Code running result

[zwyzhr]

Why is the result I run is far from the paper？


Results in the paper:

[Anand270294]

Hi, I am currently also experencing the same issue with the generated images and also the discriminator loss seems to be approaching inf after 30 epochs.

[zwyzhr]

Hi, I am currently also experencing the same issue with the generated images and also the discriminator loss seems to be approaching inf after 30 epochs.

So your result is the same as mine？

[Anand270294]

Yeah, it is, I cant seem to get the results shown the in paper but maybe it because I might be doing something wrong. So I am trying to recreate the code properly and see if it works. Apparently the numbers cant seem to be generated in my case.

[Samuelstein1224]

Hi, the experiment settings for the notebook are set to be PCA-2 Dim (i.e. 5 qubit total, see variable k) and only one set of labels. To change this, you can append the line :

valid_labels += train_labels == 3

prior to data processing.
Furthermore, increase k to 9, and then PCA to 4 Dimensions, and change the train_labels = train_labels[:10000] section to include all data.

If this does not work please post an update and I will investigate it.

[SamYJ2606]

Hi, can you please clarify what are you referring to which part of the code are you referring to with regards to the PCA dimension? There is no way to alter such a parameter other than the k value according to how the code is written.

[Samuelstein1224]

Within the code, there is a section with PCA. The n_dimensions can be changed, and accordingly with the k value. So if you use 4 PCA dimensions, k should be 9 (i.e. 9 qubits), or 5 depending on which encoding method you are using.

[SamYJ2606]

I have changed the n_dimensions to 6, and the results are still the same. May I know what was your n_dimension when you managed to obtain different results?

[BassantTolba1234]

Dear Prof,
please if you get the same result of the paper, can you kindly share it with me?
I'm waiting for your reply.
Thanks a lot.

[BassantTolba1234]

Dear Prof, @Anand270294 @SamYJ2606 @Samuelstein1224 @zwyzhr
please if you get the same result of the paper, can you kindly share it with me?
I'm waiting for your reply.
Thanks a lot.

[BassantTolba1234]

Dear Prof, @Anand270294 @SamYJ2606 @Samuelstein1224 @zwyzhr Is the proposed code used for training only ?.. please can you kindly explain and provide the code for testing please ?

[BassantTolba1234]

Dear Prof, @Anand270294 @SamYJ2606 @Samuelstein1224 @zwyzhr Is the proposed code used for training only ?.. please can you kindly share the code version for running this code on a real quantum computer not a simulator?