GitHub - samir-sarwar/samir-vllm: small LLM inference engine from scratch, w/ 6+ hours of youtube tutorials & 85+ pages of handwritten notes included

Samir Sarwar has released samir-vllm, a small from-scratch LLM inference engine, on GitHub alongside more than six hours of YouTube tutorials and over 85 pages of handwritten notes. The project targets developers who want to understand how inference engines work internally rather than treat them as black boxes, pairing a compact codebase with unusually extensive educational material. It matters because most inference stacks are large, complex systems, leaving a gap for a minimal reference implementation that is practical to read end to end. Watch whether the repository attracts contributors and course-style follow-ups that extend it beyond a teaching tool.