`with torch.no_grad()` in your `train_fusionnet_axial.py` ?

[shinianzhihou]

In your training script, you use with torch.no_grad() in line 56 to build your nest_model.

It means only fusion_model are trained during the training process?

If I want to train on my own dataset, should I remove with torch.no_grad() ?

[Vibashan]

Yes, only the fusion model is trained during the training process.

If I want to train on my own dataset, should I remove with torch.no_grad() ? Yes.