Dear @, I played around with training the new LSTM mode. According to the documentation of the network specification (https://github.com/tesseract-ocr/tesseract/wiki/VGSLSpecs), the last number in the first tupel called 'depth' corresponds to the type of input (i.e. 1 ... grayscale, 3 ... color). However one o the given examples uses '48' in this position:
[1,1,0,48 Lbx256 O1c105] Using an assumingly corrected specification [1,48,0,1 Lbx256 O1c105] causes serious runtime issues namely each iteration takes several minutes and huge amounts of memory are adressed. Any hints on what I am doing wrong here? Many thanks in advance, Kay -- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/tesseract-ocr. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/167ff50a-2fbb-463c-b637-c051b2e9da82%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
