I had raised a question regarding using Google Universal Sentence Encoder earlier, but I think that embedding is not yet ready to be used with Ludwig. I would like to ask a question on how I can use any embedding layer to vectorize my text (paragraphs). If you can share a sample yaml file, that would be great.
Ludwig already vectorizes your text and looks up embeddings and does everything for you. You can specify pretrained embeddings too. Please look at the sequential encoders for more details.