I doesn't understand this please help me

Course Q&A Deep Learning Specialization Sequence Models
,
1

why we does not use k directly instead of 2i in positional Encoding Transformer network lab

image
image1355×396 19.3 KB

2

Hello @BISHWANATH_JANA,

image does not mean k = 2i.

// means floor division. For example, if k = 3, then k // 2 = 3 // 2 = 1.

Cheers,
Raymond

1 Like
3

Thank You

1 Like