Can I specify output files paths at the command line?

[pauljurczak]

I would like to run multiple copies of sparrow without having to duplicate the whole directory structure, if possible. In order to avoid collisions, specifying different output file names would be necessary. Is that possible?

I found that the maximum performance in evals/s can be achieved by running multiple copies of the binary. Here is the performance (median evals/s) on AMD 3800X CPU (8-cores) running 0.1 hour task on the same input, SIMD enabled:

Copies x Workers	Total evals/s	CPU
1 x 6	1716	3800X
1 x 7	1799	3800X
2 x 3	1943	3800X
3 x 2	2174	3800X
6 x 1	2395	3800X