Cannot find OCR model file

[tulang-nb2022]

I am trying to run ./ocropus -rpred -m en-default.pyrnn.gz book/0001/*.png

as part of a simple tutorial but the program complains that

ERROR: Cannot find OCR model file:en-default.pyrnn.gz

ERROR: Download a model and put it into:/usr/local/share/ocropus/

ERROR: (Or override the location with OCROPUS_DATA.)

I have put that file in both "ocropus_venv\ocropy-master\models" and "ocropus_venv\share\ocropus" folders in both the gunzipped and ungunzipped forms and the system still complains. All other previous commands to perform binarization and segmentation passed as expected.

• Git revision of ocropy: most recent (just dl'ed couple of days ago)
• Python verson 2.7.13
• OS: Windows 7 64-bit

Many thanks in advance!!

[zuphilip]

Did you run ./setup.py again?

[rogerfitz]

Just had same issue. tried wget from command line and http://www.tmbdev.net/en-default.pyrnn.gz did not resolve. Tried again in browser same thing...

But then it looks like the server "woke up" and started responding instantly -- after the web browser request. Potentially took 1 minute. I've experienced this coming from using heroku free servers that go to sleep if no requests within a time window. not sure if that is the same reason why tmbdev.net is acting funny

[zuphilip]

I saved the models also here https://github.com/zuphilip/ocropy-models if problems remains.

[tulang-nb2022]

Never mind, I got it to work by specifying "current directory" with ./en-default.pyrnn

Thanks for the response though!!