pulumi-azure-native-kotlin/com.pulumi.azurenative.machinelearningservices.kotlin.outputs/ImageModelSettingsClassificationResponse

ImageModelSettingsClassificationResponse

data class ImageModelSettingsClassificationResponse(val advancedSettings: String? = null, val amsGradient: Boolean? = null, val augmentations: String? = null, val beta1: Double? = null, val beta2: Double? = null, val checkpointFrequency: Int? = null, val checkpointModel: MLFlowModelJobInputResponse? = null, val checkpointRunId: String? = null, val distributed: Boolean? = null, val earlyStopping: Boolean? = null, val earlyStoppingDelay: Int? = null, val earlyStoppingPatience: Int? = null, val enableOnnxNormalization: Boolean? = null, val evaluationFrequency: Int? = null, val gradientAccumulationStep: Int? = null, val layersToFreeze: Int? = null, val learningRate: Double? = null, val learningRateScheduler: String? = null, val modelName: String? = null, val momentum: Double? = null, val nesterov: Boolean? = null, val numberOfEpochs: Int? = null, val numberOfWorkers: Int? = null, val optimizer: String? = null, val randomSeed: Int? = null, val stepLRGamma: Double? = null, val stepLRStepSize: Int? = null, val trainingBatchSize: Int? = null, val trainingCropSize: Int? = null, val validationBatchSize: Int? = null, val validationCropSize: Int? = null, val validationResizeSize: Int? = null, val warmupCosineLRCycles: Double? = null, val warmupCosineLRWarmupEpochs: Int? = null, val weightDecay: Double? = null, val weightedLoss: Int? = null)

Settings used for training the model. For more information on the available settings please visit the official documentation: https://docs.microsoft.com/en-us/azure/machine-learning/how-to-auto-train-image-models.

Constructors

fun ImageModelSettingsClassificationResponse(advancedSettings: String? = null, amsGradient: Boolean? = null, augmentations: String? = null, beta1: Double? = null, beta2: Double? = null, checkpointFrequency: Int? = null, checkpointModel: MLFlowModelJobInputResponse? = null, checkpointRunId: String? = null, distributed: Boolean? = null, earlyStopping: Boolean? = null, earlyStoppingDelay: Int? = null, earlyStoppingPatience: Int? = null, enableOnnxNormalization: Boolean? = null, evaluationFrequency: Int? = null, gradientAccumulationStep: Int? = null, layersToFreeze: Int? = null, learningRate: Double? = null, learningRateScheduler: String? = null, modelName: String? = null, momentum: Double? = null, nesterov: Boolean? = null, numberOfEpochs: Int? = null, numberOfWorkers: Int? = null, optimizer: String? = null, randomSeed: Int? = null, stepLRGamma: Double? = null, stepLRStepSize: Int? = null, trainingBatchSize: Int? = null, trainingCropSize: Int? = null, validationBatchSize: Int? = null, validationCropSize: Int? = null, validationResizeSize: Int? = null, warmupCosineLRCycles: Double? = null, warmupCosineLRWarmupEpochs: Int? = null, weightDecay: Double? = null, weightedLoss: Int? = null)

Types

Companion

object Companion

Properties

advancedSettings

val advancedSettings: String? = null

Settings for advanced scenarios.

amsGradient

val amsGradient: Boolean? = null

Enable AMSGrad when optimizer is 'adam' or 'adamw'.

augmentations

val augmentations: String? = null

Settings for using Augmentations.

beta1

val beta1: Double? = null

Value of 'beta1' when optimizer is 'adam' or 'adamw'. Must be a float in the range 0, 1.

beta2

val beta2: Double? = null

Value of 'beta2' when optimizer is 'adam' or 'adamw'. Must be a float in the range 0, 1.

checkpointFrequency

val checkpointFrequency: Int? = null

Frequency to store model checkpoints. Must be a positive integer.

checkpointModel

val checkpointModel: MLFlowModelJobInputResponse? = null

The pretrained checkpoint model for incremental training.

checkpointRunId

val checkpointRunId: String? = null

The id of a previous run that has a pretrained checkpoint for incremental training.

