classification-cifar10-pytorch:通过PyTorch在cifar10数据集中训练几个经典分类网络

  • X2_898346
    了解作者
  • 105.1KB
    文件大小
  • zip
    文件格式
  • 0
    收藏次数
  • VIP专享
    资源类型
  • 0
    下载次数
  • 2022-06-13 23:32
    上传日期
分类-cifar10-pytorch 我正在PyTorch的cifar10数据集上测试几种经典分类网络的性能! 要求 火炬 火炬摘要 python3.x 结果 模型 我的帐户 总参数 估计总大小(MB) 可训练的参数 参数大小(MB) 保存的模型大小(MB) GPU内存使用率(MB) 92.64% 2,296,922 36.14 2,296,922 8.76 8.96 3107 94.27% 14,728,266 62.77 14,728,266 56.18 59.0 1229 94.70% 11,171,146 53.38 11,171,146 42.61 44.7 1665年 95.09% 9,128,778 99.84 9,128,778 34.82 36.7 5779 95.22% 23,520,842 155.
classification-cifar10-pytorch-master.zip
  • classification-cifar10-pytorch-master
  • models
  • resnet.py
    4.2KB
  • densenet.py
    3.6KB
  • lenet.py
    1.1KB
  • pnasnet.py
    4.2KB
  • dpn.py
    3.7KB
  • googlenet.py
    3.1KB
  • resnext.py
    3.6KB
  • senet.py
    3.9KB
  • __pycache__
  • efficientnet.cpython-36.pyc
    3.2KB
  • pnasnet.cpython-36.pyc
    4.5KB
  • mobilenet.cpython-36.pyc
    2.6KB
  • mobilenetv2.cpython-36.pyc
    3.2KB
  • resnext.cpython-36.pyc
    3.7KB
  • resnet.cpython-36.pyc
    4.6KB
  • senet.cpython-36.pyc
    3.8KB
  • shufflenet.cpython-36.pyc
    3.9KB
  • dpn.cpython-36.pyc
    3.8KB
  • preact_resnet.cpython-36.pyc
    4.7KB
  • densenet.cpython-36.pyc
    4.3KB
  • vgg.cpython-36.pyc
    1.8KB
  • lenet.cpython-36.pyc
    1.3KB
  • shufflenetv2.cpython-36.pyc
    5.3KB
  • __init__.cpython-36.pyc
    453B
  • googlenet.cpython-36.pyc
    2.8KB
  • __init__.py
    360B
  • mobilenetv2.py
    3.2KB
  • shufflenetv2.py
    5.4KB
  • vgg.py
    1.6KB
  • preact_resnet.py
    4.3KB
  • efficientnet.py
    3.2KB
  • mytest.py
    4.3KB
  • mobilenet.py
    2.1KB
  • shufflenet.py
    3.5KB
  • main.py
    4.9KB
  • gt_label.png
    19.2KB
  • utils.py
    3.4KB
  • main_ddp.py
    5.9KB
  • LICENSE
    1KB
  • vis-notebook.ipynb
    40.5KB
  • README.md
    2.5KB
内容介绍
# Classification-cifar10-pytorch I am testing several classical classification networks performance on cifar10 dataset by PyTorch! [Chinese blog](https://blog.csdn.net/laizi_laizi/article/details/103006497) # Requirements - pytorch - torchsummary - python3.x # Results | Model | My Acc. | Total params | Estimated Total Size (MB) | Trainable params | Params size (MB) |Saved model size (MB)|GPU memory usage(MB) | ----------------- | ----------- | ------ | ---|--- | --- | --- |--- | | [MobileNetV2](https://arxiv.org/abs/1801.04381) | 92.64% | 2,296,922 | 36.14 | 2,296,922 | 8.76 | 8.96 | 3107 | | [VGG16](https://arxiv.org/abs/1409.1556) | 94.27% | 14,728,266 | 62.77 |14,728,266 |56.18 |59.0 | 1229 | | [PreActResNet18](https://arxiv.org/abs/1603.05027) | 94.70% | 11,171,146 | 53.38 | 11,171,146 | 42.61 | 44.7 | 1665 | | [ResNeXt29(2x64d)](https://arxiv.org/abs/1611.05431) | 95.09% | 9,128,778 | 99.84 | 9,128,778 | 34.82 | 36.7 | 5779 | | [ResNet50](https://arxiv.org/abs/1512.03385) | 95.22% | 23,520,842 | 155.86 | 23,520,842 | 89.72 | 94.4 | 5723 | | [DPN92](https://arxiv.org/abs/1707.01629) | 95.42% | 34,236,634 | 243.50 | 34,236,634 | 130.60 | 137.5 | 10535 | | [ResNeXt29(32x4d)](https://arxiv.org/abs/1611.05431) | 95.49% | 4,774,218 | 83.22 | 4,774,218 | 18.21 | 19.2 | 5817 | | [DenseNet121](https://arxiv.org/abs/1608.06993) | 95.55% | 6,956,298 | 105.05 | 6,956,298 | 26.54 | 28.3 | 8203 | | [ResNet18](https://arxiv.org/abs/1512.03385) | 95.59% | 11,173,962 | 53.89 | 11,173,962 | 42.63 | 44.8 | 1615 | | [ResNet101](https://arxiv.org/abs/1512.03385) | 95.62% | 42,512,970 | 262.31 | 42,512,970 | 162.17 | 170.6 | 8857 | **Note**: 1. Above GPU memory usage(MB) was observed with batch size=128. 2. For PreActResNet18, I set initial learning rate=0.1, but it can't converge, so I set it's initial lr=0.01. 3. I firstly train **VGG16**, **ResNet18** and **ResNet50** with total epochs=400. But I want to get results earlier, so for remaining networks, I set total epochs=300 (besides, afterwards it just improve a little). 4. Run the scripts: `python main.py --resume --lr 0.1 --trainbs 128 --testbs 100` # Pre-trained models You can obtain pre-traind models(as above list) from here: [[Baidu Drive](https://pan.baidu.com/s/1oUfaxFnghIdClCFMf3A11Q)] [[Google Drive](https://drive.google.com/open?id=1PLwxkczvKq86ATRD7SB-5w31omuORUNV)]
评论
    相关推荐