[tesseract-ocr] No best char error rate

Hi all, I trained tesseract 4.1.1 with around 28000 data in which it's training started with best error of 4.187 and I got best error of 0.057. Now I added further data and final training data is around 1lac images. But the training starts with worst error rate of around 274 and never goes below

[tesseract-ocr] Advice for working with low-resolution images

In the last few days it has been a joyful experience getting to know Tesseract and using it in one of my projects. Thank you all. The challenge I am facing is that the OCR detection is not correct on some of the PDF files I am working with and I am writing to this group seeking advice on how I