Translution: Unifying Self-attention and Convolution for Adaptive and Relative Modeling

Open in new window