TorchKM: A GPU-Oriented Library for Kernel Learning and Model Selection

In three linesTorchKM is an open-source library for kernel machines (SVM, kernel logistic regression, kernel quantile regression) with GPU acceleration. Scikit-learn-style API, GPU-friendly linear algebra optimization, and intelligent matrix operation reuse. Code available on GitHub, PyPI installation.

Summary generated by Claude — human-verified