Distillation
概念
训练与推理
Teaching a small model to imitate a big one.
您正在查看一条概念解释记录：该术语的通俗含义，以及它所关联的各条目。
Distillation trains a compact student on a larger teacher's outputs, keeping most capability at a fraction of the serving cost — the standard way 'mini' and 'flash' tiers are made. It is also a flashpoint: training on a rival API's outputs against its terms is industrial-scale imitation, an accusation that surfaced prominently around DeepSeek's rise.
它如何与整个技术栈相连
本记录的关联链接一览——点击分组可直接跳转。
来源
Distillation paper (Hinton et al.) ↗
记录审阅截至 2026-01-15 — 查看 方法论.