lstmeval rtl Problem.

[Ishak DÖLEK]

Hello
I trained my model with lstmtraining. I want to test it with lstmeval.
When I tested with lstmeval , the ground truth text is aligned from right to left. But  OCR  text is aligned from left to right.
Hence I cannot calculate the character error rate and word error rate.
How do I solve this problem.

Thanks in advance.

Regards.