How to train mBart or any multilingual model for translation task

According to docs

you will need to set the language codes of your inputs and targets in the tokenizer by setting tokenizer.src_lang and tokenizer.tgt_lang to the right values.

How and where do i write this to the code? Is it supposed to be written at this line of code?

tokenizer = AutoTokenizer.from_pretrained(model_checkpoint, return_tensors="pt")

If so, how? Please give an example line of code. Thank you.