Accuracy -- Version Error

[wangtao666666]

why accuracy of the model tensor(0)?

[KulikovML]

I have the same problem in evaluate() function. you can replace
return torch.eq(y_preds,y_test_var).data.sum()/x_test_var.size(0)
to
return torch.eq(y_preds,y_test_var).data.numpy().sum()/x_test_var.size(0)

[s1162276945]

return torch.eq(y_preds,y_test_var).data.numpy().sum()/x_test_var.size(0)

Excuse me, but what is the reason?
I have changed the file as you say, but still show "Accuracy of the model tensor(0)"
maybe it's the problem of epoch?

[s1162276945]

why accuracy of the model tensor(0)?

I have met the same problem with you.

[kaushalshetty]

@s1162276945 , I am not maintaining the repo anymore. The code was tested in pytorch v0.3, looks like much changes are done to pytorch. Try running the code in pytorch v0.3.

[s1162276945]

@s1162276945 , I am not maintaining the repo anymore. The code was tested in pytorch v0.3, looks like much changes are done to pytorch. Try running the code in pytorch v0.3.

OK，thank you advice.

[vhientran]

Please help me to explain the model in Figure 1 (https://arxiv.org/pdf/1703.03130.pdf). How to create a representation vector from a M matrix?

[a389965626]

@s1162276945 , I am not maintaining the repo anymore. The code was tested in pytorch v0.3, looks like much changes are done to pytorch. Try running the code in pytorch v0.3.

OK，thank you advice.

can the new version pytorch work?