cspdarknet53.ra_in1k 模型卡片

CSP-DarkNet（跨阶段部分网络）图像分类模型。基于下述训练方案模板，在 timm 框架中使用 ImageNet-1k 数据集进行训练。

方案详情：

RandAugment RA 数据增强方案。灵感来源于并改进自 EfficientNet RandAugment 方案。在论文《ResNet Strikes Back》中被公布为 B 方案（https://arxiv.org/abs/2110.00476）。
优化器：RMSProp（采用 TF 1.0 行为），使用 EMA 权重平均
学习率调度：带预热的 Step 调度（指数衰减，带阶梯式下降）

模型详情

模型类型：图像分类 / 特征骨干网络
模型统计信息：
- 参数数量（M）：27.6
- GMACs：6.6
- 激活值数量（M）：16.8
- 图像尺寸：256 x 256
相关论文：
- CSPNet: A New Backbone that can Enhance Learning Capability of CNN: https://arxiv.org/abs/1911.11929
- YOLOv3: An Incremental Improvement: https://arxiv.org/abs/1804.02767
- ResNet strikes back: An improved training procedure in timm: https://arxiv.org/abs/2110.00476
原始代码库：https://github.com/huggingface/pytorch-image-models

模型使用

图像分类


# from urllib.request import urlopen
from PIL import Image
import timm
import argparse
import torch
from openmind import is_torch_npu_available


def parse_args():
    parser = argparse.ArgumentParser()
    parser.add_argument(
        "--model_name_or_path",
        type=str,
        help="Path to the model file",
        default=None,
    )
    args = parser.parse_args()
    return args


if __name__ == '__main__':
    args = parse_args()
    if is_torch_npu_available():
        device = "npu:0"
    else:
        device = "cpu"
    model_path = args.model_name_or_path
    img = model_path + "/Img/beignets-task-guide.png"
    img = Image.open(img)

    model_name = 'cspdarknet53.ra_in1k'
    checkpoint_path = model_path + '/pytorch_model.bin'
    model = timm.create_model(model_name, pretrained=False, checkpoint_path=checkpoint_path).to(device)
    model = model.eval()

    # get model specific transforms (normalization, resize)
    data_config = timm.data.resolve_model_data_config(model)
    transforms = timm.data.create_transform(**data_config, is_training=False)

    output = model(transforms(img).unsqueeze(0).to(device))  # unsqueeze single image into batch of 1

    top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)
    print("top5_probabilities", top5_probabilities)
    # top5_probabilities tensor([[12.7346, 12.3356,  7.7930,  5.9564,  5.0362]], device='npu:0',
    #    grad_fn=<TopkBackward0>)
    print("top5_class_indices", top5_class_indices)
    # top5_class_indices tensor([[960, 969, 967, 968, 505]], device='npu:0')

模型对比

在 timm 的模型结果中探索此模型的数据集和运行时指标。

引用

@article{Wang2019CSPNetAN,
  title={CSPNet: A New Backbone that can Enhance Learning Capability of CNN},
  author={Chien-Yao Wang and Hong-Yuan Mark Liao and I-Hau Yeh and Yueh-Hua Wu and Ping-Yang Chen and Jun-Wei Hsieh},
  journal={2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)},
  year={2019},
  pages={1571-1580}
}

@article{Redmon2018YOLOv3AI,
  title={YOLOv3: An Incremental Improvement},
  author={Joseph Redmon and Ali Farhadi},
  journal={ArXiv},
  year={2018},
  volume={abs/1804.02767}
}

@inproceedings{wightman2021resnet,
  title={ResNet strikes back: An improved training procedure in timm},
  author={Wightman, Ross and Touvron, Hugo and Jegou, Herve},
  booktitle={NeurIPS 2021 Workshop on ImageNet: Past, Present, and Future}
}

