model compression and optimization for deployment for Pytorch, including knowledge distillation, quantization and pruning.(知识蒸馏，量化，剪枝) - View it on GitHub