Track the number of tokens in each post and the order the embeddings

are in within the post, as an initial step towards storing off the
actual range of characters in the post that the embedding corresponds
to