Model Quantization and Pruning

Techniques that compress neural networks by reducing weight precision or removing redundant parameters to lower memory footprint