Tesstrain.sh not generating TrainedData

[kamra....@gmail.com]

Hi,

I am running the following command to create trained data:

tesstrain.sh --fonts_dir /usr/share/fonts --lang eng --linedata_only --fontlist "FreeMono" --noextract_font_properties --langdata_dir /home/administrator/Downloads/tesseract-4.0.0/langdata --my_boxtiff_dir /home/administrator/pooja/testImages/ --tessdata_dir /home/administrator/Downloads/tesseract-4.0.0/tessdata --output_dir /home/administrator/images/output_folder_1/

After this it is printing:

=== Starting training for language 'eng'

[Sat Apr 24 20:15:19 IST 2021] /usr/local/bin/text2image --fonts_dir=/usr/share/fonts --font=FreeMono --outputbase=/tmp/font_tmp.e9Fi4vFUQQ/sample_text.txt --text=/tmp/font_tmp.e9Fi4vFUQQ/sample_text.txt --fontconfig_tmpdir=/tmp/font_tmp.e9Fi4vFUQQ

Rendered page 0 to file /tmp/font_tmp.e9Fi4vFUQQ/sample_text.txt.tif

=== Phase I: Generating training images ===

Rendering using FreeMono

[Sat Apr 24 20:15:23 IST 2021] /usr/local/bin/text2image --fontconfig_tmpdir=/tmp/font_tmp.e9Fi4vFUQQ --fonts_dir=/usr/share/fonts --strip_unrenderable_words --leading=32 --xsize=3600 --char_spacing=0.0 --exposure=0 --outputbase=/tmp/eng-2021-04-24.Y8g/eng.FreeMono.exp0 --max_pages=0 --font=FreeMono --text=/home/administrator/Downloads/tesseract-4.0.0/langdata/eng/eng.training_text

Rendered page 0 to file /tmp/eng-2021-04-24.Y8g/eng.FreeMono.exp0.tif

Rendered page 1 to file /tmp/eng-2021-04-24.Y8g/eng.FreeMono.exp0.tif

=== Phase UP: Generating unicharset and unichar properties files ===

[Sat Apr 24 20:15:25 IST 2021] /usr/local/bin/unicharset_extractor --output_unicharset /tmp/eng-2021-04-24.Y8g/eng.unicharset --norm_mode 1

Usage: /usr/local/bin/unicharset_extractor [--output_unicharset filename] [--norm_mode mode] box_or_text_file [...]

Where mode means:

 1=combine graphemes (use for Latin and other simple scripts)

 2=split graphemes (use for Indic/Khmer/Myanmar)

 3=pure unicode (use for Arabic/Hebrew/Thai/Tibetan)

As per specification it should be end with:

Created starter traineddata for LSTM training of language 'eng' 

Run 'lstmtraining' command to continue LSTM training for language 'eng

Please help.

Regards,

Pooja

[kamra....@gmail.com]

Resolved. There were mismatch between trainedata used. Tesseract installed was of version 4.1.0. And i was giving path of downloaded tesseract 4.0.

It was causing issue.