OnnxCatalog.ApplyOnnxModel Méthode

Définition

Surcharges

ApplyOnnxModel(TransformsCatalog, OnnxOptions)

Créez un OnnxScoringEstimator fichier à l’aide de l’objet spécifié OnnxOptions. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

ApplyOnnxModel(TransformsCatalog, String, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné à la colonne d’entrée. Les colonnes d’entrée/sortie sont déterminées en fonction des colonnes d’entrée/sortie du modèle ONNX fourni. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

ApplyOnnxModel(TransformsCatalog, String, IDictionary<String,Int32[]>, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné à la colonne d’entrée. Les colonnes d’entrée/sortie sont déterminées en fonction des colonnes d’entrée/sortie du modèle ONNX fourni. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

ApplyOnnxModel(TransformsCatalog, String, String, String, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné à la inputColumnName colonne. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

ApplyOnnxModel(TransformsCatalog, String[], String[], String, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné aux inputColumnNames colonnes. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

ApplyOnnxModel(TransformsCatalog, String, String, String, IDictionary<String,Int32[]>, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné à la inputColumnName colonne. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

ApplyOnnxModel(TransformsCatalog, String[], String[], String, IDictionary<String,Int32[]>, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné aux inputColumnNames colonnes. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

ApplyOnnxModel(TransformsCatalog, String[], String[], String, IDictionary<String,Int32[]>, Nullable<Int32>, Boolean, Int32)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné aux inputColumnNames colonnes. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

ApplyOnnxModel(TransformsCatalog, OnnxOptions)

Créez un OnnxScoringEstimator fichier à l’aide de l’objet spécifié OnnxOptions. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

public static Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator ApplyOnnxModel (this Microsoft.ML.TransformsCatalog catalog, Microsoft.ML.Transforms.Onnx.OnnxOptions options);
static member ApplyOnnxModel : Microsoft.ML.TransformsCatalog * Microsoft.ML.Transforms.Onnx.OnnxOptions -> Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator
<Extension()>
Public Function ApplyOnnxModel (catalog As TransformsCatalog, options As OnnxOptions) As OnnxScoringEstimator

Paramètres

catalog
TransformsCatalog

Catalogue de la transformation.

options
OnnxOptions

Options pour le OnnxScoringEstimator.

Retours

Remarques

Si les options sont disponibles. La valeur GpuDeviceId est null la MLContext.GpuDeviceId valeur qui sera utilisée si ce n’est pas nullle cas.

S’applique à

ApplyOnnxModel(TransformsCatalog, String, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné à la colonne d’entrée. Les colonnes d’entrée/sortie sont déterminées en fonction des colonnes d’entrée/sortie du modèle ONNX fourni. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

public static Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator ApplyOnnxModel (this Microsoft.ML.TransformsCatalog catalog, string modelFile, int? gpuDeviceId = default, bool fallbackToCpu = false);
static member ApplyOnnxModel : Microsoft.ML.TransformsCatalog * string * Nullable<int> * bool -> Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator
<Extension()>
Public Function ApplyOnnxModel (catalog As TransformsCatalog, modelFile As String, Optional gpuDeviceId As Nullable(Of Integer) = Nothing, Optional fallbackToCpu As Boolean = false) As OnnxScoringEstimator

Paramètres

catalog
TransformsCatalog

Catalogue de la transformation.

modelFile
String

Chemin d’accès du fichier contenant le modèle ONNX.

gpuDeviceId
Nullable<Int32>

ID d’appareil GPU facultatif sur lequel exécuter l’exécution, null pour s’exécuter sur le processeur.

fallbackToCpu
Boolean

Si une erreur GPU est générée, déclenchez une exception ou une secours au processeur.

