Skip to content

Latest commit

 

History

History
166 lines (112 loc) · 7.32 KB

ssl_classification.md

File metadata and controls

166 lines (112 loc) · 7.32 KB
Raw

SSL Downstream Tasks: Classification

The benchmarks of downstream tasks are based on MMSelfSup. We provide many benchmarks to evaluate models on following downstream tasks. Here are comprehensive tutorials and examples to explain how to run all benchmarks with OpenMixup.

First, you are supposed to extract your backbone weights by tools/model_converters/extract_backbone_weights.py

python tools/model_converters/extract_backbone_weights.py {CHECKPOINT} {MODEL_FILE}

Arguments:

  • CHECKPOINT: the checkpoint file of a selfsup method named as epoch_*.pth.
  • MODEL_FILE: the output backbone weights file. If not mentioned, the PRETRAIN below uses this extracted model file.

ImageNet Linear Evaluation

The linear evaluation is one of the most general benchmarks for contrastive learning pre-training, we integrate several papers' config settings, also including multi-head linear evaluation. We write classification model in our own codebase for the multi-head function, thus, to run linear evaluation, we still use .sh script to launch training. The supported datasets are ImageNet, Places205, iNaturalist18, and CIFAR-10/100.

# distributed version
bash benchmarks/classification/dist_train_linear.sh ${CONFIG} ${PRETRAIN}

# slurm version
bash benchmarks/classification/srun_train_linear.sh ${PARTITION} ${JOB_NAME} ${CONFIG} ${PRETRAIN}

Remarks:

  • The default GPU number is 8. When changing GPUS, please also change imgs_per_gpu in the config file accordingly to ensure the total batch size (e.g., 256).
  • CONFIG: Use config files under configs/benchmarks/classification/. Specifically, imagenet (excluding imagenet_*percent folders), places205, inaturalist2018, and CIFAR-10/100 are supported.
  • PRETRAIN: the pre-trained model file (the backbone parameters only).

Example:

bash benchmarks/classification/dist_train_linear.sh \
configs/benchmarks/classification/imagenet/r50_linear_sz224_4xb64_step_ep100.py \
work_dir/pretrained_model.pth

(back to top)

ImageNet Finetune Evaluation

The fully finetuning evaluation is the popular benchmark for masked image modeling pre-training, we integrate several papers' config settings and use .sh script to launch training. The supported datasets are ImageNet and CIFAR-10/100.

# distributed version
bash benchmarks/classification/dist_train_ft_8gpu.sh ${CONFIG} ${PRETRAIN}

# slurm version
bash benchmarks/classification/srun_train_ft.sh ${PARTITION} ${JOB_NAME} ${CONFIG} ${PRETRAIN}

Remarks:

  • The default GPU number is 8. When changing GPUS, please also change imgs_per_gpu in the config file accordingly to ensure the total batch size.
  • CONFIG: Use config files under configs/benchmarks/classification/. Specifically, imagenet .
  • PRETRAIN: the pre-trained model file (the backbone parameters only).

Example:

bash benchmarks/classification/dist_train_ft_4gpu.sh \
configs/benchmarks/classification/imagenet/r50_rsb_a3_ft_sz160_4xb512_cos_fp16_ep100.py \
work_dir/pretrained_model.pth

(back to top)

ImageNet Semi-Supervised Classification

To run ImageNet semi-supervised classification, we still use .sh script as Linear Evaluation and Fine-tuning to launch training.

# distributed version
bash tools/benchmarks/classification/dist_train_semi.sh ${CONFIG} ${PRETRAIN}

# slurm version
bash tools/benchmarks/classification/slurm_train_semi.sh ${PARTITION} ${JOB_NAME} ${CONFIG} ${PRETRAIN}

Remarks:

  • The default GPU number is 4.
  • CONFIG: Use config files under configs/benchmarks/classification/imagenet/, named imagenet_*percent folders.
  • PRETRAIN: the pre-trained model file (the backbone parameters only).

(back to top)

ImageNet Nearest-Neighbor Classification

Only support CNN-style backbones (like ResNet50). To evaluate the pre-trained models using the nearest-neighbor benchmark, you can run command below.

# distributed version
bash benchmarks/classification/knn_imagenet/dist_test_knn_pretrain.sh ${SELFSUP_CONFIG} ${PRETRAIN}

# slurm version
bash tools/benchmarks/classification/knn_imagenet/slurm_test_knn_pretrain.sh ${PARTITION} ${JOB_NAME} ${SELFSUP_CONFIG} ${PRETRAIN}

Besides, if you want to evaluate the ckpt files saved by runner, you can run command below.

# distributed version
bash tools/benchmarks/classification/knn_imagenet/dist_test_knn_epoch.sh ${SELFSUP_CONFIG} ${EPOCH}

# slurm version
bash tools/benchmarks/classification/knn_imagenet/slurm_test_knn_epoch.sh ${PARTITION} ${JOB_NAME} ${SELFSUP_CONFIG} ${EPOCH}

To test with ckpt, the code uses the epoch_*.pth file, there is no need to extract weights.

Remarks:

  • ${SELFSUP_CONFIG} is the config file of the self-supervised experiment.
  • PRETRAIN: the pre-trained model file.
  • if you want to change GPU numbers, you could add GPUS_PER_NODE=4 GPUS=4 at the beginning of the command.
  • EPOCH is the epoch number of the ckpt that you want to test

(back to top)

VOC SVM / Low-shot SVM

To run these benchmarks, you should first prepare your VOC datasets. Please refer to prepare_data.md for the details of data preparation.

To evaluate the pre-trained models, you can run command below.

# distributed version
bash tools/benchmarks/classification/svm_voc07/dist_test_svm_pretrain.sh ${SELFSUP_CONFIG} ${GPUS} ${PRETRAIN} ${FEATURE_LIST}

# slurm version
bash tools/benchmarks/classification/svm_voc07/slurm_test_svm_pretrain.sh ${PARTITION} ${JOB_NAME} ${SELFSUP_CONFIG} ${PRETRAIN} ${FEATURE_LIST}

Besides, if you want to evaluate the ckpt files saved by runner, you can run command below.

# distributed version
bash tools/benchmarks/classification/svm_voc07/dist_test_svm_epoch.sh ${SELFSUP_CONFIG} ${EPOCH} ${FEATURE_LIST}

# slurm version
bash tools/benchmarks/classification/svm_voc07/slurm_test_svm_epoch.sh ${PARTITION} ${JOB_NAME} ${SELFSUP_CONFIG} ${EPOCH} ${FEATURE_LIST}

To test with ckpt, the code uses the epoch_*.pth file, there is no need to extract weights.

Remarks:

  • ${SELFSUP_CONFIG} is the config file of the self-supervised experiment.
  • ${FEATURE_LIST} is a string to specify features from layer1 to layer5 to evaluate; e.g., if you want to evaluate layer5 only, then FEATURE_LIST is "feat5", if you want to evaluate all features, then FEATURE_LIST is "feat1 feat2 feat3 feat4 feat5" (separated by space). If left empty, the default FEATURE_LIST is "feat5".
  • PRETRAIN: the pre-trained model file.
  • if you want to change GPU numbers, you could add GPUS_PER_NODE=4 GPUS=4 at the beginning of the command.
  • EPOCH is the epoch number of the ckpt that you want to test

(back to top)