adding one evaluation script for bop instead of one for each dataset the new file is evaluate_zs6d.py

speed up by 50% for the find_template_cpu method and benchmarking for the find_correspondences method