Retours

Exemples

using System;
using System.IO;
using System.Linq;
using Microsoft.ML;
using Microsoft.ML.Data;

namespace Samples.Dynamic
{
    public static class ApplyOnnxModel
    {
        public static void Example()
        {
            // Download the squeeznet image model from ONNX model zoo, version 1.2
            // https://github.com/onnx/models/tree/master/squeezenet or
            // https://s3.amazonaws.com/download.onnx/models/opset_8/squeezenet.tar.gz
            // or use Microsoft.ML.Onnx.TestModels nuget.
            var modelPath = @"squeezenet\00000001\model.onnx";

            // Create ML pipeline to score the data using OnnxScoringEstimator
            var mlContext = new MLContext();

            // Generate sample test data.
            var samples = GetTensorData();
            // Convert training data to IDataView, the general data type used in
            // ML.NET.
            var data = mlContext.Data.LoadFromEnumerable(samples);
            // Create the pipeline to score using provided onnx model.
            var pipeline = mlContext.Transforms.ApplyOnnxModel(modelPath);
            // Fit the pipeline and get the transformed values
            var transformedValues = pipeline.Fit(data).Transform(data);
            // Retrieve model scores into Prediction class
            var predictions = mlContext.Data.CreateEnumerable<Prediction>(
                transformedValues, reuseRowObject: false);

            // Iterate rows
            foreach (var prediction in predictions)
            {
                int numClasses = 0;
                foreach (var classScore in prediction.softmaxout_1.Take(3))
                {
                    Console.WriteLine("Class #" + numClasses++ + " score = " +
                        classScore);
                }
                Console.WriteLine(new string('-', 10));
            }

            // Results look like below...
            // Class #0 score = 4.544065E-05
            // Class #1 score = 0.003845858
            // Class #2 score = 0.0001249467
            // ----------
            // Class #0 score = 4.491953E-05
            // Class #1 score = 0.003848222
            // Class #2 score = 0.0001245592
            // ----------
        }

        // inputSize is the overall dimensions of the model input tensor.
        private const int inputSize = 224 * 224 * 3;

        // A class to hold sample tensor data. Member name should match
        // the inputs that the model expects (in this case, data_0)
        public class TensorData
        {
            [VectorType(inputSize)]
            public float[] data_0 { get; set; }
        }

        // Method to generate sample test data. Returns 2 sample rows.
        public static TensorData[] GetTensorData()
        {
            // This can be any numerical data. Assume image pixel values.
            var image1 = Enumerable.Range(0, inputSize).Select(x => (float)x /
                inputSize).ToArray();

            var image2 = Enumerable.Range(0, inputSize).Select(x => (float)(x +
                10000) / inputSize).ToArray();

            return new TensorData[] { new TensorData() { data_0 = image1 }, new
                TensorData() { data_0 = image2 } };
        }

        // Class to contain the output values from the transformation.
        // This model generates a vector of 1000 floats.
        class Prediction
        {
            [VectorType(1000)]
            public float[] softmaxout_1 { get; set; }
        }
    }
}

Remarques

Le nom/type des colonnes d’entrée doit correspondre exactement au nom/au type des entrées du modèle ONNX. Le nom/type des colonnes de sortie produites correspond au nom/au type des sorties du modèle ONNX. Si la valeur gpuDeviceId est null la MLContext.GpuDeviceId valeur sera utilisée si ce n’est pas nullle cas.

S’applique à

ApplyOnnxModel(TransformsCatalog, String, IDictionary<String,Int32[]>, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné à la colonne d’entrée. Les colonnes d’entrée/sortie sont déterminées en fonction des colonnes d’entrée/sortie du modèle ONNX fourni. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

public static Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator ApplyOnnxModel (this Microsoft.ML.TransformsCatalog catalog, string modelFile, System.Collections.Generic.IDictionary<string,int[]> shapeDictionary, int? gpuDeviceId = default, bool fallbackToCpu = false);
static member ApplyOnnxModel : Microsoft.ML.TransformsCatalog * string * System.Collections.Generic.IDictionary<string, int[]> * Nullable<int> * bool -> Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator
<Extension()>
Public Function ApplyOnnxModel (catalog As TransformsCatalog, modelFile As String, shapeDictionary As IDictionary(Of String, Integer()), Optional gpuDeviceId As Nullable(Of Integer) = Nothing, Optional fallbackToCpu As Boolean = false) As OnnxScoringEstimator