@misc{rw2019timm,
  author = {Ross Wightman},
  title = {PyTorch Image Models},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub repository},
  doi = {10.5281/zenodo.4414861},
  howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
}

cspdarknet53.ra_in1k 模型卡片

CSP-DarkNet（跨阶段部分网络）图像分类模型。基于下述训练方案模板，在 timm 框架中使用 ImageNet-1k 数据集进行训练。

方案详情：

RandAugment RA 数据增强方案。灵感来源于并改进自 EfficientNet RandAugment 方案。在论文《ResNet Strikes Back》中被公布为 B 方案（https://arxiv.org/abs/2110.00476）。
优化器：RMSProp（采用 TF 1.0 行为），使用 EMA 权重平均
学习率调度：带预热的 Step 调度（指数衰减，带阶梯式下降）

模型详情

模型类型：图像分类 / 特征骨干网络
模型统计信息：
- 参数数量（M）：27.6
- GMACs：6.6
- 激活值数量（M）：16.8
- 图像尺寸：256 x 256
相关论文：
- CSPNet: A New Backbone that can Enhance Learning Capability of CNN: https://arxiv.org/abs/1911.11929
- YOLOv3: An Incremental Improvement: https://arxiv.org/abs/1804.02767
- ResNet strikes back: An improved training procedure in timm: https://arxiv.org/abs/2110.00476
原始代码库：https://github.com/huggingface/pytorch-image-models

模型使用

图像分类


# from urllib.request import urlopen
from PIL import Image
import timm
import argparse
import torch
from openmind import is_torch_npu_available


def parse_args():
    parser = argparse.ArgumentParser()
    parser.add_argument(
        "--model_name_or_path",
        type=str,
        help="Path to the model file",
        default=None,
    )
    args = parser.parse_args()
    return args


if __name__ == '__main__':
    args = parse_args()
    if is_torch_npu_available():
        device = "npu:0"
    else:
        device = "cpu"
    model_path = args.model_name_or_path
    img = model_path + "/Img/beignets-task-guide.png"
    img = Image.open(img)

    model_name = 'cspdarknet53.ra_in1k'
    checkpoint_path = model_path + '/pytorch_model.bin'
    model = timm.create_model(model_name, pretrained=False, checkpoint_path=checkpoint_path).to(device)
    model = model.eval()

    # get model specific transforms (normalization, resize)
    data_config = timm.data.resolve_model_data_config(model)
    transforms = timm.data.create_transform(**data_config, is_training=False)

    output = model(transforms(img).unsqueeze(0).to(device))  # unsqueeze single image into batch of 1

    top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)
    print("top5_probabilities", top5_probabilities)
    # top5_probabilities tensor([[12.7346, 12.3356,  7.7930,  5.9564,  5.0362]], device='npu:0',
    #    grad_fn=<TopkBackward0>)
    print("top5_class_indices", top5_class_indices)
    # top5_class_indices tensor([[960, 969, 967, 968, 505]], device='npu:0')

模型对比

在 timm 的模型结果中探索此模型的数据集和运行时指标。

引用

@article{Wang2019CSPNetAN,
  title={CSPNet: A New Backbone that can Enhance Learning Capability of CNN},
  author={Chien-Yao Wang and Hong-Yuan Mark Liao and I-Hau Yeh and Yueh-Hua Wu and Ping-Yang Chen and Jun-Wei Hsieh},
  journal={2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)},
  year={2019},
  pages={1571-1580}
}

@article{Redmon2018YOLOv3AI,
  title={YOLOv3: An Incremental Improvement},
  author={Joseph Redmon and Ali Farhadi},
  journal={ArXiv},
  year={2018},
  volume={abs/1804.02767}
}

@inproceedings{wightman2021resnet,
  title={ResNet strikes back: An improved training procedure in timm},
  author={Wightman, Ross and Touvron, Hugo and Jegou, Herve},
  booktitle={NeurIPS 2021 Workshop on ImageNet: Past, Present, and Future}
}

@misc{rw2019timm,
  author = {Ross Wightman},
  title = {PyTorch Image Models},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub repository},
  doi = {10.5281/zenodo.4414861},
  howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
}