wutong16 / DistributionBalancedLoss

[ ECCV 2020 Spotlight ] Pytorch implementation for "Distribution-Balanced Loss for Multi-Label Classification in Long-Tailed Datasets"

wutong16/DistributionBalancedLoss Issues

RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.FloatTensor [64, 109]], which is output 0 of SigmoidBackward0, is at version 1; expected version 0 instead. Hint: the backtrace further above shows the operation that failed to compute its gradient. The variable in question was changed in there or anywhere later. Good luck!
Closed 2 months ago
How to use the ResampleLoss？
Closed 4 months ago
About code implementation of eq7
Updated 2 years ago
About Class aware sampler
Closed 2 years ago2
FileNotFoundError: [Errno 2] No such file or directory: '/mnt/SSD/det/coco/annotations/instances_train2017.json'
Updated 2 years ago
Reproduction of Distribution-Balanced Loss on VOC Dataset
Closed 2 years ago2
ModuleNotFoundError: No module named 'lvis'
Closed 4 years ago2
how to generate class_freq.pkl file for my own dataset
Updated 2 years ago
bug: args.gpus is missing
Updated 3 years ago
大佬，这个方法可以用在one stage模型中吗（yolov5）
Closed 3 years ago2
How to reproduce the baseline mAP on VOC dataset
Closed 3 years ago
How to use in my own dataset?
Updated 3 years ago4
Lower mAP when training on VOC
Closed 3 years ago2
About the formula(9) is inconstense with the code implementation.
Closed 3 years ago1
--
Closed 3 years ago
Dataset?
Closed 3 years ago1
Support multi-GPU training?
Closed 4 years ago1
Question regarding the model
Closed 4 years ago1
problem of train set and experiment
Closed 4 years ago4
formulation error in Eq 7
Closed 4 years ago1