Trait SupervisedLearningStrategy 

pub trait SupervisedLearningStrategy<M>
where
    M: LearnerModel,
{
    // Required method
    fn fit(
        &self,
        training_components: TrainingComponents<M>,
        learner: Learner<M>,
        dataloader_train: Arc<dyn DataLoader<<M as TrainStep>::Input>>,
        dataloader_valid: Arc<dyn DataLoader<<M as InferenceStep>::Input>>,
        starting_epoch: usize,
    ) -> (M, AsyncProcessorTraining<LearnerEvent<<M as TrainStep>::Output>, LearnerEvent<<M as InferenceStep>::Output>>);

    // Provided method
    fn train(
        &self,
        learner: Learner<M>,
        dataloader_train: Arc<dyn DataLoader<<M as TrainStep>::Input>>,
        dataloader_valid: Arc<dyn DataLoader<<M as InferenceStep>::Input>>,
        training_components: TrainingComponents<M>,
    ) -> LearningResult<M> { ... }
}

Provides the fit function for any learning strategy

Required Methods

fn fit( &self, training_components: TrainingComponents<M>, learner: Learner<M>, dataloader_train: Arc<dyn DataLoader<<M as TrainStep>::Input>>, dataloader_valid: Arc<dyn DataLoader<<M as InferenceStep>::Input>>, starting_epoch: usize, ) -> (M, AsyncProcessorTraining<LearnerEvent<<M as TrainStep>::Output>, LearnerEvent<<M as InferenceStep>::Output>>)

Training loop for this strategy

Provided Methods

fn train( &self, learner: Learner<M>, dataloader_train: Arc<dyn DataLoader<<M as TrainStep>::Input>>, dataloader_valid: Arc<dyn DataLoader<<M as InferenceStep>::Input>>, training_components: TrainingComponents<M>, ) -> LearningResult<M>

Train the learner’s model with this strategy.

Dyn Compatibility

This trait is dyn compatible.

In older versions of Rust, dyn compatibility was called "object safety".

Implementors