The deep learning library fastai uses this. Their source code is also their documentation. They use something called "nbdev" to accomplish this. And I have also seen other projects both related and unrelated to deep learning adopt it. This idea stems from Knuth.
I am on a tangent here, but I don't think that Transformers will have very long term effects. It is highly resource intensive, and only a few company can take benefit for it.
I know HuggingFace has been a gamechanger providing API calls for big models, and providing very practically models like distilBERT, but still, Transformers don't seem practical to me.
I also find reading multiple books in parallel really helpful. In this way, you can read multiple kinds of books.
For example, I am currently reading a "mathematical book"- Jordan Ellenberg's 'How Not to Be Wrong', a big fat novel in my native language, and Rumi's poetry in English.
It get's tiring to read serious non-fiction for multiple hours but there is still appetite for reading left. You should switch to a fiction then. And read poetry and book like Meditations by Aurelias now and then, maybe not even linearly.
I have found this combination to work for me really well.
> I find the idea of a royal family far more inspiring and wholesome than the celebrity culture which has overtaken the West.
Some of the celebrities are at least self made. Rich assh0les get to live in perpetual luxury just because of birth seems unfortunate.
I respect the Kim K more than the queen of England (although I absolutely hate Kim K) because Kim K is self-made, and the queen is there because of birth.
Relative worked for the biggest bank in India. They had to use Oracle's Office suite. The relative and his colleagues swore up and down that it was worst software (not just the worst office suite) they have ever used.