Zeeshan Khan Suri

I think you have to create a package by ``` cd catkin_ws/src catkin_create_pkg rosdemo_mp3 ``` Then, you can clone this code inside this "rosdemo_mp3" directory.

Using `align_corners=True` in [Upsample](https://pytorch.org/docs/stable/generated/torch.nn.Upsample.html) solved this for me

But since the grid is being scaled, are the X,Y coordinates of the output (after depth scaling) are the mapping functions of how the image grid changes as a result...

Hi @patricklabatut, I am trying to train on a rectangular image and I get the following assertion error even after the merged PR. https://github.com/facebookresearch/dinov2/blob/e1277af2ba9496fbadf7aec6eba56e8d882d1e35/dinov2/models/vision_transformer.py#L192 ```bash Traceback (most recent call last):...

This solved it for me https://github.com/facebookresearch/dinov2/issues/316#issuecomment-1816378994

Hi, here's an unofficial training and testing pipeline of segmentation using mmsegmentation, based on the configs provided by the authors. The example implements Cityscapes but can be adopted to other...

Thank you @usryokousha for the clarification. Could you please point where the interpolation in the position embedding is happening? I cannot use the 224 model with 518 input or vise-versa....

`self.proj` was at different dtype as the input. I added `self.proj.to(x)` at this line and this resolved https://github.com/facebookresearch/dinov2/blob/da4b3825f0ed64b7398ace00c5062503811d0cff/dinov2/layers/patch_embed.py#L75 but I now get another RuntimeError ````bash Expected output.scalar_type() == at::ScalarType::Half to...

@qasfb, I had to manually cast particular layers to same dtype as the input in multiple places

I think this Error is just a surrogate. It means, internally, there was some error. To figure out what exactly caused the underlying error, you need to check the entire...