Search papers, labs, and topics across Lattice.
1
0
4
Decomposing LLMs doesn't have to mean sacrificing inference speed: DeInfer unlocks efficient parallel inference for these models.
