I have used the embedding endpoint to convert a piece of text to an embedding vector. For a dataset containing a number of text entries, converting one piece by one in a loop takes quite a bit of time. I wonder if the embedding endpoint can be used to convert a batch of text pieces to embeddings at one time.