知识卡片

普通读书笔记卡 · 6557.e · 生成模型:GPT-6

二值标记不能用单列取最大值

理解等级:中级

对形状[1,1]的Bernoulli样本沿唯一列argmax永远得到0,不能还原原来的0/1标记。应读取标量或显式阈值,而类别one-hot才适合在类别轴argmax。否则“生成的是打印体”日志可能只是解码辅助函数错误。

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

      is_handwritten = torch.argmax(is_handwritten, 1) #4 return img, digit, is_handwritten #4 #1 Choose a random example from the dataset. #2 Reshape the data. #3 Generate data that is encoded. #4 Generate (unencoded) data. Finally, we can run the training procedure. First, we’ll set up stochastic variational inference. We’ll first set up an instance of the Adam optimizer, which will handle optimization of the parameters in training_guide.

    参考来源

    - 位置:《Causal AI》第05章《连接因果与深度学习》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-5.xhtml) - 结论依据:摘录中关于二值标记不能用单列取最大值的定义、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:is_handwritten = torch.argmax(is_handwritten, 1) #4 return img, digit, is_handwritten #4 #1 Choose a random example from the dataset. #2 Reshape the data. #3 Generate data that is encoded. #4 Generate (unencoded) data. Finally, we can run the training procedure. First, we’ll set up stochastic variational inference. We’ll first set up an instance of the Adam optimizer, which will handle optimization of the parameters in training_guide.