Paramètres

catalog
TransformsCatalog

Catalogue de la transformation.

modelFile
String

Chemin d’accès du fichier contenant le modèle ONNX.

shapeDictionary
IDictionary<String,Int32[]>

Formes ONNX à utiliser sur celles chargées à partir de modelFile. Pour les clés, utilisez des noms comme indiqué dans le modèle ONNX, par exemple « input ». La déclaration des formes avec ce paramètre est particulièrement utile pour l’utilisation des entrées et sorties de dimension variable.

gpuDeviceId
Nullable<Int32>

ID d’appareil GPU facultatif sur lequel exécuter l’exécution, null pour s’exécuter sur le processeur.

fallbackToCpu
Boolean

Si une erreur GPU est générée, déclenchez une exception ou une secours au processeur.

Retours

Exemples

using System;
using System.IO;
using System.Linq;
using Microsoft.ML;
using Microsoft.ML.Data;

namespace Samples.Dynamic
{
    public static class ApplyOnnxModel
    {
        public static void Example()
        {
            // Download the squeeznet image model from ONNX model zoo, version 1.2
            // https://github.com/onnx/models/tree/master/squeezenet or
            // https://s3.amazonaws.com/download.onnx/models/opset_8/squeezenet.tar.gz
            // or use Microsoft.ML.Onnx.TestModels nuget.
            var modelPath = @"squeezenet\00000001\model.onnx";

            // Create ML pipeline to score the data using OnnxScoringEstimator
            var mlContext = new MLContext();

            // Generate sample test data.
            var samples = GetTensorData();
            // Convert training data to IDataView, the general data type used in
            // ML.NET.
            var data = mlContext.Data.LoadFromEnumerable(samples);
            // Create the pipeline to score using provided onnx model.
            var pipeline = mlContext.Transforms.ApplyOnnxModel(modelPath);
            // Fit the pipeline and get the transformed values
            var transformedValues = pipeline.Fit(data).Transform(data);
            // Retrieve model scores into Prediction class
            var predictions = mlContext.Data.CreateEnumerable<Prediction>(
                transformedValues, reuseRowObject: false);

            // Iterate rows
            foreach (var prediction in predictions)
            {
                int numClasses = 0;
                foreach (var classScore in prediction.softmaxout_1.Take(3))
                {
                    Console.WriteLine("Class #" + numClasses++ + " score = " +
                        classScore);
                }
                Console.WriteLine(new string('-', 10));
            }

            // Results look like below...
            // Class #0 score = 4.544065E-05
            // Class #1 score = 0.003845858
            // Class #2 score = 0.0001249467
            // ----------
            // Class #0 score = 4.491953E-05
            // Class #1 score = 0.003848222
            // Class #2 score = 0.0001245592
            // ----------
        }

        // inputSize is the overall dimensions of the model input tensor.
        private const int inputSize = 224 * 224 * 3;

        // A class to hold sample tensor data. Member name should match
        // the inputs that the model expects (in this case, data_0)
        public class TensorData
        {
            [VectorType(inputSize)]
            public float[] data_0 { get; set; }
        }

        // Method to generate sample test data. Returns 2 sample rows.
        public static TensorData[] GetTensorData()
        {
            // This can be any numerical data. Assume image pixel values.
            var image1 = Enumerable.Range(0, inputSize).Select(x => (float)x /
                inputSize).ToArray();

            var image2 = Enumerable.Range(0, inputSize).Select(x => (float)(x +
                10000) / inputSize).ToArray();

            return new TensorData[] { new TensorData() { data_0 = image1 }, new
                TensorData() { data_0 = image2 } };
        }

        // Class to contain the output values from the transformation.
        // This model generates a vector of 1000 floats.
        class Prediction
        {
            [VectorType(1000)]
            public float[] softmaxout_1 { get; set; }
        }
    }
}

Remarques

Le nom/type des colonnes d’entrée doit correspondre exactement au nom/au type des entrées du modèle ONNX. Le nom/type des colonnes de sortie produites correspond au nom/au type des sorties du modèle ONNX. Si la valeur gpuDeviceId est null la MLContext.GpuDeviceId valeur sera utilisée si ce n’est pas nullle cas.

