[Question] Which fonts were used to train the publicly available pre-trained model?

[rishiyama]

Regarding fonts.csv in manga-ocr/manga_ocr_dev, it states:

List of fonts with metadata used by synthetic data generator. Provided file is just an example, you have to generate similar file for your own set of fonts, using manga_ocr_dev/synthetic_data_generator/scan_fonts.py script. Note that label will be filled with regular by default. You have to label your special fonts manually.

I would appreciate it if you could list the types of fonts used for training.

I understand that fonts.csv only provides an example and that users need to generate their own font list using the scan_fonts.py script. However, could you provide a list of fonts that were actually used for training?

Thank you. I truly appreciate your contributions!

[phu2007vis]

same question