Low Precision RNNs: Quantizing RNNs Without Losing Accuracy | Read Paper on Bytez