S’applique à

ApplyOnnxModel(TransformsCatalog, String, String, String, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné à la inputColumnName colonne. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

public static Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator ApplyOnnxModel (this Microsoft.ML.TransformsCatalog catalog, string outputColumnName, string inputColumnName, string modelFile, int? gpuDeviceId = default, bool fallbackToCpu = false);
static member ApplyOnnxModel : Microsoft.ML.TransformsCatalog * string * string * string * Nullable<int> * bool -> Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator
<Extension()>
Public Function ApplyOnnxModel (catalog As TransformsCatalog, outputColumnName As String, inputColumnName As String, modelFile As String, Optional gpuDeviceId As Nullable(Of Integer) = Nothing, Optional fallbackToCpu As Boolean = false) As OnnxScoringEstimator

Paramètres

catalog
TransformsCatalog

Catalogue de la transformation.

outputColumnName
String

Colonne de sortie résultant de la transformation.

inputColumnName
String

Colonne d’entrée.

modelFile
String

Chemin d’accès du fichier contenant le modèle ONNX.

gpuDeviceId
Nullable<Int32>

ID d’appareil GPU facultatif sur lequel exécuter l’exécution, null pour s’exécuter sur le processeur.

fallbackToCpu
Boolean

Si une erreur GPU est générée, déclenchez une exception ou une secours au processeur.

Retours

Exemples

using System;
using System.Linq;
using Microsoft.ML;
using Microsoft.ML.Data;
using Microsoft.ML.Transforms.Image;

namespace Samples.Dynamic
{
    public static class ApplyOnnxModelWithInMemoryImages
    {
        // Example of applying ONNX transform on in-memory images.
        public static void Example()
        {
            // Download the squeeznet image model from ONNX model zoo, version 1.2
            // https://github.com/onnx/models/tree/master/vision/classification/squeezenet or use
            // Microsoft.ML.Onnx.TestModels nuget.
            // It's a multiclass classifier. It consumes an input "data_0" and
            // produces an output "softmaxout_1".
            var modelPath = @"squeezenet\00000001\model.onnx";

            // Create ML pipeline to score the data using OnnxScoringEstimator
            var mlContext = new MLContext();

            // Create in-memory data points. Its Image/Scores field is the
            // input /output of the used ONNX model.
            var dataPoints = new ImageDataPoint[]
            {
                new ImageDataPoint(red: 255, green: 0, blue: 0), // Red color
                new ImageDataPoint(red: 0, green: 128, blue: 0)  // Green color
            };

            // Convert training data to IDataView, the general data type used in
            // ML.NET.
            var dataView = mlContext.Data.LoadFromEnumerable(dataPoints);

            // Create a ML.NET pipeline which contains two steps. First,
            // ExtractPixle is used to convert the 224x224 image to a 3x224x224
            // float tensor. Then the float tensor is fed into a ONNX model with an
            // input called "data_0" and an output called "softmaxout_1". Note that
            // "data_0" and "softmaxout_1" are model input and output names stored
            // in the used ONNX model file. Users may need to inspect their own
            // models to get the right input and output column names.
            // Map column "Image" to column "data_0"
            // Map column "data_0" to column "softmaxout_1"
            var pipeline = mlContext.Transforms.ExtractPixels("data_0", "Image")
                .Append(mlContext.Transforms.ApplyOnnxModel("softmaxout_1",
                "data_0", modelPath));

            var model = pipeline.Fit(dataView);
            var onnx = model.Transform(dataView);

