Context I found the translations of the Old Testament often lacking to the extent it was not conveying the original sense when compared to the original ancient Greek text. So it is only logical to use the original text for studies. Surprisingly it seems the ancient Greek text was not available in a format suitable for NLP until yet. [According to Wikipedia](https://en.wikipedia.org/wiki/Septuagint) this Koine Greek text is believed to be itself a translation from Hebrew. Yet it is the most ancient written collection of the Old Testament books we have. The Hebrew versions we have were written only hundreds of years later as copies of a now lost original Hebrew source. The New Testament is making direct references to this Greek text and so it is of more importance to have it available for studies. Content cut -d";" -f2 septuagint.csv|wc 1180 lines 653478 words 7692966 characters Acknowledgements Thanks to @Dan at OpenDataScience Slack for pointing me to the plain text source for Septuagint online. Inspiration It would be great to create an NLP dataset class out of this text with for example collection of stopwords, own names, morphological library etc. Current Spacy library is capable of processing the text but is not capturing Named Entities to a large extent.