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

trained model are not working #13

Open
margokhokhlova opened this issue Mar 16, 2022 · 3 comments
Open

trained model are not working #13

margokhokhlova opened this issue Mar 16, 2022 · 3 comments

Comments

@margokhokhlova
Copy link

Please, can you advice what I might be doing wrong?
I am trying to run the eval script, I changed the path to the saved weigths but I am getting this:

RuntimeError: Error(s) in loading state_dict for EncoderNet:
Missing key(s) in state_dict: "conv.weight", "conv.bias", "bn.weight", "bn.bias", "bn.running_mean", "bn.running_var", "en_layer1.0.conv1.weight", "en_layer1.0.conv1.bias", "en_layer1.0.bn1.weight", "en_layer1.0.bn1.bias", "en_layer1.0.bn1.running_mean", "en_layer1.0.bn1.running_var", "en_layer1.0.conv2.weight", "en_layer1.0.conv2.bias", "en_layer1.0.bn2.weight", "en_layer1.0.bn2.bias", "en_layer1.0.bn2.running_mean", "en_layer1.0.bn2.running_var", "en_layer2.0.conv1.weight", "en_layer2.0.conv1.bias", "en_layer2.0.bn1.weight", "en_layer2.0.bn1.bias", "en_layer2.0.bn1.running_mean", "en_layer2.0.bn1.running_var", "en_layer2.0.conv2.weight", "en_layer2.0.conv2.bias", "en_layer2.0.bn2.weight", "en_layer2.0.bn2.bias", "en_layer2.0.bn2.running_mean", "en_layer2.0.bn2.running_var", "en_layer2.0.downsample.0.weight", "en_layer2.0.downsample.0.bias", "en_layer2.0.downsample.1.weight", "en_layer2.0.downsample.1.bias", "en_layer2.0.downsample.1.running_mean", "en_layer2.0.downsample.1.running_var", "en_layer3.0.conv1.weight", "en_layer3.0.conv1.bias", "en_layer3.0.bn1.weight", "en_layer3.0.bn1.bias", "en_layer3.0.bn1.running_mean", "en_layer3.0.bn1.running_var", "en_layer3.0.conv2.weight", "en_layer3.0.conv2.bias", "en_layer3.0.bn2.weight", "en_layer3.0.bn2.bias", "en_layer3.0.bn2.running_mean", "en_layer3.0.bn2.running_var", "en_layer3.0.downsample.0.weight", "en_layer3.0.downsample.0.bias", "en_layer3.0.downsample.1.weight", "en_layer3.0.downsample.1.bias", "en_layer3.0.downsample.1.running_mean", "en_layer3.0.downsample.1.running_var", "en_layer4.0.conv1.weight", "en_layer4.0.conv1.bias", "en_layer4.0.bn1.weight", "en_layer4.0.bn1.bias", "en_layer4.0.bn1.running_mean", "en_layer4.0.bn1.running_var", "en_layer4.0.conv2.weight", "en_layer4.0.conv2.bias", "en_layer4.0.bn2.weight", "en_layer4.0.bn2.bias", "en_layer4.0.bn2.running_mean", "en_layer4.0.bn2.running_var", "en_layer4.0.downsample.0.weight", "en_layer4.0.downsample.0.bias", "en_layer4.0.downsample.1.weight", "en_layer4.0.downsample.1.bias", "en_layer4.0.downsample.1.running_mean", "en_layer4.0.downsample.1.running_var", "en_layer5.0.conv1.weight", "en_layer5.0.conv1.bias", "en_layer5.0.bn1.weight", "en_layer5.0.bn1.bias", "en_layer5.0.bn1.running_mean", "en_layer5.0.bn1.running_var", "en_layer5.0.conv2.weight", "en_layer5.0.conv2.bias", "en_layer5.0.bn2.weight", "en_layer5.0.bn2.bias", "en_layer5.0.bn2.running_mean", "en_layer5.0.bn2.running_var", "en_layer5.0.downsample.0.weight", "en_layer5.0.downsample.0.bias", "en_layer5.0.downsample.1.weight", "en_layer5.0.downsample.1.bias", "en_layer5.0.downsample.1.running_mean", "en_layer5.0.downsample.1.running_var", "en_layer5.1.conv1.weight", "en_layer5.1.conv1.bias", "en_layer5.1.bn1.weight", "en_layer5.1.bn1.bias", "en_layer5.1.bn1.running_mean", "en_layer5.1.bn1.running_var", "en_layer5.1.conv2.weight", "en_layer5.1.conv2.bias", "en_layer5.1.bn2.weight", "en_layer5.1.bn2.bias", "en_layer5.1.bn2.running_mean", "en_layer5.1.bn2.running_var".
