Datasets:
AlgorithmicResearchGroup
/
arxiv_research_code

Dataset card Data Studio Files
xet
Community
Search is not available for this dataset
repo
stringlengths
2
152
⌀
file
stringlengths
15
239
code
stringlengths
0
58.4M
file_length
int64
0
58.4M
avg_line_length
float64
0
1.81M
max_line_length
int64
0
12.7M
extension_type
stringclasses
364 values
null
OpenOOD-main/scripts/ood/logitnorm/cifar100_test_logitnorm.sh
#!/bin/bash # sh scripts/ood/logitnorm/cifar100_test_logitnorm.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_logitnorm_e100_lr0.1_alpha0.04_default \ --postprocessor msp \ --save-score --save-csv
486
31.466667
84
sh
null
OpenOOD-main/scripts/ood/logitnorm/cifar100_train_logitnorm.sh
#!/bin/bash # sh scripts/ood/logitnorm/cifar100_train_logitnorm.sh python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/train/train_logitnorm.yml \ configs/preprocessors/base_preprocessor.yml \ --seed 0
294
28.5
54
sh
null
OpenOOD-main/scripts/ood/logitnorm/cifar10_test_logitnorm.sh
#!/bin/bash # sh scripts/ood/logitnorm/cifar10_test_logitnorm.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 cifar10 \ --root ./results/cifar10_resnet18_32x32_logitnorm_e100_lr0.1_alpha0.04_default \ --postprocessor msp \ --save-score --save-csv
483
31.266667
83
sh
null
OpenOOD-main/scripts/ood/logitnorm/cifar10_train_logitnorm.sh
#!/bin/bash # sh scripts/ood/logitnorm/cifar10_train_logitnorm.sh python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/train/train_logitnorm.yml \ configs/preprocessors/base_preprocessor.yml \ --seed 0
291
28.2
53
sh
null
OpenOOD-main/scripts/ood/logitnorm/imagenet200_test_logitnorm.sh
#!/bin/bash # sh scripts/ood/logitnorm/imagenet200_test_logitnorm.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 # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_logitnorm_e90_lr0.1_alpha0.04_default \ --postprocessor msp \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_logitnorm_e90_lr0.1_alpha0.04_default \ --postprocessor msp \ --save-score --save-csv --fsood
738
29.791667
88
sh
null
OpenOOD-main/scripts/ood/logitnorm/imagenet200_train_logitnorm.sh
#!/bin/bash # sh scripts/ood/logitnorm/imagenet200_train_logitnorm.sh python main.py \ --config configs/datasets/imagenet200/imagenet200.yml \ configs/networks/resnet18_224x224.yml \ configs/pipelines/train/train_logitnorm.yml \ configs/preprocessors/base_preprocessor.yml \ --optimizer.num_epochs 90 \ --dataset.train.batch_size 128 \ --num_gpus 2 --num_workers 16 \ --merge_option merge \ --seed 0
437
30.285714
59
sh
null
OpenOOD-main/scripts/ood/logitnorm/imagenet_test_logitnorm.sh
#!/bin/bash # sh scripts/ood/logitnorm/imagenet_test_logitnorm.sh ############################################ # we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood_imagenet.py # available architectures: # resnet50 # ood python scripts/eval_ood_imagenet.py \ --ckpt-path ./results/imagenet_resnet50_logitnorm_e30_lr0.001_alpha0.04_default/s0/best.ckpt \ --arch resnet50 \ --postprocessor msp \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood_imagenet.py \ --ckpt-path ./results/imagenet_resnet50_logitnorm_e30_lr0.001_alpha0.04_default/s0/best.ckpt \ --arch resnet50 \ --postprocessor msp \ --save-score --save-csv --fsood
723
27.96
96
sh
null
OpenOOD-main/scripts/ood/logitnorm/imagenet_train_logitnorm.sh
#!/bin/bash # sh scripts/ood/logitnorm/imagenet_train_logitnorm.sh python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/networks/resnet50.yml \ configs/pipelines/train/train_logitnorm.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint ./results/pretrained_weights/resnet50_imagenet1k_v1.pth \ --optimizer.lr 0.001 \ --optimizer.num_epochs 30 \ --dataset.train.batch_size 128 \ --num_gpus 2 --num_workers 16 \ --merge_option merge \ --seed 0
562
32.117647
82
sh
null
OpenOOD-main/scripts/ood/mcd/cifar100_test_mcd.sh
#!/bin/bash # sh scripts/ood/mcd/cifar100_test_mcd.sh # NOTE!!!! # need to manually change the checkpoint path # remember to use the last_*.ckpt because mcd only trains for the last 10 epochs # and the best.ckpt (according to accuracy) is typically not within the last 10 epochs # therefore using best.ckpt is equivalent to early stopping with standard cross-entropy loss python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/mcd_net.yml \ configs/preprocessors/base_preprocessor.yml \ configs/pipelines/test/test_ood.yml \ configs/postprocessors/mcd.yml \ --network.backbone.name resnet18_32x32 \ --network.pretrained True \ --network.checkpoint 'results/cifar100_oe_mcd_mcd_e100_lr0.1_default/s0/last_epoch100_acc0.7510.ckpt' \ --num_workers 8 \ --seed 0
877
40.809524
107
sh
null
OpenOOD-main/scripts/ood/mcd/cifar100_train_mcd.sh
#!/bin/bash # sh scripts/ood/mcd/cifar100_train_mcd.sh GPU=1 CPU=1 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_oe.yml \ configs/networks/mcd_net.yml \ configs/preprocessors/base_preprocessor.yml \ configs/pipelines/train/baseline.yml \ configs/pipelines/train/train_mcd.yml \ --network.backbone.name resnet18_32x32 \ --network.pretrained False \ --dataset.image_size 32 \ --optimizer.num_epochs 100 \ --num_workers 8 \ --seed 0
724
28
53
sh
null
OpenOOD-main/scripts/ood/mcd/cifar10_test_mcd.sh
#!/bin/bash # sh scripts/ood/mcd/cifar10_test_mcd.sh # NOTE!!!! # need to manually change the checkpoint path # remember to use the last_*.ckpt because mcd only trains for the last 10 epochs # and the best.ckpt (according to accuracy) is typically not within the last 10 epochs # therefore using best.ckpt is equivalent to early stopping with standard cross-entropy loss python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/mcd_net.yml \ configs/preprocessors/base_preprocessor.yml \ configs/pipelines/test/test_ood.yml \ configs/postprocessors/mcd.yml \ --network.backbone.name resnet18_32x32 \ --network.pretrained True \ --network.checkpoint 'results/cifar10_oe_mcd_mcd_e100_lr0.1_default/s0/last_epoch100_acc0.9420.ckpt' \ --num_workers 8 \ --seed 0
871
40.52381
106
sh
null
OpenOOD-main/scripts/ood/mcd/cifar10_train_mcd.sh
#!/bin/bash # sh scripts/ood/mcd/cifar10_train_mcd.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} \ -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_oe.yml \ configs/networks/mcd_net.yml \ configs/preprocessors/base_preprocessor.yml \ configs/pipelines/train/baseline.yml \ configs/pipelines/train/train_mcd.yml \ --network.backbone.name resnet18_32x32 \ --network.pretrained False \ --dataset.image_size 32 \ --optimizer.num_epochs 100 \ --num_workers 8 \ --seed ${SEED}
762
27.259259
51
sh
null
OpenOOD-main/scripts/ood/mcd/imagenet200_test_mcd.sh
#!/bin/bash # sh scripts/ood/mcd/imagenet200_test_mcd.sh # NOTE!!!! # need to manually change the checkpoint path # remember to use the last_*.ckpt because mcd only trains for the last 10 epochs # and the best.ckpt (according to accuracy) is typically not within the last 10 epochs # therefore using best.ckpt is equivalent to early stopping with standard cross-entropy loss SCHEME="ood" # "ood" or "fsood" python main.py \ --config configs/datasets/imagenet200/imagenet200.yml \ configs/datasets/imagenet200/imagenet200_${SCHEME}.yml \ configs/networks/mcd_net.yml \ configs/preprocessors/base_preprocessor.yml \ configs/pipelines/test/test_ood.yml \ configs/postprocessors/mcd.yml \ --network.backbone.name resnet18_224x224 \ --network.pretrained True \ --network.checkpoint 'results/imagenet200_oe_mcd_mcd_e90_lr0.1_default/s0/last_epoch90_acc0.8410.ckpt' \ --num_workers 8 \ --evaluator.ood_scheme ${SCHEME} \ --seed 0
972
41.304348
108
sh
null
OpenOOD-main/scripts/ood/mcd/imagenet200_train_mcd.sh
