lamp-data
API
Generated with
lamp-data/lamp.data/IOLoops

IOLoops

lamp.data.IOLoops
object IOLoops

Contains a training loops and helpers around it

The two training loops implemented here are:

Attributes

Graph
Supertypes
class Object
trait Matchable
class Any
Self type
IOLoops.type

Members list

Type members

Classlikes

case class TrainingLoopContext(epoch: Int, lastValidationLoss: Option[Double], minValidationLoss: Option[Double])

Attributes

Companion
object
Supertypes
trait Serializable
trait Product
trait Equals
class Object
trait Matchable
class Any
Show all
object TrainingLoopContext

Attributes

Companion
class
Supertypes
trait Product
trait Mirror
class Object
trait Matchable
class Any
Self type

Value members

Concrete methods

def epochs[I, M <: GenericModule[I, Variable] : Load, LRState, BatchStreamState, BatchStreamBuffers](model: SupervisedModel[I, M], optimizerFactory: Seq[(STen, PTag)] => Optimizer, trainBatchesOverEpoch: TrainingLoopContext => BatchStream[(I, STen), BatchStreamState, BatchStreamBuffers], validationBatchesOverEpoch: Option[TrainingLoopContext => BatchStream[(I, STen), BatchStreamState, BatchStreamBuffers]], epochs: Int, trainingCallback: TrainingCallback, validationCallback: ValidationCallback, checkpointState: Option[(SimpleLoopState, LRState) => IO[Unit]], validationFrequency: Int, logger: Option[Logger], returnMinValidationLossModel: Seq[Int], learningRateSchedule: LearningRateSchedule[LRState], prefetch: Boolean, overlapModelWithLoad: Boolean, dataParallelModels: Seq[SupervisedModel[I, M]], initState: Option[SimpleLoopState], accumulateGradientOverNBatches: Int, learningRateScheduleInitState: Option[LRState], printOptimizerAllocations: Boolean, validationLossExponentialSmoothingFactor: Double): IO[(Int, SupervisedModel[I, M], List[(Int, Double, Option[(Double, Double)])], LRState, SimpleLoopState)]
def forwardAndDiscardBatchStream[I, M <: GenericModule[I, Variable], S, C](batchStream: BatchStream[(I, STen), S, C], buffers: Device => Resource[IO, C], model: M & GenericModule[I, Variable]): IO[Unit]
def oneEpoch[I, M <: GenericModule[I, Variable], S, C](epochCount: Long, trainingCallback: TrainingCallback, model: ModelWithOptimizer[I, M], trainBatches: BatchStream[(I, STen), S, C], logger: Option[Logger], learningRateScheduleFactor: Double, prefetch: Boolean, overlapModelWithLoad: Boolean, accumulateGradientOverNBatches: Int): IO[Double]
def parallelRunBatchStream[I, O, M <: GenericModule[I, O], S, O2 : Movable, C](batchStream: BatchStream[(I, STen), S, C], bufferPerModel: Resource[IO, List[(Device, C)]], models: Seq[M & GenericModule[I, O]])(tx: ((I, STen), O) => O2)(implicit evidence$2: Movable[O2], scope: Scope): IO[Vector[O2]]
def runBatchStream[A, B : Movable, M <: GenericModule[A, B], S, C](batchStream: BatchStream[A, S, C], buffers: Resource[IO, C], model: M & GenericModule[A, B])(implicit evidence$1: Movable[B], scope: Scope): IO[Vector[B]]
def validationOneEpoch[I, M <: GenericModule[I, Variable], S, C](model: SupervisedModel[I, M], validationBatches: BatchStream[(I, STen), S, C], validationCallback: ValidationCallback, logger: Option[Logger], epochCount: Long): IO[Double]
def withSWA[I, M <: GenericModule[I, Variable] : Load, LRState, LRStateSWA, BatchStreamState, BatchStreamBuffers](model: SupervisedModel[I, M], optimizerFactory: Seq[(STen, PTag)] => Optimizer, trainBatchesOverEpoch: TrainingLoopContext => BatchStream[(I, STen), BatchStreamState, BatchStreamBuffers], warmupEpochs: Int, swaEpochs: Int, validationBatchesOverEpoch: Option[TrainingLoopContext => BatchStream[(I, STen), BatchStreamState, BatchStreamBuffers]], trainingCallback: TrainingCallback, validationCallback: ValidationCallback, checkpointState: Option[(SimpleThenSWALoopState, Either[LRState, LRStateSWA]) => IO[Unit]], logger: Option[Logger], returnMinValidationLossModel: Seq[Int], learningRateSchedule: LearningRateSchedule[LRState], swaLearningRateSchedule: SWALearningRateSchedule[LRStateSWA], prefetch: Boolean, dataParallelModels: Seq[SupervisedModel[I, M]], initState: Option[SimpleThenSWALoopState], accumulateGradientOverNBatches: Int, learningRateScheduleInitState: Option[LRState], swaLearningRateScheduleInitState: Option[LRStateSWA], swaForwardPassAfterTraining: Boolean, validationLossExponentialSmoothingFactor: Double): IO[(Int, SupervisedModel[I, M], List[(Int, Double, Option[(Double, Double)])], SupervisedModel[I, M])]
Generated with