Optimizing for Latency

[mattare2]

Have the authors considered any approaches to reduce the latency of this approach?

Would be interested to understand if any avenues have been pursued (e.g., distilling into a more performant architecture)

Thanks!