ }Expand description
Distributed training coordinator
Implementations§
Source§impl<T: FloatBounds + Default + Sum<T> + ScalarOperand + Copy> DistributedTrainer<T>
impl<T: FloatBounds + Default + Sum<T> + ScalarOperand + Copy> DistributedTrainer<T>
Sourcepub fn new(config: DistributedConfig<T>) -> DistributedResult<Self>
pub fn new(config: DistributedConfig<T>) -> DistributedResult<Self>
Create a new distributed trainer
Sourcepub fn distribute_data<D>(
&self,
data: &Array<T, D>,
) -> DistributedResult<Vec<Array<T, D>>>where
D: Dimension + RemoveAxis,
pub fn distribute_data<D>(
&self,
data: &Array<T, D>,
) -> DistributedResult<Vec<Array<T, D>>>where
D: Dimension + RemoveAxis,
Distribute data across workers
Sourcepub fn synchronize_gradients(
&mut self,
gradients: &mut HashMap<String, Array1<T>>,
) -> DistributedResult<()>
pub fn synchronize_gradients( &mut self, gradients: &mut HashMap<String, Array1<T>>, ) -> DistributedResult<()>
Synchronize gradients across workers
Sourcepub fn scale_learning_rate(&self, base_lr: T) -> T
pub fn scale_learning_rate(&self, base_lr: T) -> T
Scale learning rate based on number of workers
Sourcepub fn update_stats(&mut self, computation_time_ms: f64, memory_usage_mb: f64)
pub fn update_stats(&mut self, computation_time_ms: f64, memory_usage_mb: f64)
Update statistics
Sourcepub fn get_stats(&self) -> &DistributedStats<T>
pub fn get_stats(&self) -> &DistributedStats<T>
Get training statistics
Sourcepub fn training_step<F>(
&mut self,
compute_fn: F,
gradients: &mut HashMap<String, Array1<T>>,
) -> DistributedResult<()>
pub fn training_step<F>( &mut self, compute_fn: F, gradients: &mut HashMap<String, Array1<T>>, ) -> DistributedResult<()>
Perform distributed training step
Auto Trait Implementations§
impl<T> Freeze for DistributedTrainer<T>where
T: Freeze,
impl<T> RefUnwindSafe for DistributedTrainer<T>where
T: RefUnwindSafe,
impl<T> Send for DistributedTrainer<T>
impl<T> Sync for DistributedTrainer<T>
impl<T> Unpin for DistributedTrainer<T>where
T: Unpin,
impl<T> UnsafeUnpin for DistributedTrainer<T>where
T: UnsafeUnpin,
impl<T> UnwindSafe for DistributedTrainer<T>where
T: UnwindSafe + RefUnwindSafe,
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> IntoEither for T
impl<T> IntoEither for T
Source§fn into_either(self, into_left: bool) -> Either<Self, Self>
fn into_either(self, into_left: bool) -> Either<Self, Self>
Converts
self into a Left variant of Either<Self, Self>
if into_left is true.
Converts self into a Right variant of Either<Self, Self>
otherwise. Read moreSource§fn into_either_with<F>(self, into_left: F) -> Either<Self, Self>
fn into_either_with<F>(self, into_left: F) -> Either<Self, Self>
Converts
self into a Left variant of Either<Self, Self>
if into_left(&self) returns true.
Converts self into a Right variant of Either<Self, Self>
otherwise. Read more