03 September 2021 0 7K Report

Given any CNN architecture to be trained on any classification dataset, if we make all the weights and parameters as binary, would the training converge much faster ?

More Titas De's questions See All
Similar questions and discussions