It is possible that you are treating the 'starter' traineddata file as the final one. Please read the training wiki page fully as the training process has been changed by Ray in his last update.
On 08-Dec-2017 10:25 AM, "ShreeDevi Kumar" <[email protected]> wrote: > Please check the last section on > > https://github.com/tesseract-ocr/tesseract/wiki/TrainingTesseract-4.00 > > Regarding combining files to know the correct syntax for building the new > traineddata file. > > > On 08-Dec-2017 8:04 AM, "J Klein" <[email protected]> wrote: > >> >> >> On Thursday, December 7, 2017 at 9:02:11 PM UTC-5, shree wrote: >>> >>> Re smaller traineddata size, it could possibly be related to the word >>> list dictionary size. >>> >>> You can unpack the original traineddata and compare the word list size >>> with the one you used. >>> >> >> >> Thank you for the hint. >> >> I ran the following (-u is 'unpack all' I think), >> >> combine_tessdata -u /usr/local/share/tessdata/eng.traineddata eng. >> >> and I got: >> >> -rw-r--r-- 1 klein staff 11689099 Dec 7 21:22 eng.lstm >> >> -rw-r--r-- 1 klein staff 4738 Dec 7 21:22 eng.lstm-number-dawg >> >> -rw-r--r-- 1 klein staff 4322 Dec 7 21:22 eng.lstm-punc-dawg >> >> -rw-r--r-- 1 klein staff 1012 Dec 7 21:22 eng.lstm-recoder >> >> -rw-r--r-- 1 klein staff 6360 Dec 7 21:22 eng.lstm-unicharset >> >> -rw-r--r-- 1 klein staff 3694794 Dec 7 21:22 eng.lstm-word-dawg >> >> -rw-r--r-- 1 klein staff 80 Dec 7 21:22 eng.version -- CONTENT >> is 4.00.00alpha:eng:synth20170629:[1,36,0,1Ct3,3,16Mp3,3Lfys >> 64Lfx96Lrx96Lfx512O1c1] >> >> >> Now I tried to unpack the one I created by adding the characters, and I >> get >> >> >> xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx eng.lstm is missing! >> >> -rw-r--r-- 1 klein staff 3506 Dec 7 21:26 eng.lstm-number-dawg >> >> -rw-r--r-- 1 klein staff 4322 Dec 7 21:26 eng.lstm-punc-dawg >> >> -rw-r--r-- 1 klein staff 1030 Dec 7 21:26 eng.lstm-recoder >> >> -rw-r--r-- 1 klein staff 9379 Dec 7 21:26 eng.lstm-unicharset >> >> -rw-r--r-- 1 klein staff 4153402 Dec 7 21:26 eng.lstm-word-dawg >> >> -rw-r--r-- 1 klein staff 12 Dec 7 21:26 eng.version -- CONTENT >> IS '4.00.00alpha' >> >> So you're right that the word-list is different. >> >> But more importantly it seems that eng.lstm isn't in the final >> eng.traineddata. Do I not understand something about how the process >> works? Is this my mistake, or a glitch! >> >> Thanks for helping me to make progress. >> >> >> >> >> -- >> 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/ms >> gid/tesseract-ocr/0dc37684-c454-4993-9387-ad641f22f016%40googlegroups.com >> <https://groups.google.com/d/msgid/tesseract-ocr/0dc37684-c454-4993-9387-ad641f22f016%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> For more options, visit https://groups.google.com/d/optout. >> > -- 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/CAG2NduXP3-B6tAd_e7K63EuDPSCfA0mRNf_NWzuZPOdCsPdvYw%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
