{"id":13738186,"url":"https://github.com/hunto/image_classification_sota","last_synced_at":"2026-02-03T13:11:25.345Z","repository":{"id":39710153,"uuid":"468667253","full_name":"hunto/image_classification_sota","owner":"hunto","description":"Training ImageNet / CIFAR models with sota strategies and fancy techniques such as ViT, KD, Rep, etc.","archived":false,"fork":false,"pushed_at":"2024-03-20T04:26:48.000Z","size":213,"stargazers_count":82,"open_issues_count":7,"forks_count":14,"subscribers_count":4,"default_branch":"main","last_synced_at":"2025-05-08T16:45:26.469Z","etag":null,"topics":["cifar","image-classification","imagenet","kd","nas","pruning","pytorch","rep","transformer","vit"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/hunto.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null}},"created_at":"2022-03-11T08:24:16.000Z","updated_at":"2025-02-10T02:55:07.000Z","dependencies_parsed_at":"2024-03-19T07:09:25.661Z","dependency_job_id":"633c5e7e-2877-451e-95b2-eaf2ba502c68","html_url":"https://github.com/hunto/image_classification_sota","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/hunto/image_classification_sota","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hunto%2Fimage_classification_sota","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hunto%2Fimage_classification_sota/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hunto%2Fimage_classification_sota/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hunto%2Fimage_classification_sota/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/hunto","download_url":"https://codeload.github.com/hunto/image_classification_sota/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hunto%2Fimage_classification_sota/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29046503,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-03T10:09:22.136Z","status":"ssl_error","status_checked_at":"2026-02-03T10:09:16.814Z","response_time":96,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.6:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["cifar","image-classification","imagenet","kd","nas","pruning","pytorch","rep","transformer","vit"],"created_at":"2024-08-03T03:02:13.624Z","updated_at":"2026-02-03T13:11:25.325Z","avatar_url":"https://github.com/hunto.png","language":"Python","funding_links":[],"categories":["Python"],"sub_categories":[],"readme":"# Image Classification SOTA  \n\n`Image Classification SOTA` is an image classification toolbox based on PyTorch.\n\n## Updates  \n### May 27, 2022  \n* Add knowledge distillation methods (KD and [DIST](https://github.com/hunto/DIST_KD)).  \n\n### March 24, 2022  \n* Support training strategies in DeiT (ViT).\n\n### March 11, 2022  \n* Release training code.\n\n## Supported Algorithms  \n### Structural Re-parameterization (Rep)  \n* DBB (CVPR 2021) [[paper]](https://arxiv.org/abs/2103.13425) [[original repo]](https://github.com/DingXiaoH/DiverseBranchBlock)\n* DyRep (CVPR 2022) [[README]](https://github.com/hunto/DyRep)\n\n### Knowledge Distillation (KD)  \n* KD [[paper]](https://arxiv.org/abs/1503.02531)  \n* DIST [[README]](https://github.com/hunto/DIST_KD) [[paper]](https://arxiv.org/abs/2205.10536)  \n\n## Requirements\n```\ntorch\u003e=1.0.1\ntorchvision\n```\n\n## Getting Started  \n### Prepare datasets  \nIt is recommended to symlink the dataset root to `image_classification_sota/data`. Then the file structure should be like  \n```\nimage_classification_sota\n├── lib\n├── tools\n├── configs\n├── data\n│   ├── imagenet\n│   │   ├── meta\n│   │   ├── train\n│   │   ├── val\n│   ├── cifar\n│   │   ├── cifar-10-batches-py\n│   │   ├── cifar-100-python\n```\n\n### Training configurations  \n* `Strategies`: The training strategies are configured using yaml file or arguments. Examples are in `configs/strategies` directory.\n\n### Train a model  \n\n* Training with a single GPU  \n    ```shell\n    python tools/train.py -c ${CONFIG} --model ${MODEL} [optional arguments]\n    ```\n\n* Training with multiple GPUs\n    ```shell\n    sh tools/dist_train.sh ${GPU_NUM} ${CONFIG} ${MODEL} [optional arguments]\n    ```\n\n* For slurm users\n    ```shell\n    sh tools/slurm_train.sh ${PARTITION} ${GPU_NUM} ${CONFIG} ${MODEL} [optional arguments]\n    ```\n\n**Examples**  \n* Train ResNet-50 on ImageNet\n    ```shell\n    sh tools/dist_train.sh 8 configs/strategies/resnet/resnet.yaml resnet50 --experiment imagenet_res50\n    ```\n\n* Train MobileNetV2 on ImageNet\n    ```shell\n    sh tools/dist_train.sh 8 configs/strategies/MBV2/mbv2.yaml nas_model --model-config configs/models/MobileNetV2/MobileNetV2.yaml --experiment imagenet_mbv2\n    ```\n\n* Train VGG-16 on CIFAR-10\n    ```shell\n    sh tools/dist_train.sh 1 configs/strategies/CIFAR/cifar.yaml nas_model --model-config configs/models/VGG/vgg16_cifar10.yaml --experiment cifar10_vgg16\n    ```\n\n## Projects based on Image Classification SOTA  \n* [CVPR 2022] [DyRep](https://github.com/hunto/DyRep): Bootstrapping Training with Dynamic Re-parameterization\n* [NeurIPS 2022] [DIST](https://github.com/hunto/DIST_KD): Knowledge Distillation from A Stronger Teacher\n* [LightViT](https://github.com/hunto/LightViT): Towards Light-Weight Convolution-Free Vision Transformers\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhunto%2Fimage_classification_sota","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fhunto%2Fimage_classification_sota","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhunto%2Fimage_classification_sota/lists"}