Bigger Neural Networks May Actually Be Simpler
中文摘要
新的压缩方法表明，当模型达到相同质量时，参数量更大的神经网络可能比小模型更简单。
English Summary
A new compression method suggests that larger neural networks may actually be simpler than smaller ones when achieving the same quality.
Original Excerpt
A new compression method suggests something wonderfully counterintuitive: when two models reach the same quality, the billion-parameter… Continue reading on Medium »