            // Convert IDataView back to IEnumerable<ImageDataPoint> so that user
            // can inspect the output, column "softmaxout_1", of the ONNX transform.
            // Note that Column "softmaxout_1" would be stored in ImageDataPont
            //.Scores because the added attributed [ColumnName("softmaxout_1")]
            // tells that ImageDataPont.Scores is equivalent to column
            // "softmaxout_1".
            var transformedDataPoints = mlContext.Data.CreateEnumerable<
                ImageDataPoint>(onnx, false).ToList();

            // The scores are probabilities of all possible classes, so they should
            // all be positive.
            foreach (var dataPoint in transformedDataPoints)
            {
                var firstClassProb = dataPoint.Scores.First();
                var lastClassProb = dataPoint.Scores.Last();
                Console.WriteLine("The probability of being the first class is " +
                    (firstClassProb * 100) + "%.");

                Console.WriteLine($"The probability of being the last class is " +
                    (lastClassProb * 100) + "%.");
            }

            // Expected output:
            //  The probability of being the first class is 0.002542659%.
            //  The probability of being the last class is 0.0292684%.
            //  The probability of being the first class is 0.02258059%.
            //  The probability of being the last class is 0.394428%.
        }

        // This class is used in Example() to describe data points which will be
        // consumed by ML.NET pipeline.
        private class ImageDataPoint
        {
            // Height of Image.
            private const int height = 224;

            // Width of Image.
            private const int width = 224;

            // Image will be consumed by ONNX image multiclass classification model.
            [ImageType(height, width)]
            public MLImage Image { get; set; }

            // Expected output of ONNX model. It contains probabilities of all
            // classes. Note that the ColumnName below should match the output name
            // in the used ONNX model file.
            [ColumnName("softmaxout_1")]
            public float[] Scores { get; set; }

            public ImageDataPoint()
            {
                Image = null;
            }

            public ImageDataPoint(byte red, byte green, byte blue)
            {
                byte[] imageData = new byte[width * height * 4]; // 4 for the red, green, blue and alpha colors
                for (int i = 0; i < imageData.Length; i += 4)
                {
                    // Fill the buffer with the Bgra32 format
                    imageData[i] = blue;
                    imageData[i + 1] = green;
                    imageData[i + 2] = red;
                    imageData[i + 3] = 255;
                }

                Image = MLImage.CreateFromPixels(width, height, MLPixelFormat.Bgra32, imageData);
            }
        }
    }
}

Remarques

Si la valeur gpuDeviceId est null la MLContext.GpuDeviceId valeur sera utilisée si ce n’est pas nullle cas.

S’applique à

ApplyOnnxModel(TransformsCatalog, String[], String[], String, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné aux inputColumnNames colonnes. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

public static Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator ApplyOnnxModel (this Microsoft.ML.TransformsCatalog catalog, string[] outputColumnNames, string[] inputColumnNames, string modelFile, int? gpuDeviceId = default, bool fallbackToCpu = false);
static member ApplyOnnxModel : Microsoft.ML.TransformsCatalog * string[] * string[] * string * Nullable<int> * bool -> Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator
<Extension()>
Public Function ApplyOnnxModel (catalog As TransformsCatalog, outputColumnNames As String(), inputColumnNames As String(), modelFile As String, Optional gpuDeviceId As Nullable(Of Integer) = Nothing, Optional fallbackToCpu As Boolean = false) As OnnxScoringEstimator

Paramètres

catalog
TransformsCatalog

Catalogue de la transformation.

outputColumnNames
String[]

Colonnes de sortie résultant de la transformation.

inputColumnNames
String[]

Colonnes d’entrée.

modelFile
String

Chemin d’accès du fichier contenant le modèle ONNX.

gpuDeviceId
Nullable<Int32>

ID d’appareil GPU facultatif sur lequel exécuter l’exécution, null pour s’exécuter sur le processeur.

fallbackToCpu
Boolean

Si une erreur GPU est générée, déclenchez une exception ou une secours au processeur.

