EVA icon indicating copy to clipboard operation
EVA copied to clipboard

EVA Series: Visual Representation Fantasies from BAAI

Results 85 EVA issues
Sort by recently updated
recently updated
newest added

In using your https://github.com/baaivision/EVA/tree/master/EVA-02/det/docker/Dockerfile compile mirror, found that there are many places of version conflicts. For example, the base image uses python3.6, but when compiling detectron2, it requires python>=3.7; pytorch...

Thank you for your amazing work. I notice that your model architecture is slightly different from ViT-S or DeIT-S. However, the paper doesn't provide the performance of EVA-02-S trained by...

Hi,Sir How to download the init_weights for seg model, I can't open the weights link in China when jumps to hunggingface.co Thank you.

Hi guys, Thanks so much for opening up and sharing your amazing work. I notice that `EVA-02` segmentation weights are shared for `ADE20k`, but not for `COCO-stuff-164k`. But do you...

I have tried running EVA01/EVA02 on Linux and Windows. All the tests go through. But the demo code is dysfuctional. You claim that you adopted the Detectron2 framework. In Detectron2...

Hi, authors. I am struggling with using the EVA-02 detection models for inference. Do you have example code available? I ran the following command ``` python demo/demo.py \ --config-file ./projects/ViTDet/configs/eva2_o365_to_lvis/eva2_o365_to_lvis_cascade_mask_rcnn_vitdet_l_8attn_1536_lrd0p8.py...

Is there any simple code available for running inference of EVA-02 Detection models? Thanks

It is shown that the robust accuracy of the fine-tuned EVA ViT on ObjectNet is obviously lower than that of the EVA CLIP model. Is that because the former is...

I am trying to evaluate the EVA model https://huggingface.co/BAAI/EVA/blob/main/eva_coco_seg.pth in COCO instance segementation. I have only GTX 2080 and it always is out of memory even after I have set...