search

pyli0628 / MPG

43 stars 9 forks source link

Question about fintuning model #6

Open ge374 opened 11 months ago

ge374 commented 11 months ago

when I run fintune model, Imet this problem : Traceback (most recent call last): File "/home/yehlab/home/xinxin/DDI_MPG/MPG/DDI/finetune_snap.py", line 244, in main() File "/home/yehlab/home/xinxin/DDI_MPG/MPG/DDI/finetune_snap.py", line 218, in main train(args, model, device, train_loader, optimizer,criterion) File "/home/yehlab/home/xinxin/DDI_MPG/MPG/DDI/finetune_snap.py", line 67, in train pred = model(batch) ^^^^^^^^^^^^ File "/home/yehlab/anaconda3/envs/DDI_MPG_python37/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/anaconda3/envs/DDI_MPG_python37/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl return forward_call(*args, *kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/home/xinxin/DDI_MPG/MPG/DDI/model.py", line 294, in forward node_representation = self.gnn(x, edge_index, edge_attr,node_seg,edge_seg) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/anaconda3/envs/DDI_MPG_python37/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl return self._call_impl(args, kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/anaconda3/envs/DDI_MPG_python37/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl return forward_call(*args, kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/home/xinxin/DDI_MPG/MPG/DDI/model.py", line 226, in forward x = gnn(x,edge_index,edge_attr) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/anaconda3/envs/DDI_MPG_python37/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, *kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/anaconda3/envs/DDI_MPG_python37/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl return forward_call(args, kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/home/xinxin/DDI_MPG/MPG/DDI/model.py", line 178, in forward attention_output = self.attention.forward(x, edge_index, edge_attr) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/home/xinxin/DDI_MPG/MPG/DDI/model.py", line 147, in forward return self.propagate(edge_index, size=size, x=x, pseudo=pseudo) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/anaconda3/envs/DDI_MPG_python37/lib/python3.11/site-packages/torch_geometric/nn/conv/message_passing.py", line 463, in propagate out = self.message(**msg_kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/home/xinxin/DDI_MPG/MPG/DDI/model.py", line 158, in message alpha = softmax(alpha, edge_index_i, size_i) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/yehlab/anaconda3/envs/DDI_MPG_python37/lib/python3.11/site-packages/torch_geometric/utils/softmax.py", line 56, in softmax count = ptr[1:] - ptr[:-1]


TypeError: 'int' object is not subscriptable
pyli0628 commented 5 months ago

_softmax error is usually due to a version issue with pytorch geometry. It is possible that you have used a higher version of pyg than the required environment. Please check.