ArmMachineLearningModelFactory.ImageModelSettingsObjectDetection Method
Definition
Important
Some information relates to prerelease product that may be substantially modified before it’s released. Microsoft makes no warranties, express or implied, with respect to the information provided here.
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.
public static Azure.ResourceManager.MachineLearning.Models.ImageModelSettingsObjectDetection ImageModelSettingsObjectDetection(string advancedSettings = default, bool? amsGradient = default, string augmentations = default, float? beta1 = default, float? beta2 = default, int? checkpointFrequency = default, Azure.ResourceManager.MachineLearning.Models.MachineLearningFlowModelJobInput checkpointModel = default, string checkpointRunId = default, bool? distributed = default, bool? earlyStopping = default, int? earlyStoppingDelay = default, int? earlyStoppingPatience = default, bool? enableOnnxNormalization = default, int? evaluationFrequency = default, int? gradientAccumulationStep = default, int? layersToFreeze = default, float? learningRate = default, Azure.ResourceManager.MachineLearning.Models.LearningRateScheduler? learningRateScheduler = default, string modelName = default, float? momentum = default, bool? nesterov = default, int? numberOfEpochs = default, int? numberOfWorkers = default, Azure.ResourceManager.MachineLearning.Models.StochasticOptimizer? optimizer = default, int? randomSeed = default, float? stepLRGamma = default, int? stepLRStepSize = default, int? trainingBatchSize = default, int? validationBatchSize = default, float? warmupCosineLRCycles = default, int? warmupCosineLRWarmupEpochs = default, float? weightDecay = default, int? boxDetectionsPerImage = default, float? boxScoreThreshold = default, int? imageSize = default, int? maxSize = default, int? minSize = default, Azure.ResourceManager.MachineLearning.Models.MachineLearningModelSize? modelSize = default, bool? multiScale = default, float? nmsIouThreshold = default, string tileGridSize = default, float? tileOverlapRatio = default, float? tilePredictionsNmsThreshold = default, float? validationIouThreshold = default, Azure.ResourceManager.MachineLearning.Models.ValidationMetricType? validationMetricType = default);static member ImageModelSettingsObjectDetection : string * Nullable<bool> * string * Nullable<single> * Nullable<single> * Nullable<int> * Azure.ResourceManager.MachineLearning.Models.MachineLearningFlowModelJobInput * string * Nullable<bool> * Nullable<bool> * Nullable<int> * Nullable<int> * Nullable<bool> * Nullable<int> * Nullable<int> * Nullable<int> * Nullable<single> * Nullable<Azure.ResourceManager.MachineLearning.Models.LearningRateScheduler> * string * Nullable<single> * Nullable<bool> * Nullable<int> * Nullable<int> * Nullable<Azure.ResourceManager.MachineLearning.Models.StochasticOptimizer> * Nullable<int> * Nullable<single> * Nullable<int> * Nullable<int> * Nullable<int> * Nullable<single> * Nullable<int> * Nullable<single> * Nullable<int> * Nullable<single> * Nullable<int> * Nullable<int> * Nullable<int> * Nullable<Azure.ResourceManager.MachineLearning.Models.MachineLearningModelSize> * Nullable<bool> * Nullable<single> * string * Nullable<single> * Nullable<single> * Nullable<single> * Nullable<Azure.ResourceManager.MachineLearning.Models.ValidationMetricType> -> Azure.ResourceManager.MachineLearning.Models.ImageModelSettingsObjectDetectionPublic Shared Function ImageModelSettingsObjectDetection (Optional advancedSettings As String = Nothing, Optional amsGradient As Nullable(Of Boolean) = Nothing, Optional augmentations As String = Nothing, Optional beta1 As Nullable(Of Single) = Nothing, Optional beta2 As Nullable(Of Single) = Nothing, Optional checkpointFrequency As Nullable(Of Integer) = Nothing, Optional checkpointModel As MachineLearningFlowModelJobInput = Nothing, Optional checkpointRunId As String = Nothing, Optional distributed As Nullable(Of Boolean) = Nothing, Optional earlyStopping As Nullable(Of Boolean) = Nothing, Optional earlyStoppingDelay As Nullable(Of Integer) = Nothing, Optional earlyStoppingPatience As Nullable(Of Integer) = Nothing, Optional enableOnnxNormalization As Nullable(Of Boolean) = Nothing, Optional evaluationFrequency As Nullable(Of Integer) = Nothing, Optional gradientAccumulationStep As Nullable(Of Integer) = Nothing, Optional layersToFreeze As Nullable(Of Integer) = Nothing, Optional learningRate As Nullable(Of Single) = Nothing, Optional learningRateScheduler As Nullable(Of LearningRateScheduler) = Nothing, Optional modelName As String = Nothing, Optional momentum As Nullable(Of Single) = Nothing, Optional nesterov As Nullable(Of Boolean) = Nothing, Optional numberOfEpochs As Nullable(Of Integer) = Nothing, Optional numberOfWorkers As Nullable(Of Integer) = Nothing, Optional optimizer As Nullable(Of StochasticOptimizer) = Nothing, Optional randomSeed As Nullable(Of Integer) = Nothing, Optional stepLRGamma As Nullable(Of Single) = Nothing, Optional stepLRStepSize As Nullable(Of Integer) = Nothing, Optional trainingBatchSize As Nullable(Of Integer) = Nothing, Optional validationBatchSize As Nullable(Of Integer) = Nothing, Optional warmupCosineLRCycles As Nullable(Of Single) = Nothing, Optional warmupCosineLRWarmupEpochs As Nullable(Of Integer) = Nothing, Optional weightDecay As Nullable(Of Single) = Nothing, Optional boxDetectionsPerImage As Nullable(Of Integer) = Nothing, Optional boxScoreThreshold As Nullable(Of Single) = Nothing, Optional imageSize As Nullable(Of Integer) = Nothing, Optional maxSize As Nullable(Of Integer) = Nothing, Optional minSize As Nullable(Of Integer) = Nothing, Optional modelSize As Nullable(Of MachineLearningModelSize) = Nothing, Optional multiScale As Nullable(Of Boolean) = Nothing, Optional nmsIouThreshold As Nullable(Of Single) = Nothing, Optional tileGridSize As String = Nothing, Optional tileOverlapRatio As Nullable(Of Single) = Nothing, Optional tilePredictionsNmsThreshold As Nullable(Of Single) = Nothing, Optional validationIouThreshold As Nullable(Of Single) = Nothing, Optional validationMetricType As Nullable(Of ValidationMetricType) = Nothing) As ImageModelSettingsObjectDetectionParameters
- advancedSettings
- String
Settings for advanced scenarios.
