search

DOsinga / deep_learning_cookbook

Deep Learning Cookbox
Apache License 2.0
689 stars 338 forks source link

doubt on movie recommendation tutorial #39

Closed anne1994 closed 5 years ago

anne1994 commented 5 years ago

Hi,

I am reading your book, and its great so far !

In the function movie_embedding_model-> embedding layer , we have used input_dim as the len(top_links) or len(movie_to_idx) .. how is this input dimension used?, what does it do?

In batch we send a tuple of ({linkids, movieids}, label ). In the book you mentioned that : allocates a vector of embedding_size for each possible input. I understand it to allocate vector of embedding_size to each possible ID sent in batch. How are we using the input_dim ?

DO correct me if i am thinking wrong.

Thanks. Anne.

anne1994 commented 5 years ago

ok i think input dim is len(top_links) for each id in batch, but the output dim allocated would be embedding_size. can embedding size be same as input_dim?