I have trained the embeddings for content-based recommendations using the command:
starspace train -trainFile prod_by_order__products.txt -model docspace -trainMode 1 -fileFormat labelDoc
I would like to get embeddings for the given products. As far as I understood, this can be done via:
embed_doc docspace input.txt
Unfortunately, looks like the application requires basedoc. Why does it need the basedoc to compute embeddings? I don't see any use of basedoc in the embed_doc sources.
Start to load a trained starspace model. STARSPACE-2018-2 Model loaded. Must provide base labels when label is featured.
I have trained the embeddings for content-based recommendations using the command:
starspace train -trainFile prod_by_order__products.txt -model docspace -trainMode 1 -fileFormat labelDocI would like to get embeddings for the given products. As far as I understood, this can be done via:
embed_doc docspace input.txtUnfortunately, looks like the application requires basedoc. Why does it need the basedoc to compute embeddings? I don't see any use of basedoc in the embed_doc sources.
Start to load a trained starspace model. STARSPACE-2018-2 Model loaded. Must provide base labels when label is featured.Is it a bug or I'm doing something wrong?