SD Occasions Open-Supply Challenge of the Week: NeoML


This week’s highlighted open-source venture is NeoML, a library for constructing, coaching, and deploying machine studying fashions.

In response to ABBYY, the corporate that created NeoML, the venture is finest fitted to functions operating in cloud environments, on desktop, and on cellular gadgets. Builders can use the library for object identification, classification, semantic segmentation, verification, and predictive modeling. 

An instance use case is a financial institution utilizing NeoML to develop fashions to handle credit score threat and predict buyer churn, ABBYY defined. 

At the moment NeoML helps C++, Java, and Goal-C. ABBYY plans on including help for Python quickly. The library additionally helps over 100 layer sorts, has greater than 20 conventional machine studying algorithms, and is absolutely cross-platform. 

As well as, NeoML helps the Open Neural Community Change (ONNX), which is an ecosystem for interoperable machine studying fashions. ONNX improves instrument compatibility, which makes it simpler for builders to make use of varied instruments to fulfill their targets, ABBYY defined. 

“The launch of NeoML displays our dedication to contribute to industry-wide AI innovation,” stated Ivan Yamshchikov, AI Evangelist at ABBYY. “ABBYY has a confirmed monitor report of technological innovation with over 400 patents and patent functions. Sharing our framework permits builders to leverage its inference velocity, cross-platform capabilities, and particularly its potential on cellular gadgets, whereas their suggestions and contribution will develop and enhance the library. We’re thrilled to advertise developments in AI and help machine studying being utilized to more and more high-value and impactful use circumstances.”

The venture will be discovered on GitHub