Reproducible shuffling of training data (fix portability issue #85 for macOS)

tesseract-ocr / tesstrain

Train Tesseract LSTM with make

Apache License 2.0

637 stars 188 forks source link

Reproducible shuffling of training data (fix portability issue #85 for macOS) #86

Closed stweil closed 5 years ago

stweil commented 5 years ago

sort -R depends on OpenSSL and does not work on macOS because of missing named pipes, so replace it with a well defined pseudo random shuffling Python script.

Signed-off-by: Stefan Weil sw@weilnetz.de

stweil commented 5 years ago

This fixes issue #85.

stweil commented 5 years ago

We already depend on python3 for generate_line_box.py, so I prefer to keep that uniform.