Show HN: Go-CoreML – Go Bindings for Apple's CoreML with Neural Engine Support (github.com)

A new project, Go-CoreML, has emerged, providing Go bindings to Apple's CoreML framework designed for optimized machine learning inference on Apple Silicon devices. This integration allows developers to leverage Apple's Neural Engine (ANE) and utilize Metal GPU acceleration, enhancing the performance of ML models on macOS. Key features include a programmatic model construction interface with the Machine Learning Intermediate Language (MIL), low-level interaction with CoreML for tensor manipulation and model inference, and support for various ML operations like relu, matmul, and pooling. The significance of Go-CoreML lies in its potential to bridge the gap between the Go programming language and Apple's robust ML ecosystem, opening avenues for Go developers to build and deploy high-performance machine learning applications. As the API is currently in alpha, further refinements are expected, with plans for deeper integration with GoMLX. This could stimulate faster adoption of machine learning capabilities within the Go community, making high-performance AI accessible to a broader range of developers and applications. The project requires macOS 12.0+ and Go 1.21+, making it aligned with modern development environments.