Search papers, labs, and topics across Lattice.
1
0
3
FourTune slashes memory overhead by 2.25x while matching the performance of full-precision fine-tuning in diffusion models.