- augmentations
- String
Settings for using Augmentations.
Value of 'beta1' when optimizer is 'adam' or 'adamw'. Must be a float in the range [0, 1].
Value of 'beta2' when optimizer is 'adam' or 'adamw'. Must be a float in the range [0, 1].
Frequency to store model checkpoints. Must be a positive integer.
- checkpointModel
- MachineLearningFlowModelJobInput
The pretrained checkpoint model for incremental training.
- checkpointRunId
- String
The id of a previous run that has a pretrained checkpoint for incremental training.
Minimum number of epochs or validation evaluations to wait before primary metric improvement is tracked for early stopping. Must be a positive integer.
Minimum number of epochs or validation evaluations with no primary metric improvement before the run is stopped. Must be a positive integer.
Frequency to evaluate validation dataset to get metric scores. Must be a positive integer.
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.
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.
- learningRateScheduler
- Nullable<LearningRateScheduler>
Learning rate scheduler enum.
- modelName
- String
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.
Value of momentum when optimizer is 'sgd'. Must be a float in the range [0, 1].
- optimizer
- Nullable<StochasticOptimizer>
Stochastic optimizer for image models.
Value of gamma when learning rate scheduler is 'step'. Must be a float in the range [0, 1].
Value of step size when learning rate scheduler is 'step'. Must be a positive integer.
Value of cosine cycle when learning rate scheduler is 'warmup_cosine'. Must be a float in the range [0, 1].
Value of warmup epochs when learning rate scheduler is 'warmup_cosine'. Must be a positive integer.
Value of weight decay when optimizer is 'sgd', 'adam', or 'adamw'. Must be a float in the range[0, 1].
Maximum number of detections per image, for all classes. Must be a positive integer. Note: This settings is not supported for the 'yolov5' algorithm.
During inference, only return proposals with a classification score greater than BoxScoreThreshold. Must be a float in the range[0, 1].
Image size for train and validation. Must be a positive integer. Note: The training run may get into CUDA OOM if the size is too big. Note: This settings is only supported for the 'yolov5' algorithm.
Maximum size of the image to be rescaled before feeding it to the backbone. Must be a positive integer. Note: training run may get into CUDA OOM if the size is too big. Note: This settings is not supported for the 'yolov5' algorithm.
Minimum size of the image to be rescaled before feeding it to the backbone. Must be a positive integer. Note: training run may get into CUDA OOM if the size is too big. Note: This settings is not supported for the 'yolov5' algorithm.
- modelSize
- Nullable<MachineLearningModelSize>
Image model size.
Enable multi-scale image by varying image size by +/- 50%. Note: training run may get into CUDA OOM if no sufficient GPU memory. Note: This settings is only supported for the 'yolov5' algorithm.
IOU threshold used during inference in NMS post processing. Must be a float in the range [0, 1].
- tileGridSize
- String
The grid size to use for tiling each image. Note: TileGridSize must not be None to enable small object detection logic. A string containing two integers in mxn format. Note: This settings is not supported for the 'yolov5' algorithm.
Overlap ratio between adjacent tiles in each dimension. Must be float in the range [0, 1). Note: This settings is not supported for the 'yolov5' algorithm.
The IOU threshold to use to perform NMS while merging predictions from tiles and image. Used in validation/ inference. Must be float in the range [0, 1]. Note: This settings is not supported for the 'yolov5' algorithm.
IOU threshold to use when computing validation metric. Must be float in the range [0, 1].
- validationMetricType
- Nullable<ValidationMetricType>
Metric computation method to use for validation metrics in image tasks.
Returns
A new ImageModelSettingsObjectDetection instance for mocking.
