mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2025-06-07 18:02:40 +08:00
Code Issues Actions 10 Packages Projects Releases Wiki Activity

Updated TrainingTesseract 4.00 (markdown)

Amit D 2019-06-12 06:51:27 +03:00
parent 364b5ae735
commit 480fbf4e15
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
TrainingTesseract-4.00.md

@ -332,7 +332,7 @@ scripts.
For Stochastic Gradient Descent to work properly, the training data is supposed For Stochastic Gradient Descent to work properly, the training data is supposed
to be randomly shuffled across all the sample files, so the trainer can read its to be randomly shuffled across all the sample files, so the trainer can read its
way through each file in turn and go back to the first one when it reaches the way through each file in turn and go back to the first one when it reaches the
end. This is entirely contrary to the way base Tesseract was trained! end.
If using the rendering code, (via `tesstrain.sh`) then it will shuffle the If using the rendering code, (via `tesstrain.sh`) then it will shuffle the
sample text lines within each file, but you will get a set of files, each sample text lines within each file, but you will get a set of files, each