Search papers, labs, and topics across Lattice.
Tsinghua University
1
0
3
Achieving comparable text-to-image quality with a linearized model that accelerates inference by up to 1.47 times, all while leveraging pretrained weights.
