search

tkipf / gcn

Implementation of Graph Convolutional Networks in TensorFlow
MIT License
7.1k stars 2k forks source link

version of Tensorflow2 #187

Open WuZhongSEU opened 3 years ago

WuZhongSEU commented 3 years ago

Hello,this is a very good GNN approach, which is highly integrated.I wonder to know if there is a version of Tensorflow2.Thank you.

simbadu1999 commented 3 years ago

几个tf1的函数修改一下就可以了，请问你知道怎么预处理新的数据集吗？

YuQiao0303 commented 3 years ago

亲测有效： 将所有.py文件中的 import tensorflow as tf 改为： import tensorflow.compat.v1 as tf tf.disable_eager_execution()

Hello, I've tried the following way and it worked for my tensorflow2:

In all .py files, modify "import tensorflow as tf" into the following two lines: import tensorflow.compat.v1 as tf tf.disable_eager_execution()

ratmcu commented 3 years ago

I created a pull request here for tf 2 : https://github.com/tkipf/keras-gcn/pull/66

banbsyip commented 2 years ago

几个tf1的函数修改一下就可以了，请问你知道怎么预处理新的数据集吗？

有个数据问题想问下，如果有一条数据没有label，那怎么存储这个label呢？是null还是别的字符啊？这个问题搜了很久，没有找到答案

simbadu1999 commented 2 years ago

这是来自QQ邮箱的自动回复邮件。您好，我是杜承豪，已收到您的邮件，谢谢啦，我会尽快给您回复。