Fetching the paper…

Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding · Around