Correct answer to last question should be
“Equal to dimension of the embedding times your vocabulary”
instead of
“Equal to your vocabulary times the dimension of the embedding”
Or?
Hi Fabian_Muller,
In practice this depends on the chosen implementation. In the case of this course, as you can see in the video ‘Other Layers’, the implementation uses a number of rows equal to the size of the vocabulary and a number of columns equal to the dimension of the embedding.