Embedding layer - where is 3rd argument?

Course Q&A TensorFlow Developer Professional Natural Language Processing in TensorFlow
1

image
image1884×166 85.2 KB

In previous video, instructor showed that 3 arguments went inside Embedding layer

  1. vocab size,
  2. embedding dimension
  3. maximum length of sentence.

Q1. But in this week’s videos, he’s using only 2 arguments. Why is maximum length not specified as 3rd argument like was done in last week?
Q2. What is the reason for choosing 64 cells in LSTM? Why not 20 or 25 or 100 or some other number?

3

To Q1: Have you compared the summary() of model regarding the layer of Embedding?

When possible, would you mind sharing a public notebook (copy of lab) ?