Closed zixufang closed 5 years ago
Hi zixufang,
I'm glad this project is useful to you. We used the Daily Dialogue dataset (yanran2017dailydialog). Please bear in mind that this work is no longer actively updated nor fully tested!
thank you. So, you put up a new model called "DPAC". You didn't write a paper about your model? I didn't find relevant paper about it .
We did, I will add it to the project. We never sent it to any conference, it was just for the project evaluation for the University of Amsterdam.
May I ask ,which discriminator have you tested? The "discriminator.py" or "discriminator_LM.py" or "discriminator_LM2.py"? I came across model collapse when I was training with SEQGAN.It has disturbed me for a few days. I would appreciate it if you can give me the info!!
I think we tested discriminator.py (binary discriminator) and discriminator_LM2.py (language model based discriminator). If I recall correctly, LM1 was replaced by LM2 because we had problems with the implementation.
Hi,I am so happy to find this project. But I am confused about the dataset. May I ask where the dataset come from? And is it a public dataset? I find no other way to contact you, so I can only use this way to seek your help.