逐步了解 LLM 的内部机制——从分词到注意力机制再到推理优化。
Learn LLM internals step by step - from tokenization to attention to inference optimization.