Post-Training Sparsity-Aware Quantization

1 code implementation • NeurIPS 2021 • Gil Shomron, Freddy Gabbay, Samer Kurzum, Uri Weiser

Moreover, instead of quantizing activation-by-activation to 4 bits, we focus on pairs of 8-bit activations and examine whether one of the two is equal to zero.

Quantization