adam9500370

Could you share your training settings (i.e., optimizer, learning rate, image size, ... in config file)?

Hi, @victor123499 . Change 'exponent' tensor type to float: ``` scale_weight = torch.pow(scale * torch.ones(n_inp), torch.arange(n_inp).float()) ```

Maybe you can check whether your target labels are in the range [0, n_classes-1].

Hi, @jinfagang . You can set `multi_scale_cross_entropy` loss function in config file. ``` loss: name: 'multi_scale_cross_entropy' ``` And change 'exponent' tensor type to float and set the corresponding device (in...

Could you share your training settings (i.e., # of classes (dataset), optimizer, learning rate, image size, ...)?

Due to `size_average: False` for loss calculation, you may get a very large loss value (summation of cross entropy loss for all the pixels of all the images in each...

Due to high proportion of pixels for `road` class in the Cityscapes dataset, you may need to do class balancing to set higher loss weights for the rare classes. (reference:...

Replace ``` scale_weight = torch.pow(scale * torch.ones(n_inp), torch.arange(n_inp).float()).to('cuda' if input.is_cuda else 'cpu') ``` with ``` scale_weight = torch.pow(scale * torch.ones(n_inp), torch.arange(n_inp).float()).to('cuda' if target.is_cuda else 'cpu') ``` to avoid handling different...

Due to high proportion of pixels for `background` class in the Pascal VOC dataset, if you train the model from scratch, the model might tend to only learn `background` class....

Hi, @sanweiliti . You can do the followings: * Download the converted Caffe pretrained weights [here](https://github.com/meetshah1995/pytorch-semseg/pull/81) * Set `img_norm=False` and `version="pascal"` arguments in data_loader (due to data preprocessing of original...