Open MORzyuan opened 5 years ago
If you train for the lstm engine, you should use lstm.train.
If you train for the lstm engine, you should use lstm.train.
I am not aimed to train lstm engine. And also for the 3.05 version, just as the Environment 2 says, this problem still exists.
Environment 1
Tesseract Version: tesseract 4.1.0 leptonica-1.78.0 libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.0.10 : zlib 1.2.11 : libwebp 1.0.2 : libopenjp2 2.3.1 Found AVX2 Found AVX Found SSE Found libarchive 3.3.3 zlib/1.2.11 liblzma/5.2.4 bz2lib/1.0.6
Commit Number:
Platform: ProductName: Mac OS X ProductVersion: 10.13.6 BuildVersion: 17G65
Environment 2
Tesseract Version: tesseract 3.05.02 leptonica-1.78.0 libjpeg 8d (libjpeg-turbo 1.4.2) : libpng 1.2.54 : libtiff 4.0.6 : zlib 1.2.8
Commit Number:
Platform: Distributor ID: Ubuntu Description: Ubuntu 16.04.5 LTS Release: 16.04 Codename: xenial
The case reported below have been tested under both of these two enviroments.
Current Behavior:
Training progress progress fell into silence but didn't exit.
tesseract allhz.NewspaperSung.exp0.tif allhz.NewspaperSung.exp0 box.train
The filecommonhz.NespaperSung.exp0.tif
was generated by the following commandtext2image --text=training_all.txt --outputbase=allhz.NewspaperSung.exp0 --fonts_dir=../font/ --font='Old Newspapers Sung' --writing_mode vertical --xsize=1600 --ysize=2000 --resolution=300
wheretraining_all.txt
is the random combination of Han characters, shown as the attached file training_all.txt, and the font file can be downloaded here: http://js.xiazaicc.com/down1/mgbzzt_downcc.zipThe case happened 3 times as the same page 404(404 is not a lovely code!! TAT), it seems not a coincidence.
Best!