知识卡片
多标签按固定宽度多热向量组织
每条评论可对应多个0或1,输出列应与标签列表逐项一致,允许全零。包装器的text与labels字段只是组织格式,仍需检查各标签含义、缺失值与分类损失,不能把一个单一类别编号塞进多标签协议。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
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
There’s one last data extract, transform, load (ETL) task for you to deal with: adding a wrapper function to ensure the batches of examples passed to your transformer have the right shape and content. You are going to use the simpletransformers library, which provides wrappers for various Hugging Face models designed for classification tasks, including multi-label classification—not to be confused with multiclass or multi-output classification
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:摘录中关于多标签按固定宽度多热向量组织的定义、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:There’s one last data extract, transform, load (ETL) task for you to deal with: adding a wrapper function to ensure the batches of examples passed to your transformer have the right shape and content. You are going to use the simpletransformers library, which provides wrappers for various Hugging Face models designed for classification tasks, including multi-label classification—not to be confused with multiclass or multi-output classification