How can I use DDPM for prediction of next term in sequence?

Deep Learning
1

If I have vector of size (N, D) , it may be embeddings of the words of a sentence so their order is important. Can I train a diffusion model to predict the (N+1)th term ?