The checkpoint on iNat2018

[Duconnor]

Thank you guys for your impressive work and releasing the code. I just wonder when will the checkpoint on iNat2018 be released? I'm looking forward to that. Thanks a lot!

[jiequancui]

Hi, thanks for your interest in our work.

We have already released the log file for checking.
If you met problems with reproduction, welcome to discuss them with me.

We are being dedicated to an improved version of PaCo. It will take some time to re-train it and release the models.

[Duconnor]

Thanks for your quick reply, I've successfully reproduce the result. But I do think there is one tiny bug here

Parametric-Contrastive-Learning/LT/paco_lt.py

Line 530 in 4609959

total_logits = torch.empty((0, 1000)).cuda()

which I think should be changed to
total_logits = torch.empty((0, args.num_classes)).cuda()

[jiequancui]

Thanks for your suggestion.

I will fix the issue.