A TrOCR model for Persian. This is just a preview to verify the model works in Hezar. This model will be trained on more and better datasets soon!

Example

from hezar import Model

model = Model.load("trocr-base-fa-v1")
text = model.predict(["image.jpg"])
print(text)