#!/bin/bash # sh scripts/ood/mcd/imagenet200_train_mcd.sh python main.py \ --config configs/datasets/imagenet200/imagenet200.yml \ configs/datasets/imagenet200/imagenet200_oe.yml \ configs/networks/mcd_net.yml \ configs/preprocessors/base_preprocessor.yml \ configs/pipelines/train/baseline.yml \ configs/pipelines/train/train_mcd.yml \ --network.backbone.name resnet18_224x224 \ --network.pretrained False \ --trainer.start_epoch_ft 80 \ --optimizer.num_epochs 90 \ --dataset.train.batch_size 128 \ --num_gpus 2 --num_workers 16 \ --merge_option merge \ --seed 0
621
31.736842
59
sh
null
OpenOOD-main/scripts/ood/mds/cifar100_test_ood_mds.sh
#!/bin/bash # sh scripts/ood/mds/cifar100_test_ood_mds.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mds.yml \ --network.checkpoint 'results/cifar100_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar100 \ --root ./results/cifar100_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor mds \ --save-score --save-csv
1,106
31.558824
95
sh
null
OpenOOD-main/scripts/ood/mds/cifar10_test_ood_mds.sh
#!/bin/bash # sh scripts/ood/mds/cifar10_test_ood_mds.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mds.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar10_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 1 \ --postprocessor.postprocessor_args.temperature 1 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar10 \ --root ./results/cifar10_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor mds \ --save-score --save-csv
1,190
31.189189
96
sh
null
OpenOOD-main/scripts/ood/mds/imagenet200_test_ood_mds.sh
#!/bin/bash # sh scripts/ood/mds/imagenet200_test_ood_mds.sh ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor mds \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor mds \ --save-score --save-csv --fsood
708
28.541667
74
sh
null
OpenOOD-main/scripts/ood/mds/imagenet_test_ood_mds.sh
#!/bin/bash # sh scripts/ood/mds/imagenet_test_ood_mds.sh GPU=1 CPU=1 node=63 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/datasets/imagenet/imagenet_ood.yml \ configs/networks/resnet50.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mds.yml \ --num_workers 4 \ --ood_dataset.image_size 256 \ --dataset.test.batch_size 256 \ --dataset.val.batch_size 256 \ --network.pretrained True \ --network.checkpoint 'results/pretrained_weights/resnet50_imagenet1k_v1.pth' \ --merge_option merge ############################################ # we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood_imagenet.py # available architectures: # resnet50, swin-t, vit-b-16 # ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor mds \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor mds \ --save-score --save-csv --fsood
1,386
27.895833
82
sh
null
OpenOOD-main/scripts/ood/mds_ensemble/cifar100_test_ood_mds_ensemble.sh
#!/bin/bash # sh scripts/ood/mds_ensemble/cifar100_test_ood_mds_ensemble.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mds_ensemble.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar100_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 0 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar100 \ --root ./results/cifar100_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor mds_ensemble \ --save-score --save-csv
1,175
31.666667
97
sh
null
OpenOOD-main/scripts/ood/mds_ensemble/cifar10_test_ood_mds_ensemble.sh
#!/bin/bash # sh scripts/ood/mds_ensemble/cifar10_test_ood_mds_ensemble.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mds_ensemble.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar10_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 0 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar10 \ --root ./results/cifar10_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor mds_ensemble \ --save-score --save-csv
1,167
31.444444
96
sh
null
OpenOOD-main/scripts/ood/mds_ensemble/imagenet200_test_ood_mds_ensemble.sh
#!/bin/bash # sh scripts/ood/mds_ensemble/imagenet200_test_ood_mds_ensemble.sh ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor mds_ensemble \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor mds_ensemble \ --save-score --save-csv --fsood
744
30.041667
74
sh
null
OpenOOD-main/scripts/ood/mds_ensemble/imagenet_test_ood_mds.sh
#!/bin/bash # sh scripts/ood/mds_ensemble/imagenet_test_ood_mds_ensemble.sh GPU=1 CPU=1 node=39 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/datasets/imagenet/imagenet_ood.yml \ configs/networks/resnet50.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mds_ensemble.yml \ --num_workers 4 \ --ood_dataset.image_size 256 \ --dataset.test.batch_size 256 \ --dataset.val.batch_size 256 \ --network.pretrained True \ --network.checkpoint 'results/pretrained_weights/resnet50_imagenet1k_v1.pth' \ --merge_option merge ############################################ # we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood_imagenet.py # available architectures: # resnet50 # ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor mds_ensemble \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor mds_ensemble \ --save-score --save-csv --fsood
1,405
28.291667
82
sh
null
OpenOOD-main/scripts/ood/mds_ensemble/mnist_test_ood_mds_ensemble.sh
#!/bin/bash # sh scripts/ood/mds_ensemble/mnist_test_ood_mds_ensemble.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/mnist/mnist.yml \ configs/datasets/mnist/mnist_ood.yml \ configs/networks/lenet.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mds_ensemble.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/mnist_lenet_acc99.30.ckpt' \ --mark 0
714
28.791667
74
sh
null
OpenOOD-main/scripts/ood/mds_ensemble/mnist_test_osr_mds_ensemble.sh
#!/bin/bash # sh scripts/ood/mds_ensemble/mnist_test_osr_mds_ensemble.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/osr_mnist6/mnist6_seed1.yml \ configs/datasets/osr_mnist6/mnist6_seed1_ood.yml \ configs/networks/lenet.yml \ configs/pipelines/test/test_osr.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mds_ensemble.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/osr/mnist6_seed1.ckpt' \ --mark 0
734
29.625
73
sh
null
OpenOOD-main/scripts/ood/mds_ensemble/sweep_osr.py
# python scripts/ood/mds/sweep_osr.py import os config = [ [ 'osr_cifar6/cifar6_seed1.yml', 'osr_cifar6/cifar6_seed1_ood.yml', 'resnet18_32x32', 'results/checkpoints/osr/cifar6_seed1.ckpt' ], [ 'osr_cifar50/cifar50_seed1.yml', 'osr_cifar50/cifar50_seed1_ood.yml', 'resnet18_32x32', 'results/checkpoints/osr/cifar50_seed1.ckpt' ], [ 'osr_tin20/tin20_seed1.yml', 'osr_tin20/tin20_seed1_ood.yml', 'resnet18_64x64', 'results/checkpoints/osr/tin20_seed1.ckpt' ], [ 'osr_mnist6/mnist6_seed1.yml', 'osr_mnist6/mnist6_seed1_ood.yml', 'lenet', 'results/checkpoints/osr/mnist6_seed1.ckpt' ], ] for [dataset, ood_dataset, network, pth] in config: command = (f"PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:1 -n1 \ --cpus-per-task=1 --ntasks-per-node=1 \ --kill-on-bad-exit=1 --job-name=openood \ python main.py \ --config configs/datasets/{dataset} \ configs/datasets/{ood_dataset} \ configs/networks/{network}.yml \ configs/pipelines/test/test_osr.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mds.yml \ --network.pretrained True \ --network.checkpoint {pth} \ --num_workers 8 \ --merge_option merge &") os.system(command)
1,322
32.075
77
py
null
OpenOOD-main/scripts/ood/mixoe/cifar100_test_mixoe.sh
#!/bin/bash # sh scripts/ood/mixoe/cifar100_test_mixoe.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_oe_resnet18_32x32_mixoe_e10_lr0.001_alpha0.1_beta1.0_cutmix_lam1.0_default \ --postprocessor msp \ --save-score --save-csv
