The Embedding is a matrix of weights / word vectors of size (vocab size, emb_dim)
and the Input of a single training example is of size(max_len) and consists of indices. Now for a single training example this generates an output of size(max_len, emb_dim). And this output is specific to the words for which indices are given.
How does this work?
While I was doing the exercises, I got a doubt of how Embeddings generate such output? I did not get any reply to this query.
I have also successfully completed the DL specialization and I am currently trying to apply my learnings through projects.