I have an issue with generating embeddings for my dataset. It consists of about 16 000 000 reddit comments (their bodies + some negligible metadata). I have them stored as a CSV file, from which I’m generating a Pandas dataframe using pd.read_csv().