You can give MODEL_NAME any value. For specifying the path to data directory use: DATA_DIR Data directory for output files, proto model, start model, etc. Default: data
for example: if MODEL_NAME=abc and DATA_DIR=data you need to put the ground truth files (box, gt.txt & tif) in 'data/abc-ground-truth', OR set value for GROUND_TRUTH_DIR directly. (In both cases 'OUTPUT_DIR Output directory for generated files. Default: DATA_DIR/MODEL_NAME' will be 'data/abc') Check make file: https://github.com/tesseract-ocr/tesstrain/blob/main/Makefile Check tutorial video @: https://www.youtube.com/watch?v=KE4xEzFGSU8 On Thursday, August 1, 2024 at 8:42:55 PM UTC+3 surf...@gmail.com wrote: > I'm attempting to train tesseract for a new english font. ve created a > folder ./output comprising images, ground truth and box files. Does this > folder need to be in the tesseract folder? > What is the exact command to perform the training? According to the > documentation, its: > >>make training MODEL_NAME=name-of-the-resulting-model > > However, the details are thin in the doc. For example how to specify the > paths to the raw data, what to call MODEL_NAME if I'm just fine tuning and > not creating a new language. -- 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 tesseract-ocr+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/9e3378b5-a01d-4342-b11a-9d8793b583b2n%40googlegroups.com.
