Number of Attention Heads vs. Number of Transformer-encoders in Computer Vision.

Tomas Hrycej, Bernhard Bermeitinger, Siegfried Handschuh

Browse the full IC3K paper archive.