Search papers, labs, and topics across Lattice.
2
0
6
8
Achieving state-of-the-art multimodal performance with only 208.62 million unique images and a theoretical training cost of just $400K challenges the notion that larger datasets and budgets are always necessary for success.
Train billion-parameter LLMs on a single H100 GPU, no AdamW required, using a memory-efficient orthogonal transformation method.
