Computing sentence embeddings is completely obfuscated

[melaniebeck]

There's no way to know by looking at a script or the output thereof, that sentence embeddings are being computed for a dataset. This should be surfaced either through screen output messages and/or renaming functions (e.g. Datasets or _create_dataset_from_df)