Unexpected key(s) in state_dict: "module.conv.weight", "module.conv.bias", "module.bn.weight", "module.bn.bias", "module.bn.running_mean", "module.bn.running_var", "module.en_layer1.0.conv1.weight", "module.en_layer1.0.conv1.bias", "module.en_layer1.0.bn1.weight", "module.en_layer1.0.bn1.bias", "module.en_layer1.0.bn1.running_mean", "module.en_layer1.0.bn1.running_var", "module.en_layer1.0.conv2.weight", "module.en_layer1.0.conv2.bias", "module.en_layer1.0.bn2.weight", "module.en_layer1.0.bn2.bias", "module.en_layer1.0.bn2.running_mean", "module.en_layer1.0.bn2.running_var", "module.en_layer2.0.conv1.weight", "module.en_layer2.0.conv1.bias", "module.en_layer2.0.bn1.weight", "module.en_layer2.0.bn1.bias", "module.en_layer2.0.bn1.running_mean", "module.en_layer2.0.bn1.running_var", "module.en_layer2.0.conv2.weight", "module.en_layer2.0.conv2.bias", "module.en_layer2.0.bn2.weight", "module.en_layer2.0.bn2.bias", "module.en_layer2.0.bn2.running_mean", "module.en_layer2.0.bn2.running_var", "module.en_layer2.0.downsample.0.weight", "module.en_layer2.0.downsample.0.bias", "module.en_layer2.0.downsample.1.weight", "module.en_layer2.0.downsample.1.bias", "module.en_layer2.0.downsample.1.running_mean", "module.en_layer2.0.downsample.1.running_var", "module.en_layer3.0.conv1.weight", "module.en_layer3.0.conv1.bias", "module.en_layer3.0.bn1.weight", "module.en_layer3.0.bn1.bias", "module.en_layer3.0.bn1.running_mean", "module.en_layer3.0.bn1.running_var", "module.en_layer3.0.conv2.weight", "module.en_layer3.0.conv2.bias", "module.en_layer3.0.bn2.weight", "module.en_layer3.0.bn2.bias", "module.en_layer3.0.bn2.running_mean", "module.en_layer3.0.bn2.running_var", "module.en_layer3.0.downsample.0.weight", "module.en_layer3.0.downsample.0.bias", "module.en_layer3.0.downsample.1.weight", "module.en_layer3.0.downsample.1.bias", "module.en_layer3.0.downsample.1.running_mean", "module.en_layer3.0.downsample.1.running_var", "module.en_layer4.0.conv1.weight", "module.en_layer4.0.conv1.bias", "module.en_layer4.0.bn1.weight", "module.en_layer4.0.bn1.bias", "module.en_layer4.0.bn1.running_mean", "module.en_layer4.0.bn1.running_var", "module.en_layer4.0.conv2.weight", "module.en_layer4.0.conv2.bias", "module.en_layer4.0.bn2.weight", "module.en_layer4.0.bn2.bias", "module.en_layer4.0.bn2.running_mean", "module.en_layer4.0.bn2.running_var", "module.en_layer4.0.downsample.0.weight", "module.en_layer4.0.downsample.0.bias", "module.en_layer4.0.downsample.1.weight", "module.en_layer4.0.downsample.1.bias", "module.en_layer4.0.downsample.1.running_mean", "module.en_layer4.0.downsample.1.running_var", "module.en_layer5.0.conv1.weight", "module.en_layer5.0.conv1.bias", "module.en_layer5.0.bn1.weight", "module.en_layer5.0.bn1.bias", "module.en_layer5.0.bn1.running_mean", "module.en_layer5.0.bn1.running_var", "module.en_layer5.0.conv2.weight", "module.en_layer5.0.conv2.bias", "module.en_layer5.0.bn2.weight", "module.en_layer5.0.bn2.bias", "module.en_layer5.0.bn2.running_mean", "module.en_layer5.0.bn2.running_var", "module.en_layer5.0.downsample.0.weight", "module.en_layer5.0.downsample.0.bias", "module.en_layer5.0.downsample.1.weight", "module.en_layer5.0.downsample.1.bias", "module.en_layer5.0.downsample.1.running_mean", "module.en_layer5.0.downsample.1.running_var", "module.en_layer5.1.conv1.weight", "module.en_layer5.1.conv1.bias", "module.en_layer5.1.bn1.weight", "module.en_layer5.1.bn1.bias", "module.en_layer5.1.bn1.running_mean", "module.en_layer5.1.bn1.running_var", "module.en_layer5.1.conv2.weight", "module.en_layer5.1.conv2.bias", "module.en_layer5.1.bn2.weight", "module.en_layer5.1.bn2.bias", "module.en_layer5.1.bn2.running_mean", "module.en_layer5.1.bn2.running_var".

