After some trials, we don't want to name the run storing subdirectories after the hyper parameters swept. Instead we just the number assigned by hydra to the job.
I also left as comment other possibilites of naming for the subdirectories (it might be useful for some users)
After some trials, we don't want to name the run storing subdirectories after the hyper parameters swept. Instead we just the number assigned by hydra to the job.
I also left as comment other possibilites of naming for the subdirectories (it might be useful for some users)