Issue running explainer

[farahFif]

Hello,
I tried the to run experiment_replication.ipynb, it seems that the stored models and the GNN model are different (check below the error). I also tried to retrain and replace the best_model in the other folder but the results don't match what's in the papers.
Can you please help with this ?

Traceback (most recent call last):
File "explain.py", line 19, in
(auc, auc_std), inf_time = replication(config.args.explainer, extension)
File "/storage/RE-ParameterizedExplainerForGraphNeuralNetworks/ExplanationEvaluation/tasks/replication.py", line 158, in replication
return_checkpoint=True)
File "/storage/RE-ParameterizedExplainerForGraphNeuralNetworks/ExplanationEvaluation/models/model_selector.py", line 75, in model_selector
model.load_state_dict(checkpoint['model_state_dict'])
File "/home/ubuntu/miniconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1407, in load_state_dict
self.class.name, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for NodeGCN:
Missing key(s) in state_dict: "conv1.lin.weight", "conv2.lin.weight", "conv3.lin.weight".
Unexpected key(s) in state_dict: "conv1.weight", "conv2.weight", "conv3.weight".

[farahFif]

Thanks to @seakkass solution, model_selector needs to be updated when using pyg 2.0.4 to be able to use pre-trained models.

renamed_state_dict = {}
for key in checkpoint['model_state_dict']:
    if key.startswith('conv') and key.endswith('weight'):
        new_key = key[:5] + '.lin' + key[5:]
        renamed_state_dict[new_key] = (checkpoint['model_state_dict'][key]).T
    else:
        renamed_state_dict[key] = checkpoint['model_state_dict'][key]
model.load_state_dict(renamed_state_dict)

#2 (comment)