How to Adapt the Code for Training on TCR, a Real-World Chinese Dataset?

[cosmicrealm]

Hello,

I am interested in training on TCR, a real-world Chinese dataset. However, as a beginner in text super-resolution and someone not very familiar with OCR-related preprocessing or priors, I would like to know:
1. Which parts of the code should I modify to ensure compatibility with Chinese characters?
2. Are there any specific considerations or best practices I should follow when working with a Chinese dataset?

Any guidance or recommendations would be greatly appreciated.