This script is used to retrieve document embeddings for the feuilleton novels dataset (ca. 1,400 texts)
For directions on running it, see info.md
Script should run with any sentencetransformer compatible model, automatically retrieving the max_token length used for pooling.