A TrOCR model for Persian (V2)
from hezar.models import Model model = Model.load("hezarai/trocr-base-fa-v2") text = model.predict(["image.jpg"]) print(text)