controllable-text-attribute-transfer
controllable-text-attribute-transfer copied to clipboard
Code for "Controllable Unsupervised Text Attribute Transfer via Editing Entangled Latent Representation" (NeurIPS 2019)
I was going through the outputs given at https://github.com/Nrgeup/controllable-text-attribute-transfer/blob/master/outputs/my-model-yelp2.txt and observed it is not for the **W=6**. Can you publish the results for **W=6**, which I believe gives the best...
Hello! We are trying to reproduce your paper for the neurips reproducibility challenge. First off, thanks for taking the time to make a clean codebase and providing a clear way...
https://github.com/Nrgeup/controllable-text-attribute-transfer/blob/74cf2633080870a376a769aea9ce731813126b6a/method/mymodel-yelp/model2.py#L217 请问可以提供更多信息关于 AttentionScore 这个模块么?在论文中好像没有找到。
在模型代码model.py 381行中,损失函数使用了KL散度,但是在论文中,损失函数是交叉熵。请问是我理解错了吗?
Thank you for your great repo. I'm try to calculate style accuracy, Bleu score, and PPL using my code. But when i use `my-model-yelp1.txt` and `my-model-yelp2.txt`, they didn't show high...