Sovit Ranjan Rath

@My12123 Maybe you can make `save_memory = True` in `config.py` if you have not already. That will make at least the canny model to run on 8 GB GPU. I...

Yes. But it also works if we use the latest version of PyTorch, along with the latest version of PyTorch lightning.

@engrmusawarali Hi, even with the save memory option you need at least 18 GB of VRAM to train the model. That's what I found out when it was initially released....

@engrmusawarali I have not tried it yet. But planning to do it soon.

Hello. I will try to do so. I have paused the updates for this project for a few months now because of some other priority projects. I hope I can...

So, all the Torchvision detection models expect one extra background class along with the object classes. The class name is not used in the training but instead used to easily...

May I know which version of PyTorch you are using?

Can you please paste the command that you are using?

Oh. Interesting. I need some time to debug in this case.

Hello. Apologies for the delayed response. Recently I have not been able to put much time into the project due to time related constraints. However, I plan to resume again...