Core ML Tools (coremltools) README

coremltools is Apple's converter and optimizer for the Core ML model format. Its README names the supported source frameworks explicitly: TensorFlow 1.x, TensorFlow 2.x, PyTorch, and non-neural-network frameworks including scikit-learn, XGBoost and LibSVM. The capability list is: convert trained models to Core ML, read/write/optimize Core ML models, and verify conversion on macOS by making predictions with Core ML - after which the models are integrated into an app through Xcode. The README's own description of Core ML's execution model is the hardware-constraint story in one sentence: Core ML 'optimizes on-device performance by leveraging the CPU, GPU, and Neural Engine while minimizing its memory footprint and power c[onsumption]' - i.e. it is a whole-silicon-partition dispatcher, not a GPU kernel library, and it decides for you which of the three engines runs your op. No machine-detectable license (NOASSERTION); the repo's primary language is Jupyter Notebook and its CI badge still points at gitlab.com/coremltools1/coremltools, i.e. the project is developed on GitLab and mirrored to GitHub. Repo metadata via GitHub API 2026-10-04: created 2024-04-18, pushed 2026-09-25, 7005 stars, 540 forks, 17 open issues.