Feature Request: Add Marathi (Devanagari) support in PP-OCRv6

[tesla3108-cmyk]

Hello PaddleOCR Team,

Thank you for releasing the excellent PP-OCRv6 models.

I would like to request official support for Marathi language (Devanagari script) in the PP-OCRv6 unified models (medium/small/tiny).

Why this is important:

• Marathi is spoken by more than 90 million people, mainly in Maharashtra, India.
• Currently, PP-OCRv6 only supports Chinese, English, Japanese + Latin scripts.
• Devanagari languages (Hindi, Marathi, Nepali etc.) still need the older PP-OCRv5 devanagari model.
• Many real-world use cases in India (government documents, education, banking, newspapers) require accurate Marathi OCR.

Request:

Please consider adding Marathi / Devanagari support in the next version of PP-OCRv6 or as a separate lightweight model.

I am happy to help with testing or providing sample data if needed.

Thank you!