Build, fine-tune, and rigorously benchmark your own coding LLM for code generation, tool calling, and reasoning. Run locally with llama.cpp.