Quantization is a common technique. See for example https://pytorch.org/docs/stable/quantization.html