499
32.333333
105
sh
null
OpenOOD-main/scripts/ood/mixoe/cifar100_train_mixoe.sh
#!/bin/bash # sh scripts/ood/mixoe/cifar100_train_mixoe.sh SEED=0 python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_oe.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/train/baseline.yml \ configs/pipelines/train/train_mixoe.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint ./results/cifar100_resnet18_32x32_base_e100_lr0.1_default/s${SEED}/best.ckpt \ --optimizer.lr 0.001 \ --optimizer.num_epochs 10 \ --dataset.train.batch_size 128 \ --dataset.oe.batch_size 128 \ --seed ${SEED}
652
33.368421
103
sh
null
OpenOOD-main/scripts/ood/mixoe/cifar10_test_mixoe.sh
#!/bin/bash # sh scripts/ood/mixoe/cifar10_test_mixoe.sh ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar10 \ --root ./results/cifar10_oe_resnet18_32x32_mixoe_e10_lr0.001_alpha0.1_beta1.0_cutmix_lam1.0_default \ --postprocessor msp \ --save-score --save-csv
496
32.133333
104
sh
null
OpenOOD-main/scripts/ood/mixoe/cifar10_train_mixoe.sh
#!/bin/bash # sh scripts/ood/mixoe/cifar10_train_mixoe.sh SEED=0 python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_oe.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/train/baseline.yml \ configs/pipelines/train/train_mixoe.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint ./results/cifar10_resnet18_32x32_base_e100_lr0.1_default/s${SEED}/best.ckpt \ --optimizer.lr 0.001 \ --optimizer.num_epochs 10 \ --dataset.train.batch_size 128 \ --dataset.oe.batch_size 128 \ --seed ${SEED}
646
33.052632
102
sh
null
OpenOOD-main/scripts/ood/mixoe/imagenet200_test_mixoe.sh
#!/bin/bash # sh scripts/ood/mixoe/imagenet200_test_mixoe.sh ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_oe_resnet18_224x224_mixoe_e10_lr0.001_alpha0.1_beta1.0_cutmix_lam1.0_default \ --postprocessor msp \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_oe_resnet18_224x224_mixoe_e10_lr0.001_alpha0.1_beta1.0_cutmix_lam1.0_default \ --postprocessor msp \ --save-score --save-csv --fsood
774
31.291667
110
sh
null
OpenOOD-main/scripts/ood/mixoe/imagnet200_train_mixoe.sh
#!/bin/bash # sh scripts/ood/mixoe/imagenet200_train_mixoe.sh SEED=0 python main.py \ --config configs/datasets/imagenet200/imagenet200.yml \ configs/datasets/imagenet200/imagenet200_oe.yml \ configs/networks/resnet18_224x224.yml \ configs/pipelines/train/baseline.yml \ configs/pipelines/train/train_mixoe.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default/s${SEED}/best.ckpt \ --optimizer.lr 0.001 \ --optimizer.num_epochs 10 \ --dataset.train.batch_size 128 \ --num_gpus 2 --num_workers 16 \ --merge_option merge \ --seed ${SEED}
702
34.15
107
sh
null
OpenOOD-main/scripts/ood/mls/cifar100_test_ood_maxlogit.sh
#!/bin/bash # sh scripts/ood/mls/cifar100_test_ood_maxlogit.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p mediasuper -x SZ-IDC1-10-112-2-17 --gres=gpu:${GPU} \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mls.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar100_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 0 ############################################ # 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_default \ --postprocessor mls \ --save-score --save-csv
1,131
30.444444
97
sh
null
OpenOOD-main/scripts/ood/mls/cifar10_test_ood_maxlogit.sh
#!/bin/bash # sh scripts/ood/mls/cifar10_test_ood_maxlogit.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p mediasuper -x SZ-IDC1-10-112-2-17 --gres=gpu:${GPU} \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mls.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar10_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 0 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar10 \ --root ./results/cifar10_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor mls \ --save-score --save-csv
1,123
30.222222
96
sh
null
OpenOOD-main/scripts/ood/mls/imagenet200_test_ood_maxlogit.sh
#!/bin/bash # sh scripts/ood/mls/imagenet200_test_ood_maxlogit.sh ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor mls \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor mls \ --save-score --save-csv --fsood
713
28.75
74
sh
null
OpenOOD-main/scripts/ood/mls/imagenet_test_ood_maxlogit.sh
#!/bin/bash # sh scripts/ood/mls/imagenet_test_ood_maxlogit.sh GPU=1 CPU=1 node=39 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/datasets/imagenet/imagenet_ood.yml \ configs/networks/resnet50.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mls.yml \ --num_workers 4 \ --ood_dataset.image_size 256 \ --dataset.test.batch_size 256 \ --dataset.val.batch_size 256 \ --network.pretrained True \ --network.checkpoint 'results/pretrained_weights/resnet50_imagenet1k_v1.pth' \ --merge_option merge ############################################ # we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood_imagenet.py # available architectures: # resnet50, swin-t, vit-b-16 # ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor mls \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor mls \ --save-score --save-csv --fsood
1,383
27.833333
82
sh
null
OpenOOD-main/scripts/ood/mls/mnist_test_ood_maxlogit.sh
#!/bin/bash # sh scripts/ood/mls/mnist_test_ood_maxlogit.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p mediasuper -x SZ-IDC1-10-112-2-17 --gres=gpu:${GPU} \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} \ python main.py \ --config configs/datasets/mnist/mnist.yml \ configs/datasets/mnist/mnist_ood.yml \ configs/networks/lenet.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mls.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/mnist_lenet_acc99.30.ckpt' \ --mark 0
643
25.833333
70
sh
null
OpenOOD-main/scripts/ood/mls/mnist_test_osr_maxlogit.sh
#!/bin/bash # sh scripts/ood/mls/mnist_test_osr_maxlogit.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p mediasuper -x SZ-IDC1-10-112-2-17 --gres=gpu:${GPU} \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} \ python main.py \ --config configs/datasets/osr_mnist6/mnist6_seed1.yml \ configs/datasets/osr_mnist6/mnist6_seed1_ood.yml \ configs/networks/lenet.yml \ configs/pipelines/test/test_osr.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mls.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/osr/mnist6_seed1.ckpt' \ --mark 0
663
26.666667
66
sh
null
OpenOOD-main/scripts/ood/mls/sweep_osr.py
# python scripts/ood/mls/sweep_osr.py import os config = [ [ 'osr_cifar6/cifar6_seed1.yml', 'osr_cifar6/cifar6_seed1_ood.yml', 'resnet18_32x32', 'results/checkpoints/osr/cifar6_seed1.ckpt' ], [ 'osr_cifar50/cifar50_seed1.yml', 'osr_cifar50/cifar50_seed1_ood.yml', 'resnet18_32x32', 'results/checkpoints/osr/cifar50_seed1.ckpt' ], [ 'osr_tin20/tin20_seed1.yml', 'osr_tin20/tin20_seed1_ood.yml', 'resnet18_64x64', 'results/checkpoints/osr/tin20_seed1.ckpt' ], [ 'osr_mnist6/mnist6_seed1.yml', 'osr_mnist6/mnist6_seed1_ood.yml', 'lenet', 'results/checkpoints/osr/mnist6_seed1.ckpt' ], ] for [dataset, ood_dataset, network, pth] in config: command = (f"PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:1 -n1 \ --cpus-per-task=1 --ntasks-per-node=1 \ --kill-on-bad-exit=1 --job-name=openood \ python main.py \ --config configs/datasets/{dataset} \ configs/datasets/{ood_dataset} \ configs/networks/{network}.yml \ configs/pipelines/test/test_osr.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/mls.yml \ --network.pretrained True \ --network.checkpoint {pth} \ --num_workers 8 \ --merge_option merge &") os.system(command)
1,322
32.075
77
py
null
OpenOOD-main/scripts/ood/mos/cifar100_test_mos.sh
