Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

在试图使用训练好的模型进行评测时遇到了参数不对应的问题 #36

Open
charpng opened this issue Sep 26, 2024 · 0 comments

Comments

@charpng
Copy link

charpng commented Sep 26, 2024

请问在训练过程中保存下来的模型该如何加载呢?我在试图使用训练好的模型进行评测时遇到了参数不对应的问题,报错如下:
File "/DBCNN-PyTorch-master/DBCNN.py", line 53, in init
scnn.load_state_dict(torch.load(scnn_root))
File "/home/guest1/anaconda3/envs/wpf/lib/python3.10/site-packages/torch/nn/modules/module.py", line 2153, in load_state_dict
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for DataParallel:
Missing key(s) in state_dict: "module.features.0.weight", "module.features.0.bias", "module.features.1.weight", "module.features.1.bias", "module.features.1.running_mean", "module.features.1.running_var", "module.features.3.weight", "module.features.3.bias", "module.features.4.weight", "module.features.4.bias", "module.features.4.running_mean", "module.features.4.running_var", "module.features.6.weight", "module.features.6.bias", "module.features.7.weight", "module.features.7.bias", "module.features.7.running_mean", "module.features.7.running_var", "module.features.9.weight", "module.features.9.bias", "module.features.10.weight", "module.features.10.bias", "module.features.10.running_mean", "module.features.10.running_var", "module.features.12.weight", "module.features.12.bias", "module.features.13.weight", "module.features.13.bias", "module.features.13.running_mean", "module.features.13.running_var", "module.features.15.weight", "module.features.15.bias", "module.features.16.weight", "module.features.16.bias", "module.featu
Unexpected key(s) in state_dict: "module.features1.0.weight", "module.features1.0.bias", "module.features1.2.weight", "module.features1.2.bias", "module.features1.5.weight", "module.features1.5.bias", "module.features1.7.weight", "module.features1.7.bias", "module.features1.10.weight", "module.features1.10.bias", "module.features1.12.weight", "module.features1.12.bias", "module.features1.14.weight", "module.features1.14.bias", "module.features1.17.weight", "module.features1.17.bias", "module.features1.19.weight", "module.features1.19.bias", "module.features1.21.weight", "module.features1.21.bias", "module.features1.24.weight", "module.features1.24.bias", "module.features1.26.weight", "module.features1.26.bias", "module.features1.28.weight", "module.features1.28.bias", "module.features2.0.weight", "module.features2.0.bias", "module.features2.1.weight", "module.features2.1.bias", "module.features2.1.running_mean", "module.features2.1.running_var", "module.features2.1.num_batches_tracked", "module.features2.3.weight", "module.features2.3.bias", "module.features2.4.weight", "module.features2.4.bias", "module.features2.4.running_mean", "module.features2.4.running_var", "module.features2.4.num_batches_tracked", "module.features2.6.weight", "module.features2.6.bias", "module.features2.7.weight", "module.features2.7.bias", "module.features2.7.running_mean", "module.features2.7.running_var", "module.features2.7.num_batches_tracked", "module.features2.9.weight", "module.features2.9.bias", "module.features2.10.weight", "module.features2.10.bias", "module.features2.10.running_mean", "module.features2.10.running_var", "module.features2.10.num_batches_tracked", "module.features2.12.weight", "module.features2.12.bias", "module.features2.13.weight", "module.features2.13.bias", "module.features2.13.running_mean", "module.features2.13.running_var", "module.features2.13.num_batches_tracked", "module.features2.15.weight", "module.features2.15.bias", "module.features2.16.weight", "module.features2.16.bias", "module.features2.16.running_mean", "module.features2.16.running_var", "module.features2.16.num_batches_tracked", "module.features2.18.weight", "module.features2.18.bias", "module.features2.19.weight", "module.features2.19.bias", "module.features2.19.running_mean", "module.features2.19.running_var", "module.features2.19.num_baar", "module.features2.16.num_batches_tracked", "module.features2.18.weight", "module.features2.18.bias", "module.features2.19.weight", "module.features2.19.bias", "module.features2.19.running_mean", "module.features2.19.running_var", "module.features2.19.num_batches_tracked", "module.features2.21.weight", "module.features2.21.bias", "module.features2.22.weight", "module.features2.22.bias", "module.features2.22.running_mean", "module.features2.22.running_var", "module.features2.22.num_batches_tracked", "module.featurestches_tracked", "module.features2.21.weight", "module.features2.21.bias", "module.features2.22.weight", "module.features2.22.bias", "module.features2.22.running_mean", "module.features2.22.running_var", "module.features2.22.num_batches_tracked", "module.features2.24.weight", "module.features2.24.bias", "module.features2.25.weight", "module.features2.25.bias", "module.features2.25.running_mean", "module.features2.25.running_var", "module.features2.25.num_batches_tracked", "module.fc.weight", "module.fc.bias".

以下是我用来加载模型的代码,请问问题出在哪里,该如何修正呢?

model_path = '/DBCNN-PyTorch-master/fc_models/readability_params_best.pkl'
def evaluate_model(model_path, image_dir, excel_file):
        # 加载模型
    options_ = {
        'base_lr': 5e-5,
        'batch_size': 4,
        'epochs': 1,
        'weight_decay': 5e-4,
        'dataset':'tianma',
        'fc': False,
        'train_index': [],
        'test_index': []
        }
    
    model = DBCNN(scnn_root=model_path, options=options_)
    model = torch.nn.DataParallel(model).cuda()
    model.load_state_dict(torch.load(scnn_root=model_path))
    model.eval()  # 设置为评估模式
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant