Closed ShroukMansour closed 2 years ago
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
Hard to tell without sample data which allow for reproducing the problem.
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
I've finetuned an Arabic model using the same way reported at tesstrain to train with images, but when I test it it get the results fully reversed I used this command to start training
make training LANG_TYPE=RTL MODEL_NAME=ara_img TESSDATA=/home/shrouk/work/vlens/tesseract/tesstrain/data/finetune_from \ START_MODEL=ara_plusminus_addr_2 TRAIN_TYPE=FineTune
and this command to create the .traineddata file
make traineddata MODEL_NAME=ara_img CHECKPOINT_FILES=data/ara_img/checkpoints/ara_img0_1898.checkpoint
Is there anything that I have to do to avoid this?