Inference runtime offering GPU-class performance on CPUs and APIs to integrate ML into your application - View it on GitHub