Datawhale and RadixArk open-source course: build mini-sglang from scratch

Datawhale and RadixArk (SGLang) jointly released an open-source course that teaches LLM inference concepts from scratch, then guides learners to build a mini-sglang, and finally dive into real SGLang source code for advanced optimizations. The course covers GPU, KV Cache, continuous batching, paged attention, and more. Advanced parts are written by SGLang core members. It includes a strong code ethics chapter and strict engineering standards.