configurations/cluster/fas_cpu.yaml

defaults:
  - base_slurm
  - _self_
params:
  partition: shared # e.g. kempner_h100
  # account: kempner_sham_lab # e.g. kempner_sham_lab
  env_name: wm
  num_gpus: 4
  num_cpus: 48
  memory: 128G
  time: "3-00:00:00" 

launch_template: |
  #!/bin/bash
  #SBATCH -J {name}
  #SBATCH -o {log_dir}/out_%j.out
  #SBATCH -e {log_dir}/error_%j.err
  #SBATCH --mail-user={email}
  #SBATCH --mail-type=FAIL
  #SBATCH --partition={partition}
  #SBATCH --nodes=${experiment.num_nodes}
  #SBATCH --cpus-per-task=12
  #SBATCH --mem={memory}
  #SBATCH --time={time}
  
  # export NCCL_DEBUG=INFO
  # export PYTHONFAULTHANDLER=1

  cd {project_root}
  module load Mambaforge
  mamba deactivate
  mamba activate {env_name}
  srun python -m main {python_args}