Am using easy OCR from the link below https://github.com/JaidedAI/EasyOCR
I have custom dataset of 25000 images for training and 1000 images for validation in all_data folder generated. Max image height is 64 px. Now I want to fine tune english_g2.pth model using custom dataset. I need to alter en_filtered_config.yaml file to fine tune pretrained model using custom dataset of only alphabets.
After training, I have new pth file placed in home/easyocr folder. There ll be 2 more files with same name as pth file. I ll use example files downloaded from easyocr git. If there is any update to be done in those files let me know.
Folder structure remains the same as GitHub easyocr repo.