← 返回卡片

原书出处

Natural Language Processing in Action, Second Edition

Second Edition · 核验版本的摘录快照 · 导入于 2026-11-14T20:22:29.266Z

Start-of-sequence token —Typically, "<SOS>" or "<s>" End-of-sequence token —Typically, "<EOS>" or "</s>" Out-of-vocabulary (unknown) token —Typically, "<OOV>" or "<unk>" Mask token —Typically, "<mask>" Padding token —Typically, "<pad>" The start-of-sequence token is used to trigger the decoder to generate a token that is suitable for the first token in a sequence. And many generative problems will require you to have an end-of-sequence token, so the decoder

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/Natural Language Processing in Action, Second Edition/_epub-src/OEBPS/Text/chapter-9.html

EPUB SHA-256:0e9a35e49ca446e322e5524e53f39fb6bee5b2b23d9cda10ef052aca770b0b36