Brett Parker

@aslotte what is your thinking here? To have an overload method for `GetBestRun` where the user can request what the metric is based on, e.g closest to 0 (ascending), closest...

There is no unit test project for SQL server either. Required? @aslotte Looking at this further, the SQLite `[TestCategory("UnitTests")]` are the same as `MLOpsBuilderExtensionTests.cs` stored in the SQLServer integration test....

@aslotte I was thinking of a registered model. This would then give an ability to filter based on label. Potentially a great feature for the UI side down the line.

@aslotte I'd like to look into this. Looking for your guidance on what you would like for a foundation.

@aslotte what did you have in mind for this?

@aslotte do you agree with this? Are you happy for me to go ahead and start creating the examples?

On first glance, the following metrics cannot be handled currently (Investigation once examples are created); MulticlassClassification - TopKPredictionCount - returns Int32 Possible just cast to double?

https://github.com/marketplace/actions/continuous-benchmark-net Support added for Benchmark.NET

Same issue for me. Downloading the release version for Windows (v1.8.1) worked but the pip install version crashes (v1.8.6).

The commit here piggbacks the commandline button press. I think it should have its own.