How to reproduce the Parts-of-speech (POS) task？

Hello @XiangLi1999 ,

Thank you for the exciting work !

If I'm not mistaken, your README primarily focuses on the infilling task (for example, `python scripts/infill.py ...`). However, I would like to reproduce the results for the Parts-of-speech (POS) task mentioned in the paper. Could you please let me know what commands I should execute after training the Diffusion-LM?

Thanks！