Plot predictions throwing error for the CPU and CUDA type

[Heyymant]

I am getting the same error even after changing the device settings, can someone help me to rectify this error, (yea, I have tried restarting the notebook as well)

[sametucuncu]

I think you do not need to put .cpu() becuase they are not torch tensors. But if you share the whole code, I might help better.

[shisirkha]

Yeah, if he shares the entire we can help him.

[shisirkha]

Got ya first convert the list of numpy array into a torch.Tensor type like this then try.
example:

train_data = torch.from_numpy(train_data).cpu()

or do this:

train_data= torch.from_numpy(X_train).cpu()
train_label = torch.from_numpy(y_train).cpu() 
test_data = torch.from_numpy(X_test).cpu()
test_label = torch.from_numpy(y_train).cpu()

#then do this

plot_predictions(train_data,train_label,test_data,test_label)

Or if you want your variable name to be same do this:

X_train= torch.from_numpy(X_train).cpu()
y_train = torch.from_numpy(y_train).cpu() 
X_test = torch.from_numpy(X_test).cpu()
y_test = torch.from_numpy(y_train).cpu()

#then do this

plot_predictions(X_train,y_train,X_test,y_test)

use this example to define all your numpy array to a tensor.
I hope this helps
Good luck self learner