Technology ❯ Data Science ❯ Big Data ❯ Scalability
The authors report up to 450x faster training with accuracy maintained even at D=128.