Included in ML.NET 0.6 is a new API that helps developers create models. Microsoft says this is the first release of ML.NET APIs. Developers will find more flexible models and simpler code workflow. Microsoft has also included a tool for scoring pre-trained INNX Models and general performance improvements across the platform. Below is the full changelog for ML.NET 6.0:
New API for building and using machine learning models: Our main focus was releasing the first iteration of new ML.NET APIs for building and consuming models. These new, more flexible, APIs enable new tasks and code workflow that weren’t possible with the previous LearningPipeline API. We are starting to deprecate the current LearningPipeline API.
Ability to score pre-trained ONNX Models: Many scenarios like Image Classification, Speech to Text, and translation benefit from using predictions from deep learning models. In ML.NET 0.5 we added support for using TensorFlow models. Now in ML.NET 0.6 we’ve added support for getting predictions from ONNX models. Significant performance improvements for model prediction, .NET type system consistency, and more: We know that application performance is critical. In this release, we’ve increased getting model predictions performance 100x or more.
Improving Machine Learning
ML.NET is a framework Microsoft used internally for nearly a decade. Earlier this year, the company expanded the platform to Windows, Bing, and Azure. Developers were given access to the solution for the first time. With the platform, .NET developers create their own models and infuse custom machine learning without hassle. They don’t need to be an expert in the field or have any prior experience. It simplifies the process and supports a couple of tasks already.
