ML.NET 3.0: Microsoft upgrades cross-platform framework to enhance deep learning and AI computing efficiency

Microsoft announced the launch of the cross-platform machine learning framework ML.NET 3.0 on November 29. This update mainly enhances the deep learning function, improves the data processing capabilities of ML.NET, and adds functions such as Intel oneDAL accelerated training technology and automatic machine learning

Required The rewritten content is: ▲ Picture source Microsoft

This site noticed that ML.NET 3.0 provides a number of deep learning functions, including "object detection" and "named entity recognition" and "Question and Answer Processing" etc.

Among them, "object detection" can locate and classify different types of entities in images. According to the official introduction, object detection is a computer vision task, and " "Image Classification" is closely related, but the classification is relatively more refined. When the image contains different types of objects, officials recommend using related functions.

The named entity recognition and question and answer processing are based on Microsoft's newly added TorchSharp API, which is a .NET library that claims to combine the latest technology of Microsoft Research with the Transformer neural network architecture in TorchSharp, and through the existing The text classification function of TorchSharp RoBERTa is used as the basis to achieve the above functions.

In addition, shortly after Microsoft released ML.NET 2.0, it announced that it would provide support for Intel's oneDAL accelerated training technology. This feature is currently available in ML.NET Debuted in 3.0, it can significantly accelerate the data analysis and machine learning process.

Microsoft has also updated ML.NET 3.0 Automatic Machine Learning (AutoML) function, which brings statement similarity, question and answer processing, object detection and other functions to assist developers in choosing the most suitable Suitable models and parameters make it easier for developers to design machine learning models.

This site also discovered that ML.NET now has continuous resource monitoring capabilities, and can monitor RAM and hard disk space usage through AutoML.IMonitor , making it easier for developers to control Long-term experiments can prevent the running process from crashing due to insufficient RAM or ROM, and it also allows developers to visually view various parameters of the process.

ML.NET 3.0 also integrates Tensor Primitives, a new set of APIs specifically used for tensor operations, which can further promote the application of .NET in artificial intelligence mathematical operations. This API not only uses the hardware's internal instruction set to accelerate computing efficiency, but also combines the principles and concepts of Generic Math, claiming to be a "powerful tool for developers to process complex mathematics and cumbersome data."

The content that needs to be rewritten is: ▲ Picture source Microsoft

The above is the detailed content of ML.NET 3.0: Microsoft upgrades cross-platform framework to enhance deep learning and AI computing efficiency. For more information, please follow other related articles on the PHP Chinese website!