in the end of SGD's step function where Ts is the binary mask from quantization. I was wondering if same thing can be achieved by just setting the quantized weights with required_grad = False (so would be in the quantization_scheduler class), then we don't need to hook into the optimizer code.
I'm wondering if you have tried this and whether that will work?
Hello,
I noticed to implement INQ, you plugin
in the end of SGD's
step
function whereTs
is the binary mask from quantization. I was wondering if same thing can be achieved by just setting the quantized weights withrequired_grad = False
(so would be in thequantization_scheduler
class), then we don't need to hook into the optimizer code.I'm wondering if you have tried this and whether that will work?