RuntimeError: mat1 and mat2 shapes cannot be multiplied (8x64 and 128x128)

[LongchaoDa]

Hi, I was running the code and happened with the following error:

ERROR: mat1 and mat2 shapes cannot be multiplied (8x64 and 128x128)

This also comes with command: python3 main.py --name airfoil --episodes 30 --steps 15

I believe it is caused by the initialization of self.eval_net, and might be some dimension issue:

Would you mind re-executing it and see if it has the same error from your end? (This error comes after 6 steps of successful running in the first episode, so you might want to wait for a while after running)

Detailed trackback information:

Traceback (most recent call last):
File "main.py", line 229, in
train(params)
File "main.py", line 165, in train
model_cluster1.learn(optimizer_c1)
File "/data/liliu/Vaccine_Study/pre_analysis/Feature_reconst/Group-wise-Reinforcement-Feature-Generation-for-Optimal-and-Explainable-Representation-Space-Reconst/model.py", line 303, in learn
q_eval = self.eval_net(net_input)
File "/.conda/envs/daily/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/.conda/envs/daily/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "/pre_analysis/Feature_reconst/Group-wise-Reinforcement-Feature-Generation-for-Optimal-and-Explainable-Representation-Space-Reconst/model.py", line 48, in forward
x = self.fc1(x)
File "/.conda/envs/daily/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/.conda/envs/daily/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "/.conda/envs/daily/lib/python3.8/site-packages/torch/nn/modules/linear.py", line 114, in forward
return F.linear(input, self.weight, self.bias)
RuntimeError: mat1 and mat2 shapes cannot be multiplied (8x64 and 128x128)