Experiments with SpuCo Package Group inference tuning Robust training tuning E2E tuning ERM tuning Run guild queues for gpu affinitization for i in {2..3}; do guild run queue -b --gpus="$i" -y; done