Tesseract Segmentation fault during fine tuning with fast traineddata

[ajinkya933]

Environment

• Tesseract Version: 4.0.0-324-gb67f

• Platform: Ubuntu16.04 64-bit

Current Behavior:

I am following TrainingTesseract 4.00 tutorial (https://github.com/tesseract-ocr/tesseract/wiki/TrainingTesseract-4.00). I am on a point where you Fine Tune for ± a few characters. This is the error that I am observing:

ajinkya@ajinkya-H310M-S2:~/Documents/tesseract$ src/training/lstmtraining --model_output ~/tesstutorial/trainplusminus/plusminus --continue_from ~/tesstutorial/trainplusminus/eng.lstm --traineddata ~/tesstutorial/trainplusminus/eng/eng.traineddata --old_traineddata tessdata/best/eng.traineddata --train_listfile ~/tesstutorial/trainplusminus/eng.training_files.txt --max_iterations 3600
Loaded file /home/ajinkya/tesstutorial/trainplusminus/eng.lstm, unpacking...
Warning: LSTMTrainer deserialized an LSTMRecognizer!
Code range changed from 111 to 111!
Num (Extended) outputs,weights in Series:
1,36,0,1:1, 0
Num (Extended) outputs,weights in Series:
C3,3:9, 0
Ft16:16, 160
Total weights = 160
[C3,3Ft16]:16, 160
Mp3,3:16, 0
Lfys64:64, 20736
Lfx96:96, 61824
Lrx96:96, 74112
Lfx512:512, 1247232
Fc111:111, 0
Total weights = 1404064
Previous null char=110 mapped to 110
Continuing from /home/ajinkya/tesstutorial/trainplusminus/eng.lstm
Loaded 72/72 pages (1-72) of document /home/ajinkya/tesstutorial/trainplusminus/eng.Arial_Bold.exp0.lstmf
Loaded 72/72 pages (1-72) of document /home/ajinkya/tesstutorial/trainplusminus/eng.Arial_Bold_Italic.exp0.lstmf
Loaded 72/72 pages (1-72) of document /home/ajinkya/tesstutorial/trainplusminus/eng.Arial.exp0.lstmf
Loaded 72/72 pages (1-72) of document /home/ajinkya/tesstutorial/trainplusminus/eng.Century_Schoolbook_L_Bold_Italic.exp0.lstmf
Loaded 72/72 pages (1-72) of document /home/ajinkya/tesstutorial/trainplusminus/eng.Arial_Italic.exp0.lstmf
Loaded 72/72 pages (1-72) of document /home/ajinkya/tesstutorial/trainplusminus/eng.Century_Schoolbook_L_Bold.exp0.lstmf
Loaded 72/72 pages (1-72) of document /home/ajinkya/tesstutorial/trainplusminus/eng.Century_Schoolbook_L_Italic.exp0.lstmf
Loaded 72/72 pages (1-72) of document /home/ajinkya/tesstutorial/trainplusminus/eng.Courier_New_Bold.exp0.lstmf
Loaded 72/72 pages (1-72) of document /home/ajinkya/tesstutorial/trainplusminus/eng.Century_Schoolbook_L_Medium.exp0.lstmf
Segmentation fault (core dumped)

In order to fix this error I have referred (#1447) issue.
Ive tried 1) downloading data from wget from https://github.com/tesseract-ocr/tessdata/raw/master/eng.traineddata and placed this data in tessdata/best/eng.traineddata, however this does not resolve the error

Let me know if there is anything else I should do to fix this error?

Expected Behavior:

Suggested Fix:

[stweil]

You'll need https://github.com/tesseract-ocr/tessdata_best/raw/master/eng.traineddata.

It is a known bug that Tesseract crashes when training with fast models (either from tessdata_fast or tessdata) is tried. See issue #1573.

[ajinkya933]

as directed I placed above file in tessdata/best/eng.traineddata . However the error persists. Should I redo the entire steps from beginning of the tutorial ? to solve this problem

[stweil]

Ive tried 1) downloading data from wget from https://github.com/tesseract-ocr/tessdata/raw/master/eng.traineddata [...]

That's the wrong download link.

[ajinkya933]

I solved this error by

A) Deleting everything and then reinstalling tesseract from here:
git clone https://github.com/tesseract-ocr/tesseract.git

B) cd tesseract/tessdata

C) sudo mkdir best

D) cd best

E) wget https://github.com/tesseract-ocr/tessdata_best/raw/master/eng.traineddata
wget https://github.com/tesseract-ocr/tessdata_best/raw/master/heb.traineddata
wget https://github.com/tesseract-ocr/tessdata_best/raw/master/chi_sim.traineddata

Now you have all the correct stuff needed to run this tutorial. Geez !! the files required to run this tutorial are all over the world. Someone please combine this into a single github repo wherein you can git clone and follow the instructions.