Joint Pruning and Channel-wise Mixed-Precision Quantization for Efficient Deep Neural Networks

Open in new window