Retours

Remarques

Si la valeur gpuDeviceId est null la MLContext.GpuDeviceId valeur sera utilisée si ce n’est pas nullle cas.

S’applique à

ApplyOnnxModel(TransformsCatalog, String, String, String, IDictionary<String,Int32[]>, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné à la inputColumnName colonne. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

public static Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator ApplyOnnxModel (this Microsoft.ML.TransformsCatalog catalog, string outputColumnName, string inputColumnName, string modelFile, System.Collections.Generic.IDictionary<string,int[]> shapeDictionary, int? gpuDeviceId = default, bool fallbackToCpu = false);
static member ApplyOnnxModel : Microsoft.ML.TransformsCatalog * string * string * string * System.Collections.Generic.IDictionary<string, int[]> * Nullable<int> * bool -> Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator
<Extension()>
Public Function ApplyOnnxModel (catalog As TransformsCatalog, outputColumnName As String, inputColumnName As String, modelFile As String, shapeDictionary As IDictionary(Of String, Integer()), Optional gpuDeviceId As Nullable(Of Integer) = Nothing, Optional fallbackToCpu As Boolean = false) As OnnxScoringEstimator

Paramètres

catalog
TransformsCatalog

Catalogue de la transformation.

outputColumnName
String

Colonne de sortie résultant de la transformation.

inputColumnName
String

Colonne d’entrée.

modelFile
String

Chemin d’accès du fichier contenant le modèle ONNX.

shapeDictionary
IDictionary<String,Int32[]>

Formes ONNX à utiliser sur celles chargées à partir de modelFile. Pour les clés, utilisez des noms comme indiqué dans le modèle ONNX, par exemple « input ». La déclaration des formes avec ce paramètre est particulièrement utile pour l’utilisation des entrées et sorties de dimension variable.

gpuDeviceId
Nullable<Int32>

ID d’appareil GPU facultatif sur lequel exécuter l’exécution, null pour s’exécuter sur le processeur.

fallbackToCpu
Boolean

Si une erreur GPU est générée, déclenchez une exception ou une secours au processeur.

Retours

Exemples

using System;
using System.Linq;
using Microsoft.ML;
using Microsoft.ML.Data;
using Microsoft.ML.Transforms.Image;

namespace Samples.Dynamic
{
    public static class ApplyOnnxModelWithInMemoryImages
    {
        // Example of applying ONNX transform on in-memory images.
        public static void Example()
        {
            // Download the squeeznet image model from ONNX model zoo, version 1.2
            // https://github.com/onnx/models/tree/master/vision/classification/squeezenet or use
            // Microsoft.ML.Onnx.TestModels nuget.
            // It's a multiclass classifier. It consumes an input "data_0" and
            // produces an output "softmaxout_1".
            var modelPath = @"squeezenet\00000001\model.onnx";

            // Create ML pipeline to score the data using OnnxScoringEstimator
            var mlContext = new MLContext();

            // Create in-memory data points. Its Image/Scores field is the
            // input /output of the used ONNX model.
            var dataPoints = new ImageDataPoint[]
            {
                new ImageDataPoint(red: 255, green: 0, blue: 0), // Red color
                new ImageDataPoint(red: 0, green: 128, blue: 0)  // Green color
            };

            // Convert training data to IDataView, the general data type used in
            // ML.NET.
            var dataView = mlContext.Data.LoadFromEnumerable(dataPoints);

