Exploring prompt tuning with pseudolabels for multiple modalities, learning settings, and training strategies. - View it on GitHub