@KOuldamer
Copy link

Please, can you advice what I might be doing wrong? I am trying to run the eval script, I changed the path to the saved weigths but I am getting this:

RuntimeError: Error(s) in loading state_dict for EncoderNet: Missing key(s) in state_dict: "conv.weight", "conv.bias", "bn.weight", "bn.bias", "bn.running_mean", "bn.running_var", "en_layer1.0.conv1.weight", "en_layer1.0.conv1.bias", "en_layer1.0.bn1.weight", "en_layer1.0.bn1.bias", "en_layer1.0.bn1.running_mean", "en_layer1.0.bn1.running_var", "en_layer1.0.conv2.weight", "en_layer1.0.conv2.bias", "en_layer1.0.bn2.weight", "en_layer1.0.bn2.bias", "en_layer1.0.bn2.running_mean", "en_layer1.0.bn2.running_var", "en_layer2.0.conv1.weight", "en_layer2.0.conv1.bias", "en_layer2.0.bn1.weight", "en_layer2.0.bn1.bias", "en_layer2.0.bn1.running_mean", "en_layer2.0.bn1.running_var", "en_layer2.0.conv2.weight", "en_layer2.0.conv2.bias", "en_layer2.0.bn2.weight", "en_layer2.0.bn2.bias", "en_layer2.0.bn2.running_mean", "en_layer2.0.bn2.running_var", "en_layer2.0.downsample.0.weight", "en_layer2.0.downsample.0.bias", "en_layer2.0.downsample.1.weight", "en_layer2.0.downsample.1.bias", "en_layer2.0.downsample.1.running_mean", "en_layer2.0.downsample.1.running_var", "en_layer3.0.conv1.weight", "en_layer3.0.conv1.bias", "en_layer3.0.bn1.weight", "en_layer3.0.bn1.bias", "en_layer3.0.bn1.running_mean", "en_layer3.0.bn1.running_var", "en_layer3.0.conv2.weight", "en_layer3.0.conv2.bias", "en_layer3.0.bn2.weight", "en_layer3.0.bn2.bias", "en_layer3.0.bn2.running_mean", "en_layer3.0.bn2.running_var", "en_layer3.0.downsample.0.weight", "en_layer3.0.downsample.0.bias", "en_layer3.0.downsample.1.weight", "en_layer3.0.downsample.1.bias", "en_layer3.0.downsample.1.running_mean", "en_layer3.0.downsample.1.running_var", "en_layer4.0.conv1.weight", "en_layer4.0.conv1.bias", "en_layer4.0.bn1.weight", "en_layer4.0.bn1.bias", "en_layer4.0.bn1.running_mean", "en_layer4.0.bn1.running_var", "en_layer4.0.conv2.weight", "en_layer4.0.conv2.bias", "en_layer4.0.bn2.weight", "en_layer4.0.bn2.bias", "en_layer4.0.bn2.running_mean", "en_layer4.0.bn2.running_var", "en_layer4.0.downsample.0.weight", "en_layer4.0.downsample.0.bias", "en_layer4.0.downsample.1.weight", "en_layer4.0.downsample.1.bias", "en_layer4.0.downsample.1.running_mean", "en_layer4.0.downsample.1.running_var", "en_layer5.0.conv1.weight", "en_layer5.0.conv1.bias", "en_layer5.0.bn1.weight", "en_layer5.0.bn1.bias", "en_layer5.0.bn1.running_mean", "en_layer5.0.bn1.running_var", "en_layer5.0.conv2.weight", "en_layer5.0.conv2.bias", "en_layer5.0.bn2.weight", "en_layer5.0.bn2.bias", "en_layer5.0.bn2.running_mean", "en_layer5.0.bn2.running_var", "en_layer5.0.downsample.0.weight", "en_layer5.0.downsample.0.bias", "en_layer5.0.downsample.1.weight", "en_layer5.0.downsample.1.bias", "en_layer5.0.downsample.1.running_mean", "en_layer5.0.downsample.1.running_var", "en_layer5.1.conv1.weight", "en_layer5.1.conv1.bias", "en_layer5.1.bn1.weight", "en_layer5.1.bn1.bias", "en_layer5.1.bn1.running_mean", "en_layer5.1.bn1.running_var", "en_layer5.1.conv2.weight", "en_layer5.1.conv2.bias", "en_layer5.1.bn2.weight", "en_layer5.1.bn2.bias", "en_layer5.1.bn2.running_mean", "en_layer5.1.bn2.running_var". Unexpected key(s) in state_dict: "module.conv.weight", "module.conv.bias", "module.bn.weight", "module.bn.bias", "module.bn.running_mean", "module.bn.running_var", "module.en_layer1.0.conv1.weight", "module.en_layer1.0.conv1.bias", "module.en_layer1.0.bn1.weight", "module.en_layer1.0.bn1.bias", "module.en_layer1.0.bn1.running_mean", "module.en_layer1.0.bn1.running_var", "module.en_layer1.0.conv2.weight", "module.en_layer1.0.conv2.bias", "module.en_layer1.0.bn2.weight", "module.en_layer1.0.bn2.bias", "module.en_layer1.0.bn2.running_mean", "module.en_layer1.0.bn2.running_var", "module.en_layer2.0.conv1.weight", "module.en_layer2.0.conv1.bias", "module.en_layer2.0.bn1.weight", "module.en_layer2.0.bn1.bias", "module.en_layer2.0.bn1.running_mean", "module.en_layer2.0.bn1.running_var", "module.en_layer2.0.conv2.weight", "module.en_layer2.0.conv2.bias", "module.en_layer2.0.bn2.weight", "module.en_layer2.0.bn2.bias", "module.en_layer2.0.bn2.running_mean", "module.en_layer2.0.bn2.running_var", "module.en_layer2.0.downsample.0.weight", "module.en_layer2.0.downsample.0.bias", "module.en_layer2.0.downsample.1.weight", "module.en_layer2.0.downsample.1.bias", "module.en_layer2.0.downsample.1.running_mean", "module.en_layer2.0.downsample.1.running_var", "module.en_layer3.0.conv1.weight", "module.en_layer3.0.conv1.bias", "module.en_layer3.0.bn1.weight", "module.en_layer3.0.bn1.bias", "module.en_layer3.0.bn1.running_mean", "module.en_layer3.0.bn1.running_var", "module.en_layer3.0.conv2.weight", "module.en_layer3.0.conv2.bias", "module.en_layer3.0.bn2.weight", "module.en_layer3.0.bn2.bias", "module.en_layer3.0.bn2.running_mean", "module.en_layer3.0.bn2.running_var", "module.en_layer3.0.downsample.0.weight", "module.en_layer3.0.downsample.0.bias", "module.en_layer3.0.downsample.1.weight", "module.en_layer3.0.downsample.1.bias", "module.en_layer3.0.downsample.1.running_mean", "module.en_layer3.0.downsample.1.running_var", "module.en_layer4.0.conv1.weight", "module.en_layer4.0.conv1.bias", "module.en_layer4.0.bn1.weight", "module.en_layer4.0.bn1.bias", "module.en_layer4.0.bn1.running_mean", "module.en_layer4.0.bn1.running_var", "module.en_layer4.0.conv2.weight", "module.en_layer4.0.conv2.bias", "module.en_layer4.0.bn2.weight", "module.en_layer4.0.bn2.bias", "module.en_layer4.0.bn2.running_mean", "module.en_layer4.0.bn2.running_var", "module.en_layer4.0.downsample.0.weight", "module.en_layer4.0.downsample.0.bias", "module.en_layer4.0.downsample.1.weight", "module.en_layer4.0.downsample.1.bias", "module.en_layer4.0.downsample.1.running_mean", "module.en_layer4.0.downsample.1.running_var", "module.en_layer5.0.conv1.weight", "module.en_layer5.0.conv1.bias", "module.en_layer5.0.bn1.weight", "module.en_layer5.0.bn1.bias", "module.en_layer5.0.bn1.running_mean", "module.en_layer5.0.bn1.running_var", "module.en_layer5.0.conv2.weight", "module.en_layer5.0.conv2.bias", "module.en_layer5.0.bn2.weight", "module.en_layer5.0.bn2.bias", "module.en_layer5.0.bn2.running_mean", "module.en_layer5.0.bn2.running_var", "module.en_layer5.0.downsample.0.weight", "module.en_layer5.0.downsample.0.bias", "module.en_layer5.0.downsample.1.weight", "module.en_layer5.0.downsample.1.bias", "module.en_layer5.0.downsample.1.running_mean", "module.en_layer5.0.downsample.1.running_var", "module.en_layer5.1.conv1.weight", "module.en_layer5.1.conv1.bias", "module.en_layer5.1.bn1.weight", "module.en_layer5.1.bn1.bias", "module.en_layer5.1.bn1.running_mean", "module.en_layer5.1.bn1.running_var", "module.en_layer5.1.conv2.weight", "module.en_layer5.1.conv2.bias", "module.en_layer5.1.bn2.weight", "module.en_layer5.1.bn2.bias", "module.en_layer5.1.bn2.running_mean", "module.en_layer5.1.bn2.running_var".

