You need to relax the unit cell first, because it's not necessarily the same as Co-Si. If 12.5 at% Gd-doped Co-Si you've built is stable, the calculations will converge to a specific negative energy. After that, you can calculate the band structure. The tags you've used in the INCAR file are not for ionic relaxation.

Thanks, Ehsan Moradpur-Tari ! The steps I am following are: k-points convergence --> Energy vs volume curve --> cell relaxation --> formation energy --> SC calculations --> dos calculations --> band-structure calculation. So, at the moment I am trying to get the Energy vs volume curve without any relaxation, as I found those calculations to be very intensive and my INCAR file does not have tags for relaxation. To be honest I did not get k-point convergence as well. I assumed that the k-points will converge at the same k-mesh as that of the CoSi unit cell. Would I be correct in assuming that? or Should I consider the structure to be unstable if my k-points don't converge?

You can use these tags for relaxation:

ISIF = 7 #volume relaxation

ISIF = 3 #unit cell relaxation

Other tags should be set accordingly:

NSW = 100

...

You can also use automatic mesh for Kpoints.