distributed

val distributed: Boolean? = null

Whether to use distributed training.

earlyStopping

val earlyStopping: Boolean? = null

Enable early stopping logic during training.

earlyStoppingDelay

val earlyStoppingDelay: Int? = null

Minimum number of epochs or validation evaluations to wait before primary metric improvement is tracked for early stopping. Must be a positive integer.

earlyStoppingPatience

val earlyStoppingPatience: Int? = null

Minimum number of epochs or validation evaluations with no primary metric improvement before the run is stopped. Must be a positive integer.

enableOnnxNormalization

val enableOnnxNormalization: Boolean? = null

Enable normalization when exporting ONNX model.

evaluationFrequency

val evaluationFrequency: Int? = null

Frequency to evaluate validation dataset to get metric scores. Must be a positive integer.

gradientAccumulationStep

val gradientAccumulationStep: Int? = null

Gradient accumulation means running a configured number of "GradAccumulationStep" steps without updating the model weights while accumulating the gradients of those steps, and then using the accumulated gradients to compute the weight updates. Must be a positive integer.

layersToFreeze

val layersToFreeze: Int? = null

Number of layers to freeze for the model. Must be a positive integer. For instance, passing 2 as value for 'seresnext' means freezing layer0 and layer1. For a full list of models supported and details on layer freeze, please see: https://docs.microsoft.com/en-us/azure/machine-learning/how-to-auto-train-image-models.

learningRate

val learningRate: Double? = null

Initial learning rate. Must be a float in the range 0, 1.

learningRateScheduler

val learningRateScheduler: String? = null

Type of learning rate scheduler. Must be 'warmup_cosine' or 'step'.

modelName

val modelName: String? = null

Name of the model to use for training. For more information on the available models please visit the official documentation: https://docs.microsoft.com/en-us/azure/machine-learning/how-to-auto-train-image-models.

momentum

val momentum: Double? = null

Value of momentum when optimizer is 'sgd'. Must be a float in the range 0, 1.

nesterov

val nesterov: Boolean? = null

Enable nesterov when optimizer is 'sgd'.

numberOfEpochs

val numberOfEpochs: Int? = null

Number of training epochs. Must be a positive integer.

numberOfWorkers

val numberOfWorkers: Int? = null

Number of data loader workers. Must be a non-negative integer.

optimizer

val optimizer: String? = null

Type of optimizer.

randomSeed

val randomSeed: Int? = null

Random seed to be used when using deterministic training.

stepLRGamma

val stepLRGamma: Double? = null

Value of gamma when learning rate scheduler is 'step'. Must be a float in the range 0, 1.

stepLRStepSize

val stepLRStepSize: Int? = null

Value of step size when learning rate scheduler is 'step'. Must be a positive integer.

trainingBatchSize

val trainingBatchSize: Int? = null

Training batch size. Must be a positive integer.

trainingCropSize

val trainingCropSize: Int? = null

Image crop size that is input to the neural network for the training dataset. Must be a positive integer.

validationBatchSize

val validationBatchSize: Int? = null

Validation batch size. Must be a positive integer.

validationCropSize

val validationCropSize: Int? = null

Image crop size that is input to the neural network for the validation dataset. Must be a positive integer.

validationResizeSize

val validationResizeSize: Int? = null

Image size to which to resize before cropping for validation dataset. Must be a positive integer.

warmupCosineLRCycles

val warmupCosineLRCycles: Double? = null

Value of cosine cycle when learning rate scheduler is 'warmup_cosine'. Must be a float in the range 0, 1.

warmupCosineLRWarmupEpochs

val warmupCosineLRWarmupEpochs: Int? = null

Value of warmup epochs when learning rate scheduler is 'warmup_cosine'. Must be a positive integer.

weightDecay

val weightDecay: Double? = null

Value of weight decay when optimizer is 'sgd', 'adam', or 'adamw'. Must be a float in the range0, 1.

weightedLoss

val weightedLoss: Int? = null

Weighted loss. The accepted values are 0 for no weighted loss. 1 for weighted loss with sqrt.(class_weights). 2 for weighted loss with class_weights. Must be 0 or 1 or 2.