#!/bin/bash # sh scripts/ood/mos/cifar100_test_mos.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ SEED=0 python main.py \ --config configs/datasets/cifar100/cifar100_double_label.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_mos.yml \ configs/postprocessors/mos.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint results/cifar100_double_label_resnet18_32x32_mos_e30_lr0.003/s${SEED}/best.ckpt \ --num_workers 8 \ --seed ${SEED}
613
25.695652
106
sh
null
OpenOOD-main/scripts/ood/mos/cifar100_train_mos.sh
#!/bin/bash # sh scripts/ood/mos/cifar100_train_mos.sh # GPU=1 # CPU=0 # node=73 # jobname=openood # PYTHONPATH='.':$PYTHONPATH \ SEED=0 python main.py \ --config configs/datasets/cifar100/cifar100_double_label.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/train/train_mos.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint ./results/cifar100_resnet18_32x32_base_e100_lr0.1_default/s${SEED}/best.ckpt \ --optimizer.num_epochs 30 \ --merge_option merge \ --seed ${SEED}
576
23.041667
103
sh
null
OpenOOD-main/scripts/ood/mos/cifar10_test_mos.sh
#!/bin/bash # sh scripts/ood/mos/cifar10_test_mos.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ SEED=0 python main.py \ --config configs/datasets/cifar10/cifar10_double_label.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_mos.yml \ configs/postprocessors/mos.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint results/cifar10_double_label_resnet18_32x32_mos_e30_lr0.003/s${SEED}/best.ckpt \ --num_workers 8 \ --seed ${SEED}
607
25.434783
105
sh
null
OpenOOD-main/scripts/ood/mos/cifar10_train_mos.sh
#!/bin/bash # sh scripts/ood/mos/cifar10_train_mos.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ SEED=0 python main.py \ --config configs/datasets/cifar10/cifar10_double_label.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/train/train_mos.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint ./results/cifar10_resnet18_32x32_base_e100_lr0.1_default/s${SEED}/best.ckpt \ --optimizer.num_epochs 30 \ --merge_option merge \ --seed ${SEED}
568
24.863636
102
sh
null
OpenOOD-main/scripts/ood/mos/imagenet200_test_mos.sh
#!/bin/bash # sh scripts/ood/mos/imagenet200_test_mos.sh SEED=0 # ood python main.py \ --config configs/datasets/imagenet200/imagenet200_double_label.yml \ configs/datasets/imagenet200/imagenet200_ood.yml \ configs/networks/resnet18_224x224.yml \ configs/pipelines/test/test_mos.yml \ configs/postprocessors/mos.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint results/imagenet200_double_label_resnet18_224x224_mos_e10_lr0.003/s${SEED}/best.ckpt \ --num_workers 8 \ --seed ${SEED} # full-spectrum ood python main.py \ --config configs/datasets/imagenet200/imagenet200_double_label.yml \ configs/datasets/imagenet200/imagenet200_double_label_fsood.yml \ configs/networks/resnet18_224x224.yml \ configs/pipelines/test/test_mos.yml \ configs/postprocessors/mos.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint results/imagenet200_double_label_resnet18_224x224_mos_e10_lr0.003/s${SEED}/best.ckpt \ --evaluator.ood_scheme fsood \ --num_workers 8 \ --seed ${SEED}
1,149
34.9375
111
sh
null
OpenOOD-main/scripts/ood/mos/imagenet200_train_mos.sh
#!/bin/bash # sh scripts/ood/mos/imagenet200_train_mos.sh SEED=0 python main.py \ --config configs/datasets/imagenet200/imagenet200_double_label.yml \ configs/networks/resnet18_224x224.yml \ configs/pipelines/train/train_mos.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default/s${SEED}/best.ckpt \ --optimizer.num_epochs 10 \ --dataset.train.batch_size 128 \ --num_gpus 2 --num_workers 16 \ --merge_option merge \ --seed ${SEED}
585
33.470588
107
sh
null
OpenOOD-main/scripts/ood/mos/imagenet_test_mos.sh
#!/bin/bash # sh scripts/ood/mos/imagenet_test_mos.sh SEED=0 # ood python main.py \ --config configs/datasets/imagenet/imagenet_double_label.yml \ configs/datasets/imagenet/imagenet_ood.yml \ configs/networks/resnet50.yml \ configs/pipelines/test/test_mos.yml \ configs/postprocessors/mos.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint results/imagenet_double_label_resnet50_mos_e5_lr0.003/s0/best.ckpt \ --num_workers 8 \ --seed 0 # full-spectrum ood python main.py \ --config configs/datasets/imagenet/imagenet_double_label.yml \ configs/datasets/imagenet/imagenet_double_label_fsood.yml \ configs/networks/resnet50.yml \ configs/pipelines/test/test_mos.yml \ configs/postprocessors/mos.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint results/imagenet_double_label_resnet50_mos_e5_lr0.003/s0/best.ckpt \ --num_workers 8 \ --seed 0 \ --evaluator.ood_scheme fsood
1,058
32.09375
93
sh
null
OpenOOD-main/scripts/ood/mos/imagenet_train_mos.sh
#!/bin/bash # sh scripts/ood/mos/imagenet_train_mos.sh python main.py \ --config configs/datasets/imagenet/imagenet_double_label.yml \ configs/networks/resnet50.yml \ configs/pipelines/train/train_mos.yml \ configs/preprocessors/base_preprocessor.yml \ --network.pretrained True \ --network.checkpoint ./results/pretrained_weights/resnet50_imagenet1k_v1.pth \ --optimizer.num_epochs 5 \ --dataset.train.batch_size 128 \ --num_gpus 2 --num_workers 16 \ --merge_option merge \ --seed 0
529
32.125
82
sh
null
OpenOOD-main/scripts/ood/msp/cifar100_test_ood_msp.sh
#!/bin/bash # sh scripts/ood/msp/cifar100_test_ood_msp.sh # GPU=1 # CPU=1 # node=36 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ #srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ #--cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ #--kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/msp.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar100_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 0 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar100 \ --root ./results/cifar100_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor msp \ --save-score --save-csv
1,135
31.457143
97
sh
null
OpenOOD-main/scripts/ood/msp/cifar10_test_ood_msp.sh
#!/bin/bash # sh scripts/ood/msp/cifar10_test_ood_msp.sh # GPU=1 # CPU=1 # node=36 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ #srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ #--cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ #--kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/msp.yml \ --num_workers 8 \ --network.checkpoint './results/cifar10_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 0 \ --merge_option merge ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar10 \ --root ./results/cifar10_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor msp \ --save-score --save-csv
1,156
31.138889
98
sh
null
OpenOOD-main/scripts/ood/msp/imagenet200_test_ood_msp.sh
#!/bin/bash # sh scripts/ood/msp/imagenet200_test_ood_msp.sh ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor msp \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor msp \ --save-score --save-csv --fsood
708
28.541667
74
sh
null
OpenOOD-main/scripts/ood/msp/imagenet_test_ood_msp.sh
#!/bin/bash # sh scripts/ood/msp/imagenet_test_ood_msp.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ #srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ #--cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ #--kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/datasets/imagenet/imagenet_ood.yml \ configs/networks/resnet50.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/msp.yml \ --num_workers 10 \ --ood_dataset.image_size 256 \ --dataset.test.batch_size 256 \ --dataset.val.batch_size 256 \ --network.pretrained True \ --network.checkpoint 'results/pretrained_weights/resnet50_imagenet1k_v1.pth' \ --merge_option merge ############################################ # we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood_imagenet.py # available architectures: # resnet50, swin-t, vit-b-16 # ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor msp \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor msp \ --save-score --save-csv --fsood
