[ICLR 2023 Spotlight] Vision Transformer Adapter for Dense Predictions
Release the model and log for panoptic segmentation.
We win the champion of WSDM Cup 2023 Toloka VQA Challenge using ViT-Adapter.
Pretrained weights and logs of ViT-Adapter.
htc++_beit_adapter_large_fpn_3x_coco
mask2former_beit_adapter_base_512_40k_cocostuff10k
mask2former_beit_adapter_base_480_40k_pascal_context_59
mask2former_beit_adapter_large_896_80k_cocostuff164k
mask2former_beit_adapter_large_512_40k_cocostuff10k
upernet_beit_adapter_large_512_80k_cocostuff10k