Thank you > On Dec 11, 2020, at 7:13 AM, shree <shreesh...@gmail.com> wrote: > > For Sanskrit in Devanagari and IAST, you can try the traineddata files from > https://github.com/Shreeshrii/tesstrain-Sanskrit-IAST > <https://github.com/Shreeshrii/tesstrain-Sanskrit-IAST> > > For Sanskrit alone, you can try the traineddata file from > https://github.com/Shreeshrii/tesstrain-sanPlusMinus > <https://github.com/Shreeshrii/tesstrain-sanPlusMinus> Thanks I’m not sure exactly what to do with these links or the files they access?
> These have the float models, to improve speed they can be compressed using > `combine_tessdata -c` Sorry but I don’t know what all this means? > I would appreciate feedback on how well these work compared to the official > `san` and `Devanagari` files. I would be happy to give feedback. I have been using san. But was unaware that you can also use Devanagari. What is the difference? > I had done some training for grantha using the Noto fonts. But to be usable, > I need more training data of actual line images and their groundtruth > transcription. If you can provide that, I'll be happy to retrain it. I would be happy to provide more examples of Grantha. If you tell me how to make “actual line images” and “groundtruth transcription”? I can make images of Grantha. Let me know the format? I can also provide the transliteration in IAST or ISO15919 or some other Indic script like Devanagari. Sorry if I show my lack of understanding here. -- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To unsubscribe from this group and stop receiving emails from it, send an email to tesseract-ocr+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/6B83B287-A7FD-484F-B958-B2988C83286B%40gmail.com.