1,383
27.244898
82
sh
null
OpenOOD-main/scripts/ood/msp/mnist_test_fsood_msp.sh
#!/bin/bash # sh scripts/ood/msp/mnist_test_fsood_msp.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/mnist/mnist.yml \ configs/datasets/mnist/mnist_fsood.yml \ configs/networks/lenet.yml \ configs/pipelines/test/test_fsood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/msp.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/mnist_lenet_acc98.50.ckpt' \ --mark 0
649
26.083333
73
sh
null
OpenOOD-main/scripts/ood/msp/mnist_test_ood_msp.sh
#!/bin/bash # sh scripts/ood/msp/mnist_test_ood_msp.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/mnist/mnist.yml \ configs/datasets/mnist/mnist_ood.yml \ configs/networks/lenet.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/msp.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/mnist_lenet_acc98.50.ckpt' \ --mark 0
643
25.833333
73
sh
null
OpenOOD-main/scripts/ood/msp/osr_cifar6_test_msp.sh
#!/bin/bash # sh scripts/ood/msp/osr_cifar6_test_msp.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/osr_cifar6/cifar6_seed1.yml \ configs/datasets/osr_cifar6/cifar6_seed1.yml \ configs/preprocessors/base_preprocessor.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_acc.yml \ --num_workers 8 \ --network.checkpoint './results/checkpoints/osr/cifar6_seed1.ckpt'
620
28.571429
71
sh
null
OpenOOD-main/scripts/ood/msp/osr_mnist6_test_msp.sh
#!/bin/bash # sh scripts/ood/msp/mnist_test_osr_msp.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/osr_mnist6/mnist6_seed1.yml \ configs/datasets/osr_mnist6/mnist6_seed1_ood.yml \ configs/networks/lenet.yml \ configs/pipelines/test/test_osr.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/msp.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/osr/mnist6_seed1.ckpt' \ --mark 0
656
27.565217
71
sh
null
OpenOOD-main/scripts/ood/msp/sweep_osr.py
# python scripts/ood/msp/sweep_osr.py import os config = [ [ 'osr_cifar6/cifar6_seed1.yml', 'osr_cifar6/cifar6_seed1_ood.yml', 'resnet18_32x32', 'results/checkpoints/osr/cifar6_seed1.ckpt' ], [ 'osr_cifar50/cifar50_seed1.yml', 'osr_cifar50/cifar50_seed1_ood.yml', 'resnet18_32x32', 'results/checkpoints/osr/cifar50_seed1.ckpt' ], [ 'osr_tin20/tin20_seed1.yml', 'osr_tin20/tin20_seed1_ood.yml', 'resnet18_64x64', 'results/checkpoints/osr/tin20_seed1.ckpt' ], [ 'osr_mnist6/mnist6_seed1.yml', 'osr_mnist6/mnist6_seed1_ood.yml', 'lenet', 'results/checkpoints/osr/mnist6_seed1.ckpt' ], ] for [dataset, ood_dataset, network, pth] in config: command = (f"PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:1 -n1 \ --cpus-per-task=1 --ntasks-per-node=1 \ --kill-on-bad-exit=1 --job-name=openood \ python main.py \ --config configs/datasets/{dataset} \ configs/datasets/{ood_dataset} \ configs/networks/{network}.yml \ configs/pipelines/test/test_osr.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/msp.yml \ --network.pretrained True \ --network.checkpoint {pth} \ --num_workers 8 \ --merge_option merge &") os.system(command)
1,322
32.075
77
py
null
OpenOOD-main/scripts/ood/npos/cifar100_test_npos.sh
#!/bin/bash # sh scripts/ood/npos/cifar100_test_npos.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_npos_net_npos_e100_lr0.1_default \ --postprocessor npos \ --save-score --save-csv
456
29.466667
63
sh
null
OpenOOD-main/scripts/ood/npos/cifar100_train_npos.sh
#!/bin/bash # sh scripts/ood/npos/cifar100_train_npos.sh python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/networks/npos_net.yml \ configs/pipelines/train/train_npos.yml \ configs/preprocessors/base_preprocessor.yml \ --preprocessor.name cider \ --network.backbone.name resnet18_32x32 \ --dataset.train.batch_size 256 \ --trainer.trainer_args.temp 0.1 \ --trainer.trainer_args.sample_from 600 \ --trainer.trainer_args.K 300 \ --trainer.trainer_args.cov_mat 0.1 \ --trainer.trainer_args.start_epoch_KNN 40 \ --trainer.trainer_args.ID_points_num 200 \ --optimizer.num_epochs 100 \ --optimizer.lr 0.1 \ --seed 0
699
32.333333
53
sh
null
OpenOOD-main/scripts/ood/npos/cifar10_test_npos.sh
#!/bin/bash # sh scripts/ood/npos/cifar10_test_npos.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 cifar10 \ --root ./results/cifar10_npos_net_npos_e100_lr0.1_default \ --postprocessor npos \ --save-score --save-csv
453
29.266667
62
sh
null
OpenOOD-main/scripts/ood/npos/cifar10_train_npos.sh
#!/bin/bash # sh scripts/ood/npos/cifar10_train_npos.sh python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/networks/npos_net.yml \ configs/pipelines/train/train_npos.yml \ configs/preprocessors/base_preprocessor.yml \ --preprocessor.name cider \ --network.backbone.name resnet18_32x32 \ --dataset.train.batch_size 256 \ --trainer.trainer_args.temp 0.1 \ --trainer.trainer_args.sample_from 600 \ --trainer.trainer_args.K 300 \ --trainer.trainer_args.cov_mat 0.1 \ --trainer.trainer_args.start_epoch_KNN 40 \ --trainer.trainer_args.ID_points_num 200 \ --optimizer.num_epochs 100 \ --optimizer.lr 0.1 \ --seed 0
696
32.190476
51
sh
null
OpenOOD-main/scripts/ood/npos/imagenet200_test_npos.sh
#!/bin/bash # sh scripts/ood/npos/imagenet200_test_npos.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 # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_npos_net_npos_e90_lr0.1_default \ --postprocessor npos \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_npos_net_npos_e90_lr0.1_default \ --postprocessor npos \ --save-score --save-csv --fsood
684
27.541667
65
sh
null
OpenOOD-main/scripts/ood/npos/imagenet200_train_npos.sh
#!/bin/bash # sh scripts/ood/npos/imagenet200_train_npos.sh # NPOS trainer cannot work with multiple GPUs (DDP) currently python main.py \ --config configs/datasets/imagenet200/imagenet200.yml \ configs/networks/npos_net.yml \ configs/pipelines/train/train_npos.yml \ configs/preprocessors/base_preprocessor.yml \ --preprocessor.name cider \ --network.backbone.name resnet18_224x224 \ --dataset.train.batch_size 256 \ --trainer.trainer_args.temp 0.1 \ --trainer.trainer_args.sample_from 1000 \ --trainer.trainer_args.K 400 \ --trainer.trainer_args.cov_mat 0.1 \ --trainer.trainer_args.start_epoch_KNN 40 \ --trainer.trainer_args.ID_points_num 300 \ --optimizer.num_epochs 90 \ --optimizer.lr 0.1 \ --num_gpus 1 --num_workers 16 \ --merge_option merge \ --seed 0
835
33.833333
61
sh
null
OpenOOD-main/scripts/ood/npos/imagenet_train_npos.sh
#!/bin/bash # sh scripts/ood/npos/imagenet_train_npos.sh # NPOS trainer cannot work with multiple GPUs (DDP) currently # we observed CUDA OOM error on Quadro RTX 6000 24GB GPU python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/networks/npos_net.yml \ configs/pipelines/train/train_npos.yml \ configs/preprocessors/base_preprocessor.yml \ --preprocessor.name cider \ --network.backbone.name resnet50 \ --network.backbone.pretrained True \ --network.backbone.checkpoint ./results/pretrained_weights/resnet50_imagenet1k_v1.pth \ --optimizer.lr 0.001 \ --optimizer.num_epochs 30 \ --dataset.train.batch_size 128 \ --trainer.trainer_args.temp 0.1 \ --trainer.trainer_args.sample_from 1000 \ --trainer.trainer_args.K 400 \ --trainer.trainer_args.cov_mat 0.1 \ --trainer.trainer_args.start_epoch_KNN 1 \ --trainer.trainer_args.ID_points_num 300 \ --num_gpus 1 --num_workers 16 \ --merge_option merge \ --seed 0
1,009
36.407407
91
sh
null
OpenOOD-main/scripts/ood/odin/cifar100_test_ood_odin.sh
