The RTC takes ethics seriously, and we adopt the position that language models are a derivative work of their dataset. Therefore, by using the Novum Corpus Kirkensis, you take on the obligation to release the resulting work under the most restrictive license in the whole: CC-BY-SA-4.0.

The Novum Corpus Kirkensis

The Novum Corpus Kirkensis is a sequel to the Corpus Kirkensis which preserves the principles upon which its predecessor was built: full legality, ethical provenance, and license compatibility. The bulk of the data is under the Creative Commons, with non-CC rows being made up of text that falls either under more permissive compatible licenses, the "matter-of-fact" copyright exception, or can be included as part of a collection.

Much like the previous installment, this one was not built with the intent to make models more "knowledgeable" about the world, but to have them generate interesting content. Therefore, it is heavily reliant on internet fan communities and "taboo" works, along with religious sources. To make the collection apt for general training of tiny language models, it also features quite a bit of code, along with various textbooks.

Source credits

Community sources

This collection would not be possible without the hard work of communities who release the fruits of their collective efforts under the Creative Commons:

It would also not be possible without the morally reprehensible but entertaining work of the following:

Archival sources

The following archival sources were used:

Individual Authors

Some individual authors have contributed to a substantial part of this corpus, and are uniquely recognized on that basis:

MorganAW
Mark Leach, who deserves more credit for writing what is possibly the second-largest piece of western literature (right behind that one Loud House fanfic) and releasing all of it under the Creative Commons: Marienbad my Love.

Code sources

The following collections directly containing computer code were used:

For specific credits related to all the code in this collection, please verify their individual credits. GitHub repository names are also featured in our "Authors" field.

Size

How many tokens? How many tokens? I want tokens, how many tokens are there? How much data can I extract from this? If each token was a piece of paper how many trees could I kill? Can you please tell me how many tokens are in this dataset? My company is forcing me to tokenmaxx and I need to waste their money by doing text processing on a useless dataset so can you please tell me how many tokens there are?

We estimate an amount of around 2.5 billion, depending on the specific vocabulary you're using.