Search papers, labs, and topics across Lattice.
1
0
2
Amortizing distillation across model size and variant can produce a continuum of optimized LLMs with a single distillation run, revolutionizing model deployment efficiency.
