I really appreciate your work and am grateful for your open-sourcing of the code.
I was able to successfully reproduce the experimental results on CIFAR100-LT using your code, but I was wondering if you could provide me with some additional information on the hyperparameter settings for the larger datasets, ImageNet-LT and iNaturalist.
Specifically, I would greatly appreciate it if you could share the following:
For Stage 1:
the initial learning rate, epoch number, weight decay settings,
and any data augmentation techniques used (such as color jittering for ImageNet).
For Stage 2:
the initial learning rate, epoch number, weight decay settings,
and the hyperparameters in CBLoss (loss type, beta, and gamma).
Thank you once again for your fantastic work and for your generosity in sharing your code with the community.
I really appreciate your work and am grateful for your open-sourcing of the code.
I was able to successfully reproduce the experimental results on CIFAR100-LT using your code, but I was wondering if you could provide me with some additional information on the hyperparameter settings for the larger datasets, ImageNet-LT and iNaturalist.
Specifically, I would greatly appreciate it if you could share the following:
For Stage 1: the initial learning rate, epoch number, weight decay settings, and any data augmentation techniques used (such as color jittering for ImageNet).
For Stage 2: the initial learning rate, epoch number, weight decay settings, and the hyperparameters in CBLoss (loss type, beta, and gamma).
Thank you once again for your fantastic work and for your generosity in sharing your code with the community.