#!/bin/bash # sh scripts/ood/odin/cifar100_test_ood_odin.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/odin.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar100_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 0 ############################################ # 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_default \ --postprocessor odin \ --save-score --save-csv
1,143
30.777778
97
sh
null
OpenOOD-main/scripts/ood/odin/cifar10_test_ood_odin.sh
#!/bin/bash # sh scripts/ood/odin/cifar10_test_ood_odin.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/odin.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar10_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 0 ############################################ # 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 cifar10 \ --root ./results/cifar10_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor odin \ --save-score --save-csv
1,135
30.555556
96
sh
null
OpenOOD-main/scripts/ood/odin/imagenet200_test_ood_odin.sh
#!/bin/bash # sh scripts/ood/odin/imagenet200_test_ood_odin.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 # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor odin \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor odin \ --save-score --save-csv --fsood
712
28.708333
74
sh
null
OpenOOD-main/scripts/ood/odin/imagenet_test_ood_odin.sh
#!/bin/bash # sh scripts/ood/odin/imagenet_test_ood_odin.sh GPU=1 CPU=1 node=39 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/datasets/imagenet/imagenet_ood.yml \ configs/networks/resnet50.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/odin.yml \ --num_workers 4 \ --ood_dataset.image_size 256 \ --dataset.test.batch_size 256 \ --dataset.val.batch_size 256 \ --network.pretrained True \ --network.checkpoint 'results/pretrained_weights/resnet50_imagenet1k_v1.pth' \ --merge_option merge ############################################ # we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood_imagenet.py # available architectures: # resnet50, swin-t, vit-b-16 # ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor odin \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor odin \ --save-score --save-csv --fsood
1,383
27.833333
82
sh
null
OpenOOD-main/scripts/ood/odin/mnist_test_ood_odin.sh
#!/bin/bash # sh scripts/ood/odin/mnist_test_ood_odin.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/mnist/mnist.yml \ configs/datasets/mnist/mnist_ood.yml \ configs/networks/lenet.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/odin.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/mnist_lenet_acc99.30.ckpt' \ --mark 0
654
26.291667
73
sh
null
OpenOOD-main/scripts/ood/odin/mnist_test_ood_odin_aps.sh
#!/bin/bash # sh scripts/ood/odin/mnist_test_ood_odin_aps.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/mnist/mnist.yml \ configs/datasets/mnist/mnist_ood.yml \ configs/networks/lenet.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/odin.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/mnist_lenet_acc98.50.ckpt' \ --mark 0
658
26.458333
73
sh
null
OpenOOD-main/scripts/ood/odin/mnist_test_osr_odin.sh
#!/bin/bash # sh scripts/ood/odin/mnist_test_osr_odin.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/osr_mnist6/mnist6_seed1.yml \ configs/datasets/osr_mnist6/mnist6_seed1_ood.yml \ configs/networks/lenet.yml \ configs/pipelines/test/test_osr.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/odin.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/osr/mnist6_seed1.ckpt' \ --mark 0
674
27.125
73
sh
null
OpenOOD-main/scripts/ood/odin/sweep_osr.py
# python scripts/ood/odin/sweep_osr.py import os config = [ # ['osr_cifar6/cifar6_seed1.yml', 'osr_cifar6/cifar6_seed1_ood.yml', 'resnet18_32x32', 'results/checkpoints/osr/cifar6_seed1.ckpt'], # ['osr_cifar50/cifar50_seed1.yml', 'osr_cifar50/cifar50_seed1_ood.yml', 'resnet18_32x32', 'results/checkpoints/osr/cifar50_seed1.ckpt'], # ['osr_tin20/tin20_seed1.yml', 'osr_tin20/tin20_seed1_ood.yml', 'resnet18_64x64', 'results/checkpoints/osr/tin20_seed1.ckpt'], [ 'osr_mnist6/mnist6_seed1.yml', 'osr_mnist6/mnist6_seed1_ood.yml', 'lenet', 'results/checkpoints/osr/mnist6_seed1.ckpt' ], ] for [dataset, ood_dataset, network, pth] in config: command = (f"PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:1 -n1 \ --cpus-per-task=1 --ntasks-per-node=1 \ --kill-on-bad-exit=1 --job-name=openood \ python main.py \ --config configs/datasets/{dataset} \ configs/datasets/{ood_dataset} \ configs/networks/{network}.yml \ configs/pipelines/test/test_osr.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/odin.yml \ --network.pretrained True \ --network.checkpoint {pth} \ --num_workers 8 \ --merge_option merge &") os.system(command)
1,264
39.806452
141
py
null
OpenOOD-main/scripts/ood/oe/cifar100_test_oe.sh
#!/bin/bash # sh scripts/ood/oe/cifar100_test_oe.sh GPU=1 CPU=1 node=63 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/msp.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar100_oe_resnet18_32x32_oe_e100_lr0.1_lam0.5_default/s0/best.ckpt' \ --mark 0 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar100 \ --root ./results/cifar100_oe_resnet18_32x32_oe_e100_lr0.1_lam0.5_default \ --postprocessor msp \ --save-score --save-csv
1,134
31.428571
105
sh
null
OpenOOD-main/scripts/ood/oe/cifar100_train_oe.sh
#!/bin/bash # sh scripts/ood/oe/cifar100_train_oe.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} \ -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_oe.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/train/baseline.yml \ configs/pipelines/train/train_oe.yml \ configs/preprocessors/base_preprocessor.yml \ --seed 0
602
26.409091
53
sh
null
OpenOOD-main/scripts/ood/oe/cifar10_test_oe.sh
#!/bin/bash # sh scripts/ood/oe/cifar10_test_oe.sh GPU=1 CPU=1 node=63 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/msp.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar10_oe_resnet18_32x32_oe_e100_lr0.1_lam0.5_default/s0/best.ckpt' \ --mark 0 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar10 \ --root ./results/cifar10_oe_resnet18_32x32_oe_e100_lr0.1_lam0.5_default \ --postprocessor msp \ --save-score --save-csv
1,126
31.2
104
sh
null
OpenOOD-main/scripts/ood/oe/cifar10_train_oe.sh
#!/bin/bash # sh scripts/ood/oe/cifar10_train_oe.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} \ -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_oe.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/train/baseline.yml \ configs/pipelines/train/train_oe.yml \ configs/preprocessors/base_preprocessor.yml \ --seed 0
597
26.181818
51
sh
null
OpenOOD-main/scripts/ood/oe/imagenet200_test_oe.sh
#!/bin/bash # sh scripts/ood/oe/imagenet200_test_oe.sh ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_oe_resnet18_224x224_oe_e90_lr0.1_lam0.5_default \ --postprocessor msp \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_oe_resnet18_224x224_oe_e90_lr0.1_lam0.5_default \ --postprocessor msp \ --save-score --save-csv --fsood
710
28.625
81
sh
null
OpenOOD-main/scripts/ood/oe/imagnet200_train_oe.sh
#!/bin/bash # sh scripts/ood/oe/imagenet200_train_oe.sh python main.py \ --config configs/datasets/imagenet200/imagenet200.yml \ configs/datasets/imagenet200/imagenet200_oe.yml \ configs/networks/resnet18_224x224.yml \ configs/pipelines/train/baseline.yml \ configs/pipelines/train/train_oe.yml \ configs/preprocessors/base_preprocessor.yml \ --optimizer.num_epochs 90 \ --dataset.train.batch_size 128 \ --num_gpus 2 --num_workers 16 \ --merge_option merge \ --seed 0
513
31.125
59
sh
null
OpenOOD-main/scripts/ood/rankfeat/cifar100_test_ood_rankfeat.sh
