Skip to content

Latest commit

 

History

History
80 lines (52 loc) · 3.05 KB

File metadata and controls

80 lines (52 loc) · 3.05 KB

ML.NET Benchmarks/Performance Tests

This project contains performance benchmarks.

Run the Performance Tests

Pre-requisite: In order to fetch dependencies which come through Git submodules the following command needs to be run before building:

git submodule update --init

Pre-requisite: On a clean repo with initialized submodules, build.cmd at the root installs the right version of dotnet.exe and builds the solution. You need to build the solution in Release.

build.cmd -configuration Release
  1. Navigate to the performance tests directory (machinelearning\test\Microsoft.ML.PerformanceTests)

  2. Run the benchmarks in Release:

    build.cmd -configuration Release -performanceTest

.NET Core 3.1

Pre-requisite: Follow the netcoreapp3.1 instructions.

Pre-requisite: To use dotnet cli from the root directory remember to set DOTNET_MULTILEVEL_LOOKUP environment variable to 0!

$env:DOTNET_MULTILEVEL_LOOKUP=0
  1. Navigate to the benchmarks directory (machinelearning\test\Microsoft.ML.PerformanceTests)

  2. Run the benchmarks in Release-netcoreapp3_1 configuration:

    build.cmd -configuration Release-netcoreapp3_1 -performanceTest

Authoring new benchmarks

  1. The type which contains benchmark(s) has to be a public, non-sealed, non-static class.
  2. Put the initialization logic into a separate public method with [GlobalSetup] attribute. You can use Target property to make it specific for selected benchmark. Example: [GlobalSetup(Target = nameof(MakeIrisPredictions))].
  3. Put the benchmarked code into a separate public method with [Benchmark] attribute. If the benchmark method computes some result, please return it from the benchmark. Harness will consume it to avoid dead code elimination.
  4. If given benchmark is a Training benchmark, please apply [Config(typeof(TrainConfig))] to the class. It will tell BenchmarkDotNet to run the benchmark only once in a dedicated process to mimic the real-world scenario for training.

Examples:

public class NonTrainingBenchmark
{
    [GlobalSetup(Target = nameof(TheBenchmark))]
    public void Setup() { /* setup logic goes here */ }

    [Benchmark]
    public SomeResult TheBenchmark() { /* benchmarked code goes here */  }
}

[Config(typeof(TrainConfig))]
public class TrainingBenchmark

Running the BenchmarksProjectIsNotBroken test

If your build is failing in the build machines, in the release configuration due to the BenchmarksProjectIsNotBroken test failing, you can debug this test locally by:

1- Building the solution in the release mode locally

build.cmd -configuration Release -performanceTest

2- Changing the configuration in Visual Studio from Debug -> Release 3- Changing the annotation in the BenchmarksProjectIsNotBroken to replace BenchmarkTheory with Theory, as below.

[Theory]
[MemberData(nameof(GetBenchmarks))]
public void BenchmarksProjectIsNotBroken(Type type)

4- Restart Visual Studio 5- Proceed to running the tests normally from the Test Explorer view.