OCR-D / ocrd-website

Creative Commons Attribution 4.0 International
24 stars 7 forks source link

recommended workflows: make runnable #297

Open bertsky opened 2 years ago

bertsky commented 2 years ago

Recently, our Calamari wrapper changed its parameterization: it no longer supports using checkpoint but requires checkpoint_dir instead. So checkpoint=qurator-gt4histocr-1.0/\*.ckpt.json just became checkpoint_dir=qurator-gt4histocr-1.0.

Also, there is still a reference to the (unregistered, outdated) Tesseract model GT4HistOCR_50000000.997_191951.

Plus IMO, to make this usable out of the box, there should be a direct mention of how to download the respective models via resmgr (prior to and independent of processing):

ocrd resmgr download ocrd-tesserocr-recognize Fraktur_GT4HistOCR.traineddata
ocrd resmgr download ocrd-calamari-recognize qurator-gt4histocr-1.0
bertsky commented 1 year ago

ping @lena-hinrichsen