MLoop.Extensibility 0.51.0

dotnet add package MLoop.Extensibility --version 0.51.0
                    
NuGet\Install-Package MLoop.Extensibility -Version 0.51.0
                    
This command is intended to be used within the Package Manager Console in Visual Studio, as it uses the NuGet module's version of Install-Package.
<PackageReference Include="MLoop.Extensibility" Version="0.51.0" />
                    
For projects that support PackageReference, copy this XML node into the project file to reference the package.
<PackageVersion Include="MLoop.Extensibility" Version="0.51.0" />
                    
Directory.Packages.props
<PackageReference Include="MLoop.Extensibility" />
                    
Project file
For projects that support Central Package Management (CPM), copy this XML node into the solution Directory.Packages.props file to version the package.
paket add MLoop.Extensibility --version 0.51.0
                    
#r "nuget: MLoop.Extensibility, 0.51.0"
                    
#r directive can be used in F# Interactive and Polyglot Notebooks. Copy this into the interactive tool or source code of the script to reference the package.
#:package MLoop.Extensibility@0.51.0
                    
#:package directive can be used in C# file-based apps starting in .NET 10 preview 4. Copy this into a .cs file before any lines of code to reference the package.
#addin nuget:?package=MLoop.Extensibility&version=0.51.0
                    
Install as a Cake Addin
#tool nuget:?package=MLoop.Extensibility&version=0.51.0
                    
Install as a Cake Tool

MLoop.Extensibility

Optional extensibility API for MLoop - enables code-based customization of AutoML pipeline.

Overview

MLoop.Extensibility provides interfaces for extending MLoop's AutoML capabilities through:

  • Hooks: Execute custom logic at lifecycle points (pre-train, post-train, etc.)
  • Custom Metrics: Optimize for business outcomes instead of generic accuracy

Features

  • ✅ Completely Optional: Works without any extensions
  • ✅ Zero-Overhead: < 1ms impact when not used
  • ✅ Type-Safe: Full C# type system with IDE support
  • ✅ Auto-Discovery: Convention-based discovery from filesystem

Quick Start

Install Package

dotnet add package MLoop.Extensibility

Create a Hook

A hook lives in .mloop/scripts/hooks/<stage>/ (pre-train, post-train, pre-predict, post-evaluate).

using System;
using System.IO;
using System.Linq;
using System.Threading.Tasks;
using MLoop.Extensibility.Hooks;

public class DataValidationHook : IMLoopHook
{
    public string Name => "Data Validation";

    public Task<HookResult> ExecuteAsync(HookContext ctx)
    {
        // The pre-train stage reads a file: its path is in the hook's metadata.
        var dataFile = ctx.GetMetadata<string>("DataFile", "");
        var rowCount = File.Exists(dataFile)
            ? File.ReadLines(dataFile).Count(line => !string.IsNullOrWhiteSpace(line)) - 1   // minus the header
            : 0;

        if (rowCount < 100)
            return Task.FromResult(HookResult.Abort($"Insufficient data: {rowCount} rows, need at least 100"));

        ctx.Logger.Info($"Data validation passed: {rowCount} rows");
        return Task.FromResult(HookResult.Continue());
    }
}

Create a Custom Metric

A metric lives in .mloop/scripts/metrics/. It is computed over the held-out rows after training and recorded with the run's metrics — a report, not an optimization target.

using System;
using System.Threading.Tasks;
using MLoop.Extensibility.Metrics;

public class ProfitMetric : IMLoopMetric
{
    public string Name => "Expected Profit";
    public string Description => "100 per true positive, -50 per false positive";

    public Task<MetricResult> CalculateAsync(MetricContext ctx)
    {
        var predicted = (bool[])ctx.Predictions;   // binary-classification: bool[]
        var actual = (bool[])ctx.Labels;
        double profit = 0;
        for (var i = 0; i < predicted.Length; i++)
            if (predicted[i]) profit += actual[i] ? 100 : -50;
        return Task.FromResult(new MetricResult { Name = Name, Value = profit });
    }
}

API Reference

Interfaces

  • IMLoopHook - Lifecycle hook interface
  • IMLoopMetric - Custom metric interface

Context Classes

  • HookContext - Provides data and ML context to hooks
  • MetricContext - Provides predictions for metric calculation
  • HookResult - Return value for hooks (Continue or Abort)

Logger

  • ILogger - Logging interface for hook output

Documentation

For complete documentation, see:

License

MIT

Product Compatible and additional computed target framework versions.
.NET net10.0 is compatible.  net10.0-android was computed.  net10.0-browser was computed.  net10.0-ios was computed.  net10.0-maccatalyst was computed.  net10.0-macos was computed.  net10.0-tvos was computed.  net10.0-windows was computed. 
Compatible target framework(s)
Included target framework(s) (in package)
Learn more about Target Frameworks and .NET Standard.

NuGet packages (1)

Showing the top 1 NuGet packages that depend on MLoop.Extensibility:

Package Downloads
MLoop.Core

MLoop Core - training (a trainer search over ML.NET trainers), prediction and model management on ML.NET

GitHub repositories

This package is not used by any popular GitHub repositories.

Version Downloads Last Updated
0.51.0 44 11/16/2026
0.50.0 103 11/8/2026
0.49.1 106 11/6/2026
0.49.0 115 11/1/2026
0.48.0 111 10/30/2026
0.47.0 113 10/27/2026
0.46.0 111 10/26/2026
0.45.1 111 10/24/2026
0.45.0 116 10/23/2026
0.44.0 114 10/23/2026
0.43.0 112 10/23/2026
0.42.0 124 10/22/2026
0.41.0 131 10/21/2026
0.40.0 123 10/21/2026
0.39.0 123 10/21/2026
0.38.0 118 10/21/2026
0.37.1 120 10/21/2026
0.37.0 117 10/21/2026
0.36.0 116 10/20/2026
0.35.0 128 10/20/2026
Loading failed