Training Triplet Model on Market-1501 dataset

[dorukcakmakci]

Hello, I got the following error while training Triplet model on Market-1501 dataset using the command provided in Training section of README.md:

I believe the error is caused by the following:

• https://github.com/CEA-LIST/adv-reid/blob/main/models/architectures/models.py#L42
• https://github.com/CEA-LIST/adv-reid/blob/main/models/architectures/models.py#L40

I think self.backbone.fc.in_features in these lines should be replaced with self.backbone.fc.out_features

[finger-monkey]

Hello, I got the following error while training Triplet model on Market-1501 dataset using the command provided in Training section of README.md:

I believe the error is caused by the following:

• https://github.com/CEA-LIST/adv-reid/blob/main/models/architectures/models.py#L42
• https://github.com/CEA-LIST/adv-reid/blob/main/models/architectures/models.py#L40

I think self.backbone.fc.in_features in these lines should be replaced with self.backbone.fc.out_features

Did you run it successfully? Could you share with me?

[finger-monkey]

I have the same ploblem about "Training Triplet Model on Market-1501 dataset".

Hello, I got the following error while training Triplet model on Market-1501 dataset using the command provided in Training section of README.md:

I believe the error is caused by the following:

• https://github.com/CEA-LIST/adv-reid/blob/main/models/architectures/models.py#L42
• https://github.com/CEA-LIST/adv-reid/blob/main/models/architectures/models.py#L40

I think self.backbone.fc.in_features in these lines should be replaced with self.backbone.fc.out_features

I have the same ploblem about this. And when I rectify the self.backbone.fc.in_features to self.backbone.fc.out_features, it will lead to another question as following:

you can found the accuracy never changed, Could you share me how to solve this problem.
All the best.

[finger-monkey]

could you help me?

[qbouniot]

Hello, I updated the code and corrected the error in the Triplet model. The problem was in the modification of the last layer of the pretrained resnet.
I launched it to verify and everything should work correctly now. Once you have pulled the last version of the code, can you confirm me that it works on your side ?