← 返回卡片

原书出处

Designing Large Language Model Applications

First Edition · 核验版本的摘录快照 · 导入于 2026-11-14T13:54:13.936Z

It is suspected that a large number of models may have been trained on data contaminated with the GSM8K dataset. Explore the GSM8K dataset, feed only part of the question, and check if the models you evaluated during the previous exercises correctly complete the question. Also, change the numbers in the problems and verify if the performance remains the same.

此预览只包含定位所需的摘录和上下文。

原书定位详情

来源:a19c7933-4eda-4176-aebd-30fe7e1ad292

正文位置:extracted/Designing Large Language Model Applications/_epub-src/OEBPS/ch05.html

EPUB SHA-256:1b4d2347fec2f949fcddfbaf4fad597a1ae20887a096cfae64c59a619b4781bf