zhrli

> Could you please check if the loss.item() in `python3 train_sup.py --arch resnet18 --data cifar10 --fd 128 --epo 50 0 --bs 1000 --eps 0.5 --gam1 1 --gam2 1 --lr 0.01...

and two models are different,one from huggingface you give as backbone another from graphcast_sequential.py name1 Out [11]: ['grid_feat_embedding.0.weight', 'grid_feat_embedding.0.bias', 'mesh_feat_embedding.0.weight', 'mesh_feat_embedding.0.bias', 'mesh_edge_feat_embedding.0.weight', 'mesh_edge_feat_embedding.0.bias', 'grid2mesh_edge_feat_embedding.0.weight', 'grid2mesh_edge_feat_embedding.0.bias', 'mesh2grid_edge_feat_embedding.0.weight', 'mesh2grid_edge_feat_embedding.0.bias', 'grid2mesh_edge_update.0.weight', 'grid2mesh_edge_update.0.bias', 'grid2mesh_edge_update.2.weight',...