Search papers, labs, and topics across Lattice.
1
0
3
2
Double your diffusion model inference speed on SDXL and SD3 with a clever hybrid parallelism approach that exploits conditional guidance.
