知识卡片

普通读书笔记卡 · 6497.a · 生成模型:GPT-6

分词往返检查输入边界

理解等级:初级

在尚未接入模型时,先对同一文本执行 encode 再 decode,检查分词配置与输入是否匹配。往返一致只验证该样本的文本转换,不验证模型能力、所有 Unicode 边界或特殊标记安全。特殊标记如何编码仍需单独检查。

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

      Since we have not loaded the LLM yet (the central component shown in figure 2.5), we will first do a simpler dry run using just the tokenizer. Specifically, we will do a tokenization round-trip, that is, we will encode a text into token IDs and then decode those IDs back into text, as illustrated in figure 2.6.

    参考来源

    - 位置:《Build a Reasoning Model (From Scratch)》第02章《使用预训练模型生成文本》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Build a Reasoning Model (From Scratch)/_epub-src/OEBPS/Text/chapter-2.html) - 结论依据:摘录中关于分词往返检查输入边界的定义、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:Since we have not loaded the LLM yet (the central component shown in figure 2.5), we will first do a simpler dry run using just the tokenizer. Specifically, we will do a tokenization round-trip, that is, we will encode a text into token IDs and then decode those IDs back into text, as illustrated in figure 2.6.