            // Create a ML.NET pipeline which contains two steps. First,
            // ExtractPixle is used to convert the 224x224 image to a 3x224x224
            // float tensor. Then the float tensor is fed into a ONNX model with an
            // input called "data_0" and an output called "softmaxout_1". Note that
            // "data_0" and "softmaxout_1" are model input and output names stored
            // in the used ONNX model file. Users may need to inspect their own
            // models to get the right input and output column names.
            // Map column "Image" to column "data_0"
            // Map column "data_0" to column "softmaxout_1"
            var pipeline = mlContext.Transforms.ExtractPixels("data_0", "Image")
                .Append(mlContext.Transforms.ApplyOnnxModel("softmaxout_1",
                "data_0", modelPath));

            var model = pipeline.Fit(dataView);
            var onnx = model.Transform(dataView);

            // Convert IDataView back to IEnumerable<ImageDataPoint> so that user
            // can inspect the output, column "softmaxout_1", of the ONNX transform.
            // Note that Column "softmaxout_1" would be stored in ImageDataPont
            //.Scores because the added attributed [ColumnName("softmaxout_1")]
            // tells that ImageDataPont.Scores is equivalent to column
            // "softmaxout_1".
            var transformedDataPoints = mlContext.Data.CreateEnumerable<
                ImageDataPoint>(onnx, false).ToList();

            // The scores are probabilities of all possible classes, so they should
            // all be positive.
            foreach (var dataPoint in transformedDataPoints)
            {
                var firstClassProb = dataPoint.Scores.First();
                var lastClassProb = dataPoint.Scores.Last();
                Console.WriteLine("The probability of being the first class is " +
                    (firstClassProb * 100) + "%.");

                Console.WriteLine($"The probability of being the last class is " +
                    (lastClassProb * 100) + "%.");
            }

            // Expected output:
            //  The probability of being the first class is 0.002542659%.
            //  The probability of being the last class is 0.0292684%.
            //  The probability of being the first class is 0.02258059%.
            //  The probability of being the last class is 0.394428%.
        }

        // This class is used in Example() to describe data points which will be
        // consumed by ML.NET pipeline.
        private class ImageDataPoint
        {
            // Height of Image.
            private const int height = 224;

            // Width of Image.
            private const int width = 224;

            // Image will be consumed by ONNX image multiclass classification model.
            [ImageType(height, width)]
            public MLImage Image { get; set; }

            // Expected output of ONNX model. It contains probabilities of all
            // classes. Note that the ColumnName below should match the output name
            // in the used ONNX model file.
            [ColumnName("softmaxout_1")]
            public float[] Scores { get; set; }

            public ImageDataPoint()
            {
                Image = null;
            }

            public ImageDataPoint(byte red, byte green, byte blue)
            {
                byte[] imageData = new byte[width * height * 4]; // 4 for the red, green, blue and alpha colors
                for (int i = 0; i < imageData.Length; i += 4)
                {
                    // Fill the buffer with the Bgra32 format
                    imageData[i] = blue;
                    imageData[i + 1] = green;
                    imageData[i + 2] = red;
                    imageData[i + 3] = 255;
                }

                Image = MLImage.CreateFromPixels(width, height, MLPixelFormat.Bgra32, imageData);
            }
        }
    }
}

Remarques

Si la valeur gpuDeviceId est null la MLContext.GpuDeviceId valeur sera utilisée si ce n’est pas nullle cas.

S’applique à

ApplyOnnxModel(TransformsCatalog, String[], String[], String, IDictionary<String,Int32[]>, Nullable<Int32>, Boolean)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné aux inputColumnNames colonnes. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

public static Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator ApplyOnnxModel (this Microsoft.ML.TransformsCatalog catalog, string[] outputColumnNames, string[] inputColumnNames, string modelFile, System.Collections.Generic.IDictionary<string,int[]> shapeDictionary, int? gpuDeviceId = default, bool fallbackToCpu = false);
static member ApplyOnnxModel : Microsoft.ML.TransformsCatalog * string[] * string[] * string * System.Collections.Generic.IDictionary<string, int[]> * Nullable<int> * bool -> Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator
<Extension()>
Public Function ApplyOnnxModel (catalog As TransformsCatalog, outputColumnNames As String(), inputColumnNames As String(), modelFile As String, shapeDictionary As IDictionary(Of String, Integer()), Optional gpuDeviceId As Nullable(Of Integer) = Nothing, Optional fallbackToCpu As Boolean = false) As OnnxScoringEstimator