#!/bin/bash # sh scripts/ood/rankfeat/cifar100_test_ood_rankfeat.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/rankfeat.yml \ --network.checkpoint 'results/cifar100_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar100 \ --root ./results/cifar100_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor rankfeat \ --save-score --save-csv
1,126
32.147059
95
sh
null
OpenOOD-main/scripts/ood/rankfeat/cifar10_test_ood_rankfeat.sh
#!/bin/bash # sh scripts/ood/rankfeat/cifar10_test_ood_rankfeat.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/rankfeat.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar10_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 1 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar10 \ --root ./results/cifar10_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor rankfeat \ --save-score --save-csv
1,155
31.111111
96
sh
null
OpenOOD-main/scripts/ood/rankfeat/imagenet200_test_ood_rankfeat.sh
#!/bin/bash # sh scripts/ood/rankfeat/imagenet200_test_ood_rankfeat.sh ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor rankfeat \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor rankfeat \ --save-score --save-csv --fsood
728
29.375
74
sh
null
OpenOOD-main/scripts/ood/rankfeat/imagenet_test_ood_rankfeat.sh
#!/bin/bash # sh scripts/ood/rankfeat/imagenet_test_ood_rankfeat.sh GPU=1 CPU=1 node=63 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/datasets/imagenet/imagenet_ood.yml \ configs/networks/resnet50.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/rankfeat.yml \ --num_workers 4 \ --ood_dataset.image_size 256 \ --dataset.test.batch_size 256 \ --dataset.val.batch_size 256 \ --network.pretrained True \ --network.checkpoint 'results/pretrained_weights/resnet50_imagenet1k_v1.pth' \ --merge_option merge ############################################ # we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood_imagenet.py # available architectures: # resnet50, swin-t, vit-b-16 # ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor rankfeat \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor rankfeat \ --save-score --save-csv --fsood
1,411
28.416667
82
sh
null
OpenOOD-main/scripts/ood/react/cifar100_test_ood_react.sh
#!/bin/bash # sh scripts/ood/react/cifar100_test_ood_react.sh # GPU=1 # CPU=1 # node=36 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ #srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ #--cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ #--kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/react_net.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/react.yml \ --network.pretrained False \ --network.backbone.name resnet18_32x32 \ --network.backbone.pretrained True \ --network.backbone.checkpoint 'results/cifar100_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --num_workers 8 \ --mark 0 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar100 \ --root ./results/cifar100_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor react \ --save-score --save-csv
1,266
32.342105
106
sh
null
OpenOOD-main/scripts/ood/react/cifar10_test_ood_react.sh
#!/bin/bash # sh scripts/ood/react/cifar10_test_ood_react.sh # GPU=1 # CPU=1 # node=36 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ #srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ #--cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ #--kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/react_net.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/react.yml \ --network.pretrained False \ --network.backbone.name resnet18_32x32 \ --network.backbone.pretrained True \ --network.backbone.checkpoint 'results/cifar10_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --num_workers 8 \ --mark fixed_0 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar10 \ --root ./results/cifar10_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor react \ --save-score --save-csv
1,264
32.289474
105
sh
null
OpenOOD-main/scripts/ood/react/imagenet200_test_ood_react.sh
#!/bin/bash # sh scripts/ood/react/imagenet200_test_ood_react.sh ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor react \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor react \ --save-score --save-csv --fsood
716
28.875
74
sh
null
OpenOOD-main/scripts/ood/react/imagenet_test_ood_react.sh
#!/bin/bash # sh scripts/ood/react/imagenet_test_ood_react.sh GPU=1 CPU=1 node=63 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/datasets/imagenet/imagenet_ood.yml \ configs/networks/react_net.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/react.yml \ --num_workers 4 \ --ood_dataset.image_size 256 \ --dataset.test.batch_size 256 \ --dataset.val.batch_size 256 \ --network.pretrained False \ --network.backbone.checkpoint 'results/pretrained_weights/resnet50_imagenet1k_v1.pth' \ --merge_option merge ############################################ # we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood_imagenet.py # available architectures: # resnet50, swin-t, vit-b-16 # ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor react \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor react \ --save-score --save-csv --fsood
1,399
28.166667
91
sh
null
OpenOOD-main/scripts/ood/react/mnist_test_ood_react.sh
#!/bin/bash # sh scripts/ood/react/mnist_test_ood_react.sh # GPU=1 # CPU=1 # node=36 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ #srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ #--cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ #--kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/mnist/mnist.yml \ configs/datasets/mnist/mnist_ood.yml \ configs/networks/react_net.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/react.yml \ --network.pretrained False \ --network.backbone.name lenet \ --network.backbone.pretrained True \ --network.backbone.checkpoint 'results/checkpoints/mnist_lenet_acc99.30.ckpt' \ --num_workers 8 \ --mark 0
764
28.423077
79
sh
null
OpenOOD-main/scripts/ood/react/mnist_test_osr_react.sh
#!/bin/bash # sh scripts/ood/react/mnist_test_osr_react.sh # GPU=1 # CPU=1 # node=36 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ #srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ #--cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ #--kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/osr_mnist6/mnist6_seed1.yml \ configs/datasets/osr_mnist6/mnist6_seed1_ood.yml \ configs/networks/react_net.yml \ configs/pipelines/test/test_osr.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/react.yml \ --network.pretrained False \ --network.backbone.name lenet \ --network.backbone.pretrained True \ --network.backbone.checkpoint 'results/checkpoints/osr/mnist6_seed1.ckpt' \ --num_workers 8 \ --mark 0
784
29.192308
75
sh
null
OpenOOD-main/scripts/ood/react/sweep_osr.py
# python scripts/ood/react/sweep_osr.py import os config = [ [ 'osr_cifar6/cifar6_seed1.yml', 'osr_cifar6/cifar6_seed1_ood.yml', 'resnet18_32x32', 'results/checkpoints/osr/cifar6_seed1.ckpt' ], [ 'osr_cifar50/cifar50_seed1.yml', 'osr_cifar50/cifar50_seed1_ood.yml', 'resnet18_32x32', 'results/checkpoints/osr/cifar50_seed1.ckpt' ], [ 'osr_tin20/tin20_seed1.yml', 'osr_tin20/tin20_seed1_ood.yml', 'resnet18_64x64', 'results/checkpoints/osr/tin20_seed1.ckpt' ], [ 'osr_mnist6/mnist6_seed1.yml', 'osr_mnist6/mnist6_seed1_ood.yml', 'lenet', 'results/checkpoints/osr/mnist6_seed1.ckpt' ], ] for [dataset, ood_dataset, network, pth] in config: command = (f"PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:1 -n1 \ --cpus-per-task=1 --ntasks-per-node=1 \ --kill-on-bad-exit=1 --job-name=openood \ python main.py \ --config configs/datasets/{dataset} \ configs/datasets/{ood_dataset} \ configs/networks/react_net.yml \ configs/pipelines/test/test_osr.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/react.yml \ --network.pretrained False \ --network.backbone.name {network} \ --network.backbone.checkpoint {pth} \ --num_workers 8 \ --merge_option merge &") os.system(command)
1,376
