Search papers, labs, and topics across Lattice.
Affiliation:
1
0
FlowTT slashes inference latency by over 42% by intelligently reusing computation flows in Tensor-Train embeddings.
