ViTPose Model

[gusmcarreira]

I'm trying to use this config file, "https://github.com/open-mmlab/mmpose/blob/main/configs/body_2d_keypoint/topdown_heatmap/coco/td-hm_ViTPose-large_8xb64-210e_coco-256x192.py" but in the backbone I don't get the type being "mmcls.VisionTransformer"? Any help, what should I swap it for?

Kind regards

[LareinaM]

Hi, we use the VisionTransformer implemented by mmcls (mmpretrain). Please follow their installation guide.

[seon-creator]

Hello, I also want to use that model, where is the following the installation guide?
I'm using mmpose but there is mmpretrain.
Just copy code?
Or another installation command?