Speedup deep learning models on GPU by taking advantage of efficient unstructured pruning and bit-width reduction

Re-assign community