How to Test your model

[hyeseonko]

Hi, I have two questions !! (I finished training of your nvrnn (yelp-normal, yelp-vmf, ptb-normal, ptb-vmf).)

1. How to see the generated samples? After the end of training, I could see the [Recon Loss, KL Loss, Test Loss and Test PPL] But I don't know how to see the generated samples given specific token. Do you know how to see generated samples using Yelp dataset with label?

2. How to Test your model? I couldn't find explanation for how to test your model. Is it eval_nvrnn.py? I changed the directory path and run eval_nvrnn.py, but I got this error.

C:\Users\hsko0\Anaconda3\lib\site-packages\torch\nn\functional.py:52: UserWarning: size_average and reduce args will be deprecated, please use reduction='none' instead. warnings.warn(warning.format(ret)) Traceback (most recent call last): File "D:/Generation/vMF-VAE/eval_nvrnn.py", line 336, in cur_loss, cur_kl, test_loss = player.eva() File "D:/Generation/vMF-VAE/eval_nvrnn.py", line 60, in eva self.data.test) File "D:/Generation/vMF-VAE/eval_nvrnn.py", line 91, in evaluate recon_loss, kld, aux_loss, tup, vecs = model(feed, target) ValueError: too many values to unpack (expected 5)

Thanks for reading this issue and sorry for bothering your time. Thank you so much. Sincerely, Hyeseon