About feature extraction of visual language foundation models

Thank you for your excellent work！
I have noticed that you mentioned some model has encode_text function. Is there any support in the scripts for this kind of model? I mean I don't see any codes in scripts that is about text feature extraction, or the scripts are only about image feature extraction?🤨🤨