Paramètres

catalog
TransformsCatalog

Catalogue de la transformation.

outputColumnNames
String[]

Colonnes de sortie résultant de la transformation.

inputColumnNames
String[]

Colonnes d’entrée.

modelFile
String

Chemin d’accès du fichier contenant le modèle ONNX.

shapeDictionary
IDictionary<String,Int32[]>

Formes ONNX à utiliser sur celles chargées à partir de modelFile. Pour les clés, utilisez des noms comme indiqué dans le modèle ONNX, par exemple « input ». La déclaration des formes avec ce paramètre est particulièrement utile pour l’utilisation des entrées et sorties de dimension variable.

gpuDeviceId
Nullable<Int32>

ID d’appareil GPU facultatif sur lequel exécuter l’exécution, null pour s’exécuter sur le processeur.

fallbackToCpu
Boolean

Si une erreur GPU est générée, déclenchez une exception ou une secours au processeur.

Retours

Remarques

Si la valeur gpuDeviceId est null la MLContext.GpuDeviceId valeur sera utilisée si ce n’est pas nullle cas.

S’applique à

ApplyOnnxModel(TransformsCatalog, String[], String[], String, IDictionary<String,Int32[]>, Nullable<Int32>, Boolean, Int32)

Créez un OnnxScoringEstimatormodèle Onnx préentraîné aux inputColumnNames colonnes. Pour OnnxScoringEstimator en savoir plus sur les dépendances nécessaires, consultez comment l’exécuter sur un GPU.

public static Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator ApplyOnnxModel (this Microsoft.ML.TransformsCatalog catalog, string[] outputColumnNames, string[] inputColumnNames, string modelFile, System.Collections.Generic.IDictionary<string,int[]> shapeDictionary, int? gpuDeviceId = default, bool fallbackToCpu = false, int recursionLimit = 100);
static member ApplyOnnxModel : Microsoft.ML.TransformsCatalog * string[] * string[] * string * System.Collections.Generic.IDictionary<string, int[]> * Nullable<int> * bool * int -> Microsoft.ML.Transforms.Onnx.OnnxScoringEstimator
<Extension()>
Public Function ApplyOnnxModel (catalog As TransformsCatalog, outputColumnNames As String(), inputColumnNames As String(), modelFile As String, shapeDictionary As IDictionary(Of String, Integer()), Optional gpuDeviceId As Nullable(Of Integer) = Nothing, Optional fallbackToCpu As Boolean = false, Optional recursionLimit As Integer = 100) As OnnxScoringEstimator

Paramètres

catalog
TransformsCatalog

Catalogue de la transformation.

outputColumnNames
String[]

Colonnes de sortie résultant de la transformation.

inputColumnNames
String[]

Colonnes d’entrée.

modelFile
String

Chemin d’accès du fichier contenant le modèle ONNX.

shapeDictionary
IDictionary<String,Int32[]>

Formes ONNX à utiliser sur celles chargées à partir de modelFile. Pour les clés, utilisez des noms comme indiqué dans le modèle ONNX, par exemple « input ». La déclaration des formes avec ce paramètre est particulièrement utile pour l’utilisation des entrées et sorties de dimension variable.

gpuDeviceId
Nullable<Int32>

ID d’appareil GPU facultatif sur lequel exécuter l’exécution, null pour s’exécuter sur le processeur.

fallbackToCpu
Boolean

Si une erreur GPU est générée, déclenchez une exception ou une secours au processeur.

recursionLimit
Int32

Facultatif, spécifie la limite de récursivité Protobuf CodedInputStream. La valeur par défaut est 100.

Retours

Remarques

Si la valeur gpuDeviceId est null la MLContext.GpuDeviceId valeur sera utilisée si ce n’est pas nullle cas.

S’applique à