YOLOv5使用torchsummary时出现AttributeError: ‘list‘ object has no attribute ‘size‘解决

卸载torchsummary，按照torch-summary，中间多了一个横杠，它是torchsummary的加强版。这里我们卸掉torchsummary库，安装torch-summary库，上面的问题就可以解决了。

在anaconda的环境里面，先卸载再按照

pip uninstall torchsummary
pip install torch-summary==1.4.4

我尝试了好几种方式打印yolov5 中6.0版本的网络结构，还是torchsummary好用！！

from models.yolo import *
from torchsummary import summary

FILE = Path(__file__).resolve()
ROOT = FILE.parents[1]  # YOLOv5 root directory
if str(ROOT) not in sys.path:
    sys.path.append(str(ROOT))  # add ROOT to PATH
# ROOT = ROOT.relative_to(Path.cwd())  # relative

if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    # 官方权重 torch.float16
    parser.add_argument('--weights', type=str, default=r'D:\codePythonBinocularvision\yolov5-6.0\weight\yolov5s.pt',
                        help='weights path')
    # 自训练权重 torch.float16
    # parser.add_argument('--weights', type=str, default=r'D:\codePythonBinocularvision\yolov5-6.0\weight\best.pt',
    #                     help='weights path')
    opt = parser.parse_args()

    # Load pytorch model on GPU
    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')  # 检查是否有可用的 GPU
    model = torch.load(opt.weights, map_location=device)['model']  # 将模型加载到设备上

    # 转换为FloatTensor float32
    model = model.float()

    # 将模型的权重和参数移到 GPU 上
    model.to(device)
    
    # 打印模型结构
    print(model)
    # 遍历各个层次并打印结构信息
    # for name, layer in model.named_children():
    #     print(f"\nLayer: {name}")
    #     print(layer)

    for name, parameters in model.named_parameters():
        # print(name, ':', parameters.size())
        print(parameters.dtype)

    # 官方代码打印结构
    model_info(model, True)

    # yolo.py文件结构里面的打印模型
    parser = argparse.ArgumentParser()
    parser.add_argument('--cfg', type=str, default='yolov5s.yaml', help='model.yaml')
    parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
    parser.add_argument('--profile', action='store_true', help='profile model speed')
    opt = parser.parse_args()
    opt.cfg = check_yaml(opt.cfg)  # check YAML

    print_args(FILE.stem, opt)
    set_logging()
    device = select_device(opt.device)

    # Create model
    model = Model(opt.cfg).to(device)
    model.train()

    # 使用summary函数来打印模型
    input_shape = [640, 640]
    m = Model(opt.cfg).to(device)
    summary(m, (3, input_shape[0], input_shape[1]))