Search papers, labs, and topics across Lattice.
2
0
5
AudioCALM achieves state-of-the-art performance in speech, sound, and music generation by seamlessly integrating diverse audio modalities into a single autoregressive framework.
Overlapping communication with computation in distributed Diffusion Transformer inference can yield up to 8.4x speedups, challenging the assumption that collectives are a fixed bottleneck.
