wuzhen247 / huapa Goto Github PK

According to your directory structure。run python train.py --n_class 10 --dataset imdb or python train.py --n_class 10 --data imdb? Under directory code，run python train.py --n_class 10 --data imdb? thank u.
my email [email protected]。 Can you give me some brief instructions?

I ran the original code directly, but did not achieve the results in the paper.
As to yelp13, the accuracy is 67.12 and rmse is 65.17, after 160 epochs.
As to yelp14, the accuracy and rmse achieves similar results in paper.

I'm a little confused... Is the result on yelp13 normal? What could be the factors that lead to the differences between results of the code and results in the paper?

Thank you very much!

It was killed while running. Who can tell me the possible reason? Thank you!

您好！我想请问下dataset里面都是ss文件，我想问下这个可以转换成txt文件吗？或者这个ss文件应该怎么用呢？

Hi
I can not download embedding files
could you please send me this files ?

I can not run ur code with error '' TypeError: Expected int32, got list containing Tensors of type '_Message' instead.'' so can u tell me the version about ur project. thx

我正在尝试运行您的代码来复现您论文中的成果。我检查过之后发现代码中的默认参数与您论文中所提到的配置是一样的。
但是跑了两次之后我发现这个代码的默认参数在IMDB数据集上只能跑到0.540左右的成绩（每次大概跑25000个step）。我想知道的是：

1.代码中是否有什么地方的配置是论文中没有提到的，还需要去调的吗？
2.如果有，可否提供一下这些参数的配置？

I couldn't run this project on Windows. Is there any version I can run on Windows? Thank you very much!

user_num和user_id是一样的吗？

While starting the training, the embedding file for the dataset is required. It has to be generated by ourselves? If so can you provide us the link to the dataset?

If I want to use BiLSTM and user and product attention, how could I unify them? Or which point in your model could I change in your opinion? Thanks very much! I want your advice!