recommendation results not making sense

[riyaj8888]

`model = tf.saved_model.load(model_path)

request = {"userid":user_id}

scores,rec_item_ids = model(request)
`

here rec_item_ids are all same and their scores as well .

any help why this is happening

[mustfkeskin]

My suggestion is to remove the embeddings and implement the ann module yourself.
While doing these, you need to make sure that the indices do not shift.