I have the same error

@I-am-vishalmaurya
Copy link

Add strict=False attribute to load_state_dict().

model_en.load_state_dict(torch.load('model_en.pkl'), strict=False)
model_de.load_state_dict(torch.load('model_de.pkl'), strict=False)
model_class.load_state_dict(torch.load('model_class.pkl'), strict=False)

@prateekchhikara
Copy link

You can replace the following code:

model_en.load_state_dict(torch.load('model_en.pkl'))
model_de.load_state_dict(torch.load('model_de.pkl'))
model_class.load_state_dict(torch.load('model_class.pkl'))

with:

def remove_module_prefix(state_dict):
    """
    Removes the 'module.' prefix from each key in the state dictionary.
    This is useful when loading a DataParallel model state into a non-DataParallel model.
    """
    new_state_dict = {k.replace('module.', ''): v for k, v in state_dict.items()}
    return new_state_dict

# Load your state dict with the 'module.' prefix
original_state_dict = torch.load('model_en.pkl')
updated_state_dict = remove_module_prefix(original_state_dict)
model_en.load_state_dict(updated_state_dict)

original_state_dict = torch.load('model_de.pkl')
updated_state_dict = remove_module_prefix(original_state_dict)
model_de.load_state_dict(updated_state_dict)

original_state_dict = torch.load('model_class.pkl')
updated_state_dict = remove_module_prefix(original_state_dict)
model_class.load_state_dict(updated_state_dict)

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

4 participants