If I do the relaxation of MoS2 nanoribbon, I used to relax the MoS2 monolayer with the smallest unit cell. After that, I will directly repeat the unit cell to form the nanoribbon without any further relaxation. As to the parameters for doing the relaxation, I think GGA-SG15-medium is sufficient for this system. I like to use a k-point sampling of (9 9 1) (for hexagonal unit cell) and 0.01/0.001 for force tolerance and stress tolerance, respectively.
I also want to know if it is appropriate.