知识卡片
参数规模不单独决定适用性
“大型”没有跨时期不变的参数阈值,本书采用的十亿参数口径只是工作约定。模型质量还受数据、目标、架构和适配影响,不能把参数更多直接等同于每个应用更好。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 大语言模型入门
原书出处 · 1 处
- · 查看原书依据
There is no accepted convention about when a language model is considered “large.” In fact, as the largest models get even larger, some models that would have been designated as LLMs only a couple of years ago are now termed small language models (SLMs). In this book, we will remain generous and continue to refer to all language models over a billion parameters as “large.”
参考来源
- 位置:《Designing Large Language Model Applications》第01章《大语言模型入门》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Designing Large Language Model Applications/_epub-src/OEBPS/ch01.html)
- 结论依据:摘录中关于参数规模不单独决定适用性的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:There is no accepted convention about when a language model is considered “large.” In fact, as the largest models get even larger, some models that would have been designated as LLMs only a couple of years ago are now termed small language models (SLMs). In this book, we will remain generous and continue to refer to all language models over a billion parameters as “large.”