I couldn't use your pre-training model. about lightcnn

[joannahong]

same here

[jiesonshan]

    if self._use_cuda:
        checkpoint = torch.load(self._model_path)
        restore_dict = checkpoint
    else:
        checkpoint = torch.load(model_path, map_location='cpu')
        restore_dict = {k[7:]: v for k, v in checkpoint['state_dict'].items()}
    self.model.load_state_dict(restore_dict)

[AminTaheri23]

same here @AlfredXiangWu
I downloaded All of the files for models and all of them make an Error.
this is the error I get.

loading checkpoint 'prev9.path.tar' Traceback (most recent call last): File "extract_features.py", line 115, in <module> main() File "extract_features.py", line 67, in main model.load_state_dict(checkpoint['state_dict']) File "/home/amint/virtualenvironment/lightcnn/local/lib/python2.7/site-packages/torch/nn/modules/module.py", line 839, in load_state_dict self.__class__.__name__, "\n\t".join(error_msgs))) RuntimeError: Error(s) in loading state_dict for network_9layers: Missing key(s) in state_dict: "features.0.filter.bias", "features.0.filter.weight", "features.2.conv_a.filter.bias", "features.2.conv_a.filter.weight", "features.2.conv.filter.bias", "features.2.conv.filter.weight", "features.4.conv_a.filter.bias", "features.4.conv_a.filter.weight", "features.4.conv.filter.bias", "features.4.conv.filter.weight", "features.6.conv_a.filter.bias", "features.6.conv_a.filter.weight", "features.6.conv.filter.bias", "features.6.conv.filter.weight", "features.7.conv_a.filter.bias", "features.7.conv_a.filter.weight", "features.7.conv.filter.bias", "features.7.conv.filter.weight", "fc1.filter.bias", "fc1.filter.weight", "fc2.bias", "fc2.weight". Unexpected key(s) in state_dict: "module.features.0.filter.weight", "module.features.0.filter.bias", "module.features.2.conv_a.filter.weight", "module.features.2.conv_a.filter.bias", "module.features.2.conv.filter.weight", "module.features.2.conv.filter.bias", "module.features.4.conv_a.filter.weight", "module.features.4.conv_a.filter.bias", "module.features.4.conv.filter.weight", "module.features.4.conv.filter.bias", "module.features.6.conv_a.filter.weight", "module.features.6.conv_a.filter.bias", "module.features.6.conv.filter.weight", "module.features.6.conv.filter.bias", "module.features.7.conv_a.filter.weight", "module.features.7.conv_a.filter.bias", "module.features.7.conv.filter.weight", "module.features.7.conv.filter.bias", "module.fc1.filter.weight", "module.fc1.filter.bias", "module.fc2.weight", "module.fc2.bias".

[AminTaheri23]

#37

the solution is in this issue.