Deep-Learning-Papers-Reading-Roadmap/pdfs/2 Deep Learning Method/2.2 Optimization/26 SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 1MB model size.pdf at master · zzh428/Deep-Learning-Papers-Reading-Roadmap