知识卡片

普通读书笔记卡 · 6749.a · 生成模型:GPT-6

随机输入前向只证明形状路径可通

理解等级:中级

用合法范围整数构造源目标并看输出shape,可发现词表边界、批次和张量轴错误。它不证明mask、标签对齐、翻译能力或训练正确;应把这一机械检查与数据语义和独立评价区分。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      Notice that you set the sizes of your source and target vocabularies in the constructor. You also pass the indices for the source and target padding tokens for the model to use in preparing the source, targets, and associated masking sequences. Now that you have the model defined, take a moment to do a quick sanity check to make sure there are no obvious coding

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第09章《可堆叠的Transformer》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Natural Language Processing in Action, Second Edition/_epub-src/OEBPS/Text/chapter-9.html) - 结论依据:摘录中关于随机输入前向只证明形状路径可通的定义、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:Notice that you set the sizes of your source and target vocabularies in the constructor. You also pass the indices for the source and target padding tokens for the model to use in preparing the source, targets, and associated masking sequences. Now that you have the model defined, take a moment to do a quick sanity check to make sure there are no obvious coding