No available model hosting platforms detected. Please check your network connection
#2
by
hwang136
- opened
The default model used in pipeline is PP-OCRv5_server_rec, and you can also use the local model file by argument text_recognition_model_dir
However text_recognition_model_dir did not help when I set text_recognition_model_dir explicitly it did not work at all.
Great model I have to say but this PaddleOCR dependency is so lame and full of segmentaion errors when it comes to initialization also the lack of support of local model make it impossible to integrate with other project.