| # sh scripts/uncertainty/randaugment/cifar100_test_ood_msp.sh | |
| ############################################ | |
| # alternatively, we recommend using the | |
| # new unified, easy-to-use evaluator with | |
| # the example script scripts/eval_ood.py | |
| # especially if you want to get results from | |
| # multiple runs | |
| python scripts/eval_ood.py \ | |
| --id-data cifar100 \ | |
| --root ./results/cifar100_resnet18_32x32_base_e100_lr0.1_randaugment-1-14 \ | |
| --postprocessor msp \ | |
| --save-score --save-csv | |