!int_mode_:Error:Assert failed:in file weightmatrix.cpp, line 244

[msute]

hi！ i meet the question！how can i solve？ my tesseract version is 4.0. i want to train the chi_sim.traineddata
i tried to train the eng and number, it's all this kind of mistake!

msute@ubuntu:~/Desktop/jTessBoxEditor/tesseract-ocr$ lstmtraining --debug_interval -1 --max_iterations 800 --continue_from='/home/msute/Desktop/jTessBoxEditor/tesseract-ocr/chi_sim.lstm' --model_output='/home/msute/Desktop/jTessBoxEditor/tesseract-ocr/trainingoutput' --train_listfile='/home/msute/Desktop/jTessBoxEditor/tesseract-ocr/chi_sim.training_files.txt' --traineddata='/home/msute/Desktop/jTessBoxEditor/tesseract-ocr/chi_sim.traineddata'
Loaded file /home/msute/Desktop/jTessBoxEditor/tesseract-ocr/chi_sim.lstm, unpacking...
Warning: LSTMTrainer deserialized an LSTMRecognizer!
Continuing from /home/msute/Desktop/jTessBoxEditor/tesseract-ocr/chi_sim.lstm
Loaded 1/1 pages (1-1) of document /home/msute/Desktop/jTessBoxEditor/tesseract-ocr/name.test.exp0.lstmf
Iteration 0: ALIGNED TRUTH : 惠娟潘墩镇镐源志学韩瑞峰男女壹贰叁肆伍陆柒捌玖拾源村
Iteration 0: BEST OCR TEXT : 惠娟潘墩镇镐源志学韩瑞峰男女壹武叁肆伍陆柴捌玖拾源村
File /home/msute/Desktop/jTessBoxEditor/tesseract-ocr/name.test.exp0.lstmf page 0 :
!int_mode_:Error:Assert failed:in file weightmatrix.cpp, line 244
Segmentation fault (core dumped)

[amitdo]

Duplicate of #1573.

[msute]

＃1573的副本。

what‘s your mean?

[amitdo]

Someone already reported about this issue before.

See issue 1573

Use a traineddata file from the tessdata_best repo for training.

[msute]

Someone already reported about this issue before.

See issue 1573

Use a traineddata file from the tessdata_best repo for training.

yes，my file from this website. but it still made this mistake. i just want to combine my traineddata with this source. train more font.

[amitdo]

yes，my file from this website

This error message indicates that you are training with integer model(tessdata_fast) instead of float model (from tessdata_best).

What the size of your chi_sim.traineddata file?

[msute]

yes，my file from this website

This error message indicates that you are training with integer model(tessdata_fast) instead of float model (from tessdata_best).

What the size of your chi_sim.traineddata file?

What size? how can i kown my traineddata size? If i use the method from this photo, can i solve this error message? Thank you for your patience