ConceptAI & ML
Mixture-of-Experts (MoE) inference efficiency
MoE models can achieve high throughput and efficiency at scale, particularly with large batch sizes, compared to dense models.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
GPU parallelism1Inference throughput1Mixture-of-Experts models1
Lecturers
1Connections
1The lectures that define it or assume it, and the concepts next to it.