You should edit your post and add the code directly inside.
Here you are trying to parallelize command line arguments which seems to be directory paths to check if a file exist. This is IO-bound and your bottleneck will be your disk. Parallelizing this loop will probably slow your program down due to parallel overhead and also competition over the kernels limited file descriptors.
