Attention, block and swiglu_ffn modules always raise warnings about xformers #513
Description
Thank you first of all for this amazing work.
When importing dinov2 using torch.hub.load I always face warnings concerning xformers:
>>> import torch
>>> model = torch.hub.load('facebookresearch/dinov2', "dinov2_vits14", pretrained=False)
Using cache found in /home/rfez/.cache/torch/hub/facebookresearch_dinov2_main
/home/xxxx/.cache/torch/hub/facebookresearch_dinov2_main/dinov2/layers/swiglu_ffn.py:51: UserWarning: xFormers is not available (SwiGLU)
warnings.warn("xFormers is not available (SwiGLU)")
/home/xxxx/.cache/torch/hub/facebookresearch_dinov2_main/dinov2/layers/attention.py:33: UserWarning: xFormers is not available (Attention)
warnings.warn("xFormers is not available (Attention)")
/home/xxxx/.cache/torch/hub/facebookresearch_dinov2_main/dinov2/layers/block.py:40: UserWarning: xFormers is not available (Block)
warnings.warn("xFormers is not available (Block)")I tried to debug this and found that, even if xformers is installed, a warning is issued saying that XFormers is available...!
This is a little bit annoying since there is no elegant way to silence these warnings :/
This is related to #151