In the showcase notebook for segmentation you've released the mmseg configuration files for two datasets: VOC2012 and ADE20k for ViT-Small. In an absence of the training and evaluation code for the segmentation task they provide (almost) all necessary details on augmentations and hyperparameters, which is extremely helpful in reproducing results from your paper.
Could you please release configuration files for all backbones and datasets evaluated in your paper?
Hi!
In the showcase notebook for segmentation you've released the mmseg configuration files for two datasets: VOC2012 and ADE20k for ViT-Small. In an absence of the training and evaluation code for the segmentation task they provide (almost) all necessary details on augmentations and hyperparameters, which is extremely helpful in reproducing results from your paper.
Could you please release configuration files for all backbones and datasets evaluated in your paper?