Dynamic Tuning Towards Parameter and Inference Efficiency for ViT Adaptation

Benchmark Model Rank Results
monocular-depth-estimation-on-clevr-distanceDyT ViT-B/16 (VTAB-1k)#6Accuracy: 63.2
object-counting-on-clevr-countDyT ViT-B/16 (VTAB-1k)#6Accuracy: 81.6