{"id":13499043,"url":"https://github.com/bmsookim/wide-resnet.pytorch","last_synced_at":"2025-03-29T03:32:15.079Z","repository":{"id":43210964,"uuid":"93605206","full_name":"bmsookim/wide-resnet.pytorch","owner":"bmsookim","description":"Best CIFAR-10, CIFAR-100 results with wide-residual networks using PyTorch","archived":false,"fork":false,"pushed_at":"2020-01-09T13:32:46.000Z","size":674,"stargazers_count":462,"open_issues_count":14,"forks_count":129,"subscribers_count":14,"default_branch":"master","last_synced_at":"2024-12-12T02:02:17.788Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/bmsookim.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}},"created_at":"2017-06-07T07:23:51.000Z","updated_at":"2024-12-11T15:28:15.000Z","dependencies_parsed_at":"2022-08-12T10:21:18.013Z","dependency_job_id":null,"html_url":"https://github.com/bmsookim/wide-resnet.pytorch","commit_stats":null,"previous_names":["bmsookim/wide-resnet.pytorch","meliketoy/wide-resnet.pytorch"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bmsookim%2Fwide-resnet.pytorch","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bmsookim%2Fwide-resnet.pytorch/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bmsookim%2Fwide-resnet.pytorch/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bmsookim%2Fwide-resnet.pytorch/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/bmsookim","download_url":"https://codeload.github.com/bmsookim/wide-resnet.pytorch/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":246135746,"owners_count":20729056,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","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":[],"created_at":"2024-07-31T22:00:27.417Z","updated_at":"2025-03-29T03:32:14.727Z","avatar_url":"https://github.com/bmsookim.png","language":"Python","funding_links":[],"categories":["Papers\u0026Codes"],"sub_categories":["WRN"],"readme":"\u003cp align=\"center\"\u003e\u003cimg width=\"40%\" src=\"./imgs/pytorch.png\"\u003e\u003c/p\u003e\n\nBest CIFAR-10, CIFAR-100 results with wide-residual networks using PyTorch\n\nPytorch Implementation of Sergey Zagoruyko's [Wide Residual Networks](https://arxiv.org/pdf/1605.07146v2.pdf)\n\nFor Torch implementations, see [here](https://github.com/meliketoy/wide-residual-network).\n\n## Requirements\nSee the [installation instruction](INSTALL.md) for a step-by-step installation guide.\nSee the [server instruction](SERVER.md) for server settup.\n- Install [cuda-8.0](https://developer.nvidia.com/cuda-downloads)\n- Install [cudnn v5.1](https://developer.nvidia.com/cudnn)\n- Download [Pytorch 2.7](https://pytorch.org) and clone the repository.\n```bash\npip install http://download.pytorch.org/whl/cu80/torch-0.1.12.post2-cp27-none-linux_x86_64.whl\npip install torchvision\ngit clone https://github.com/meliketoy/wide-resnet.pytorch\n```\n\n## How to run\nAfter you have cloned the repository, you can train each dataset of either cifar10, cifar100 by running the script below.\n```bash\npython main --lr 0.1 resume false --net_type [lenet/vggnet/resnet/wide-resnet] --depth 28 --widen_factor 10 --dropout_rate 0.3 --dataset [cifar10/cifar100] \n```\n\n## Implementation Details\n\n|   epoch   | learning rate |  weight decay | Optimizer | Momentum | Nesterov |\n|:---------:|:-------------:|:-------------:|:---------:|:--------:|:--------:|\n|   0 ~ 60  |      0.1      |     0.0005    | Momentum  |    0.9   |   true   |\n|  61 ~ 120 |      0.02     |     0.0005    | Momentum  |    0.9   |   true   |\n| 121 ~ 160 |     0.004     |     0.0005    | Momentum  |    0.9   |   true   |\n| 161 ~ 200 |     0.0008    |     0.0005    | Momentum  |    0.9   |   true   |\n\n## CIFAR-10 Results\n \n![alt tag](imgs/cifar10_image.png)\n\nBelow is the result of the test set accuracy for **CIFAR-10 dataset** training.\n\n**Accuracy is the average of 5 runs**\n\n| network           | dropout | preprocess | GPU:0 | GPU:1 | per epoch    | accuracy(%) |\n|:-----------------:|:-------:|:----------:|:-----:|:-----:|:------------:|:-----------:|\n| wide-resnet 28x10 |    0    |     ZCA    | 5.90G |   -   | 2 min 03 sec |    95.83    |\n| wide-resnet 28x10 |    0    |   meanstd  | 5.90G |   -   | 2 min 03 sec |    96.21    |\n| wide-resnet 28x10 |   0.3   |   meanstd  | 5.90G |   -   | 2 min 03 sec |    96.27    |\n| wide-resnet 28x20 |   0.3   |   meanstd  | 8.13G | 6.93G | 4 min 10 sec |  **96.55**  |\n| wide-resnet 40x10 |   0.3   |   meanstd  | 8.08G |   -   | 3 min 13 sec |    96.31    |\n| wide-resnet 40x14 |   0.3   |   meanstd  | 7.37G | 6.46G | 3 min 23 sec |    96.34    |\n\n## CIFAR-100 Results\n\n![alt tag](imgs/cifar100_image.png)\n\nBelow is the result of the test set accuracy for **CIFAR-100 dataset** training.\n\n**Accuracy is the average of 5 runs**\n\n| network           | dropout |  preprocess | GPU:0 | GPU:1 | per epoch    | Top1 acc(%)| Top5 acc(%) |\n|:-----------------:|:-------:|:-----------:|:-----:|:-----:|:------------:|:----------:|:-----------:|\n| wide-resnet 28x10 |    0    |     ZCA     | 5.90G |   -   | 2 min 03 sec |    80.07   |    95.02    |\n| wide-resnet 28x10 |    0    |   meanstd   | 5.90G |   -   | 2 min 03 sec |    81.02   |    95.41    |\n| wide-resnet 28x10 |   0.3   |   meanstd   | 5.90G |   -   | 2 min 03 sec |    81.49   |    95.62    |\n| wide-resnet 28x20 |   0.3   |   meanstd   | 8.13G | 6.93G | 4 min 05 sec |  **82.45** |  **96.11**  |\n| wide-resnet 40x10 |   0.3   |   meanstd   | 8.93G |   -   | 3 min 06 sec |    81.42   |    95.63    |\n| wide-resnet 40x14 |   0.3   |   meanstd   | 7.39G | 6.46G | 3 min 23 sec |    81.87   |    95.51    |\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbmsookim%2Fwide-resnet.pytorch","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fbmsookim%2Fwide-resnet.pytorch","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbmsookim%2Fwide-resnet.pytorch/lists"}