参数Parameter
Parameter
解决什么问题What problem it solves
模型里可调的数字，训练就是不断调整它们。参数越多，一般越「博学」。The numbers the training process adjusts; everything the model knows lives in them.
这是什么What it is
模型内部的数值参数，是模型学到的知识与能力的载体。
常说的「70 亿参数」指可训练参数的数量级，不等于质量。
Weights and biases inside the network. Model size usually refers to how many there are, and larger counts generally mean more capacity.
为什么需要知道Why it matters
它是你理解「为什么本地跑不动大模型」的关键。参数量直接决定要占多少显存。It grounds abstract claims: a "7B model" is a statement about parameter count, not about being better than everything else.
现在能做什么What you can do now
11 系列 16 规格的实测体积Measured sizes: 16 specs, 11 series ↗
回到名词地图看第1站的其他 34 个词 →Back to the map: the other 34 terms in station 1 →
出处Sources
- 11 系列 16 规格的实测体积Measured sizes: 16 specs, 11 series ↗