Extracting trained embeddings in a Retrieval Task

General Discussion
,
#1

Following the tutorial Recommending movies: retrieval  |  TensorFlow Recommenders

How do you extract the trained embeddings of a trained model?

I know you can use model.get_layer(‘sequential_##’).get_weights()[0] to output the embeddings. How do you index this to unique_user_ids/unique_movie_titles, also taking into account the +1 padding on the user_model and movie_model? Does it take the same order?

user_model = tf.keras.Sequential([
  tf.keras.layers.StringLookup(
      vocabulary=unique_user_ids, mask_token=None),
  # We add an additional embedding to account for unknown tokens.
  tf.keras.layers.Embedding(len(unique_user_ids) + 1, embedding_dimension)
])

movie_model = tf.keras.Sequential([
  tf.keras.layers.StringLookup(
      vocabulary=unique_movie_titles, mask_token=None),
  tf.keras.layers.Embedding(len(unique_movie_titles) + 1, embedding_dimension)
])
#3

maybe @Wei_Wei can help here