返回全部动态
生成配置详解：温度、top-k、top-p 与测试时计算
原标题：Generation configurations: temperature, top-k, top-p, and test time compute
AI 摘要
Chip Huyen 在其博客中深入探讨了 AI 模型的生成配置，包括温度、top-k、top-p 以及测试时计算。文章解释了模型生成文本的概率性采样过程，以及如何通过调整温度等参数来平衡创造性与一致性。此外，文章还讨论了增加推理计算量（如多次采样）对提升模型性能的作用，并提及了结构化输出的方法。
以上摘要由 AI 生成，可能存在误差。事实请以原文为准。
正文节选
ML models are probabilistic. Imagine that you want to know what’s the best cuisine in the world. If you ask someone this question twice, a minute apart, their answers both times should be the same. If you ask a model the same question twice, its answer can change. If the model thinks that Vietnamese cuisine has a 70% chance of being the best cuisine and Italian cuisine has a 30% chance, it’ll answer “Vietnamese” 70% of the time, and “Italian” 30%. This probabilistic nature makes AI great for cre
发布时间：2024-01-16 08:00
抓取时间：2026-08-02 00:21
来源机构：Chip Huyen