[FEA] Get only embeddings from model.item_embeddings() #255
Description
🚀 Feature request
when we run the following command we do not only get item embeddings but also item feature columns in the item_embeddings dataframe. It'd be useful to add an arg in the item_embeddings() to exclude other columns.
item_embs = model.item_embeddings(Dataset(item_features, schema=schema), batch_size=1024)
item_embs_df = item_embs.compute(scheduler="synchronous")
Motivation
We save item_embeddings for E-2-E and we only want to have the embedding columns. In doing so, we need an extra manual process to drop the item features from the item-embeddings dataframe.