tesseract-ocr / tesstrain

Train Tesseract LSTM with make
Apache License 2.0
604 stars 178 forks source link

radical-stroke.txt location changed #278

Closed aquino-a closed 2 years ago

aquino-a commented 2 years ago

It looks like 'master' was changed to 'main' the tesseract project. This breaks line 317 in the MakeFile.

NazeemaShaik commented 2 years ago

I am trying to train tesseract on my custom images. But training is getting stop in the middle. It is not able to find radical-stroke.txt in the location(https://github.com/tesseract-ocr/langdata_lstm/raw/master/radical-stroke.txt). Can anyone help me how to proceed?. Thanks

aquino-a commented 2 years ago

The last commit in the tesseract project changed 'master' to 'main', which broke all URLs using that branch. You can either edit the url or put file in the correct location manually.

stweil commented 2 years ago

This is fixed now. But we should not use the default branch here. Using the latest tagged release would be better.

So please keep this issue open until release 5.0.0 was tagged. Then we can use that.

NazeemaShaik commented 2 years ago

image

I have edited line 317 in makefile as https://github.com/tesseract-ocr/langdata_lstm/raw/main/radical-stroke.txt . But getting the above error. Can you please help me on this?. Thanks

stweil commented 2 years ago

That's unrelated to the issue here. The latest Tesseract code would have shown why your training does not work: you start it with a "fast" model instead of using a "best" model.

stale[bot] commented 2 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.