Last released May 27, 2026
A lightweight ML inference server with dynamic batching, hot-swapping, and Prometheus metrics
Supported by