Random seed not assigned: Trying to replicate the CUB results

[samarth-b]

Thanks for releasing the code.

We are trying to replicate your results on CUB dataset for comparison purposes. The write_CUB_filelist.py doesnt seem to have assigned a seed. So the filelist would change every run.

https://github.com/wyharveychen/CloserLookFewShot/blob/4ca19c75147b49a50fee7b2277971a2299c3d231/filelists/CUB/write_CUB_filelist.py#L25

Can you please release the json you generated for running the experiments in the paper?

[Tsingularity]

I also thought this random.shuffle gonna shuffle the order of classes in the list, but after taking a closer look at the overall code, I find this line actually only shuffle the images' order inside each class.

And if u also take a look at the codes above line 25, u will find they actually apply sort() to the class list in line 17. Therefore I think if u directly use their code to generate your train/val/test split for CUB, u will end up getting the same split as the author.

Please correct me if I am wrong. (I am also not 100 percent sure lol)