Thank you for releasing the code for CRTP.
I am trying to reproduce the training pipeline following the README. After setting up the environment and preparing the dataset, I ran:
python train_cls.py --id risk_category_classifier --epochs 10 --batch_size 6 --gpus 1
However, the script failed when initializing the classifier backbone. The error message is:
FileNotFoundError: [Errno 2] No such file or directory: './pdresnet50_model_best.pth'
After checking the code, it seems that classifier.py calls:
self.backbone = pdresnet50(pretrained=self.pretrained)
and pdresnet50.py loads the pretrained weights from:
'pdresnet50': './pdresnet50_model_best.pth'
However, I could not find pdresnet50_model_best.pth in the repository or in the dataset files.
Could you provide the download link for this pretrained model?
My current workaround is to change the model initialization to:
GCN_model(pretrained=False)
Thank you for your help.
Thank you for releasing the code for CRTP.
I am trying to reproduce the training pipeline following the README. After setting up the environment and preparing the dataset, I ran:
python train_cls.py --id risk_category_classifier --epochs 10 --batch_size 6 --gpus 1
However, the script failed when initializing the classifier backbone. The error message is:
FileNotFoundError: [Errno 2] No such file or directory: './pdresnet50_model_best.pth'
After checking the code, it seems that classifier.py calls:
self.backbone = pdresnet50(pretrained=self.pretrained)
and pdresnet50.py loads the pretrained weights from:
'pdresnet50': './pdresnet50_model_best.pth'
However, I could not find pdresnet50_model_best.pth in the repository or in the dataset files.
Could you provide the download link for this pretrained model?
My current workaround is to change the model initialization to:
GCN_model(pretrained=False)
Thank you for your help.