search

uw-ipd / RoseTTAFold2

MIT License
160 stars 36 forks source link

RuntimeError: mat1 and mat2 must have the same dtype, but got Half and Float #38

Closed Khas-Erdene-1 closed 1 month ago

Khas-Erdene-1 commented 2 months ago

Hi,

I'm having following problem. I run with the example files, it was fine but when I input following fasta, it gives the error. Any suggestions?

/data2/apps/RoseTTAFold2/run_RF2.sh /home/khasaa/projects/RoseTTAfold_ADAMTS13/Q76LX8.fasta -o Q76LX8 Running RoseTTAFold2 to predict structures -> Running command: python /data2/apps/RoseTTAFold2/network/predict.py -inputs Q76LX8/Q76LX8_1.msa0.a3m -prefix Q76LX8/models/model -model /data2/apps/RoseTTAFold2/network/weights/RF2_jan24.pt -db /data2/apps/RoseTTAFold2/pdb100_2021Mar03/pdb100_2021Mar03 -symm C1 Running on GPU N=2048 L=1427 recycle 0 plddt 0.308 pae 28.391 rmsd 34.750 Traceback (most recent call last): File "/data2/apps/RoseTTAFold2/network/predict.py", line 623, in pred.predict( File "/data2/apps/RoseTTAFold2/network/predict.py", line 425, in predict self.run_prediction( File "/data2/apps/RoseTTAFold2/network/predict.py", line 510, in run_prediction logits, , _, logits_pae, p_bind, xyz_prev, alpha, symmsub, pred_lddt, msa_prev, pair_prev, state_prev = self.model( File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(*args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(*args, *kwargs) File "/data2/apps/RoseTTAFold2/network/RoseTTAFoldModel.py", line 115, in forward msa, pair, R, T, alpha, state, symmsub = self.simulator( File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(*args, kwargs) File "/data2/apps/RoseTTAFold2/network/Track_module.py", line 792, in forward msa_full, pair, R_in, T_in, state, alpha, symmsub = self.extra_block[i_m]( File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(*args, *kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(args, kwargs) File "/data2/apps/RoseTTAFold2/network/Track_module.py", line 690, in forward R, T, state, alpha = self.str2str( File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(*args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(*args, *kwargs) File "/data2/apps/RoseTTAFold2/network/Track_module.py", line 600, in forward shift = self.se3(G, node, l1_feats, edge_feats) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(*args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/amp/autocast_mode.py", line 16, in decorate_autocast return func(*args, *kwargs) File "/data2/apps/RoseTTAFold2/network/SE3_network.py", line 86, in forward return self.se3(G, node_features, edge_features) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(*args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/se3_transformer-1.2.0-py3.10.egg/se3_transformer/model/transformer.py", line 187, in forward File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(*args, *kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/se3_transformer-1.2.0-py3.10.egg/se3_transformer/model/transformer.py", line 47, in forward File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(*args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(*args, *kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/se3_transformer-1.2.0-py3.10.egg/se3_transformer/model/layers/attention.py", line 158, in forward File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(*args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/se3_transformer-1.2.0-py3.10.egg/se3_transformer/model/layers/convolution.py", line 358, in forward File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/se3_transformer-1.2.0-py3.10.egg/se3_transformer/model/layers/convolution.py", line 245, in File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(*args, *kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/se3_transformer-1.2.0-py3.10.egg/se3_transformer/model/layers/convolution.py", line 181, in forward File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(*args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(*args, *kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/se3_transformer-1.2.0-py3.10.egg/se3_transformer/model/layers/convolution.py", line 119, in forward File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl return self._call_impl(args, kwargs) File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl return forward_call(*args, **kwargs) RuntimeError: The following operation failed in the TorchScript interpreter. Traceback of TorchScript (most recent call last): File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/container.py", line 217, in forward def forward(self, input): for module in self: input = module(input)


        return input
  File "/home/khasaa/miniconda3/envs/RF2/lib/python3.10/site-packages/torch/nn/modules/linear.py", line 116, in forward
    def forward(self, input: Tensor) -> Tensor:
        return F.linear(input, self.weight, self.bias)
               ~~~~~~~~ <--- HERE
RuntimeError: mat1 and mat2 must have the same dtype, but got Half and Float
jaideepm009 commented 1 month ago

having the same issues for some pdb files

Khas-Erdene-1 commented 1 month ago

Turned out our machine graphics memory is not enough for bigger proteins like bigger than 1200 amino acids, so I'm closing this issue. Also RF2 doesn't allow multi graphics unit to predict.