32.585366
77
py
null
OpenOOD-main/scripts/ood/residual/cifar100_test_ood_residual.sh
#!/bin/bash # sh scripts/ood/residual/cifar100_test_ood_residual.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p mediasuper -x SZ-IDC1-10-112-2-17 --gres=gpu:${GPU} \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/residual.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar100_resnet18_32x32_base_e100_lr0.1/best.ckpt' \ --mark 0
689
27.75
82
sh
null
OpenOOD-main/scripts/ood/residual/cifar10_test_ood_residual.sh
#!/bin/bash # sh scripts/ood/residual/cifar10_test_ood_residual.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p mediasuper -x SZ-IDC1-10-112-2-17 --gres=gpu:${GPU} \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/residual.yml \ --num_workers 8 \ --network.checkpoint 'results/checkpoints/cifar10_res18_acc94.30.ckpt' \ --mark 0
674
27.125
72
sh
null
OpenOOD-main/scripts/ood/residual/imagenet_test_ood_residual.sh
#!/bin/bash # sh scripts/ood/residual/imagenet_test_ood_residual.sh GPU=1 CPU=1 node=73 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p mediasuper -x SZ-IDC1-10-112-2-17 --gres=gpu:${GPU} \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} \ python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/datasets/imagenet/imagenet_ood.yml \ configs/networks/vit.yml \ configs/pipelines/test/test_ood.yml \ configs/postprocessors/residual.yml \ --num_workers 8 \ --network.checkpoint ./checkpoints/vit-base-p16_in21k-pre-3rdparty_ft-64xb64_in1k-384_20210928-98e8652b.pth \ --mark 0
644
28.318182
109
sh
null
OpenOOD-main/scripts/ood/rmds/cifar100_test_ood_rmds.sh
#!/bin/bash # sh scripts/ood/rmds/cifar100_test_ood_rmds.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/datasets/cifar100/cifar100_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/rmds.yml \ --network.checkpoint 'results/cifar100_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar100 \ --root ./results/cifar100_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor rmds \ --save-score --save-csv
1,110
31.676471
95
sh
null
OpenOOD-main/scripts/ood/rmds/cifar10_test_ood_rmds.sh
#!/bin/bash # sh scripts/ood/rmds/cifar10_test_ood_rmds.sh # GPU=1 # CPU=1 # node=73 # jobname=openood PYTHONPATH='.':$PYTHONPATH \ # srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ # --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ # --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/datasets/cifar10/cifar10_ood.yml \ configs/networks/resnet18_32x32.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/rmds.yml \ --num_workers 8 \ --network.checkpoint 'results/cifar10_resnet18_32x32_base_e100_lr0.1_default/s0/best.ckpt' \ --mark 1 ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs python scripts/eval_ood.py \ --id-data cifar10 \ --root ./results/cifar10_resnet18_32x32_base_e100_lr0.1_default \ --postprocessor rmds \ --save-score --save-csv
1,139
30.666667
96
sh
null
OpenOOD-main/scripts/ood/rmds/imagenet200_test_ood_rmds.sh
#!/bin/bash # sh scripts/ood/rmds/imagenet200_test_ood_rmds.sh ############################################ # alternatively, we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood.py # especially if you want to get results from # multiple runs # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor rmds \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_resnet18_224x224_base_e90_lr0.1_default \ --postprocessor rmds \ --save-score --save-csv --fsood
712
28.708333
74
sh
null
OpenOOD-main/scripts/ood/rmds/imagenet_test_ood_rmds.sh
#!/bin/bash # sh scripts/ood/rmds/imagenet_test_ood_rmds.sh GPU=1 CPU=1 node=63 jobname=openood PYTHONPATH='.':$PYTHONPATH \ srun -p dsta --mpi=pmi2 --gres=gpu:${GPU} -n1 \ --cpus-per-task=${CPU} --ntasks-per-node=${GPU} \ --kill-on-bad-exit=1 --job-name=${jobname} -w SG-IDC1-10-51-2-${node} \ python main.py \ --config configs/datasets/imagenet/imagenet.yml \ configs/datasets/imagenet/imagenet_ood.yml \ configs/networks/resnet50.yml \ configs/pipelines/test/test_ood.yml \ configs/preprocessors/base_preprocessor.yml \ configs/postprocessors/rmds.yml \ --num_workers 4 \ --ood_dataset.image_size 256 \ --dataset.test.batch_size 256 \ --dataset.val.batch_size 256 \ --network.pretrained True \ --network.checkpoint 'results/pretrained_weights/resnet50_imagenet1k_v1.pth' \ --merge_option merge ############################################ # we recommend using the # new unified, easy-to-use evaluator with # the example script scripts/eval_ood_imagenet.py # available architectures: # resnet50, swin-t, vit-b-16 # ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor rmds \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood_imagenet.py \ --tvs-pretrained \ --arch resnet50 \ --postprocessor rmds \ --save-score --save-csv --fsood
1,391
28
82
sh
null
OpenOOD-main/scripts/ood/rotpred/cifar100_test_rotpred.sh
#!/bin/bash # sh scripts/ood/rotpred/cifar100_test_rotpred.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_rot_net_rotpred_e100_lr0.1_default \ --postprocessor rotpred \ --save-score --save-csv
467
30.2
65
sh
null
OpenOOD-main/scripts/ood/rotpred/cifar100_train_rotpred.sh
#!/bin/bash # sh scripts/ood/rotpred/cifar100_train_rotpred.sh python main.py \ --config configs/datasets/cifar100/cifar100.yml \ configs/networks/rot_net.yml \ configs/pipelines/train/baseline.yml \ configs/preprocessors/base_preprocessor.yml \ --trainer.name rotpred \ --seed 0
305
26.818182
53
sh
null
OpenOOD-main/scripts/ood/rotpred/cifar10_test_rotpred.sh
#!/bin/bash # sh scripts/ood/rotpred/cifar10_test_rotpred.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 cifar10 \ --root ./results/cifar10_rot_net_rotpred_e100_lr0.1_default \ --postprocessor rotpred \ --save-score --save-csv
464
30
64
sh
null
OpenOOD-main/scripts/ood/rotpred/cifar10_train_rotpred.sh
#!/bin/bash # sh scripts/ood/rotpred/cifar10_train_rotpred.sh python main.py \ --config configs/datasets/cifar10/cifar10.yml \ configs/networks/rot_net.yml \ configs/pipelines/train/baseline.yml \ configs/preprocessors/base_preprocessor.yml \ --trainer.name rotpred \ --seed 0
302
26.545455
51
sh
null
OpenOOD-main/scripts/ood/rotpred/imagenet200_test_rotpred.sh
#!/bin/bash # sh scripts/ood/rotpred/imagenet200_test_rotpred.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 # ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_rot_net_rotpred_e90_lr0.1_default \ --postprocessor rotpred \ --save-score --save-csv #--fsood # full-spectrum ood python scripts/eval_ood.py \ --id-data imagenet200 \ --root ./results/imagenet200_rot_net_rotpred_e90_lr0.1_default \ --postprocessor rotpred \ --save-score --save-csv --fsood
700
28.208333
67
sh
null
OpenOOD-main/scripts/ood/rotpred/imagenet200_train_rotpred.sh
#!/bin/bash # sh scripts/ood/rotpred/imagenet200_train_rotpred.sh python main.py \ --config configs/datasets/imagenet200/imagenet200.yml \ configs/networks/rot_net.yml \ configs/pipelines/train/baseline.yml \ configs/preprocessors/base_preprocessor.yml \ --network.backbone.name resnet18_224x224 \ --trainer.name rotpred \ --optimizer.num_epochs 90 \ --dataset.train.batch_size 128 \ --num_gpus 2 --num_workers 16 \ --merge_option merge \ --seed 0
493
29.875
59
sh