Optimization for conv workload mapping #21
Description
Optimization: If we can optionally lower a tensor & the loop below it is not something through which we can lower for a given permutation, skip options that lower that tensor because they get the same result as not lowering the tensor.