llama2_aided_tesseract VS tools

I made a tool like that, and I bet with a more powerful LLM like GPT4, and perhaps a better baseline OCR tool (like GPT4 vision), it could work really well for this sort of thing:

https://github.com/Dicklesworthstone/llama2_aided_tesseract

This violates the "One Tool" constraint that OP requested, but the Standard Ebooks tool chain is available on Github for anyone interested: https://github.com/standardebooks/tools

The code is GPL-3 and the templates are CC0: https://github.com/standardebooks/tools/blob/master/LICENSE....

Feel free to ask on the mailing list if you have any questions, more likely to be picked up there than in a random HN thread :)

I'll give a shoutout to some other excellent software.

The first is the "Standard Ebooks"[1] toolset, which is a suite of Python scripts to create, process, and build ebooks in all common formats. The results on the Standard Ebooks site speak for themselves. They're impeccable in every way, and far better than many big name, commercially produced efforts.

GitHub: https://github.com/standardebooks/tools

This question comes up a lot. The source to our production pipeline is GPLed and freely available,[1] but the biggest part of why we produce good work is that we have a high quality manual of style.[2] Unfortunately, that second part is very specific to English, and that’s the difficult part to replicate for other languages.

[1] https://github.com/standardebooks/tools/

[2] https://standardebooks.org/manual/