I encountered error while working on tensorflow
ValueError: Items of feature_columns must be a _FeatureColumn. Given (type <class ‘tensorflow_hub.feature_column_v2._TextEmbeddingColumnV2’>): _TextEmbeddingColumnV2(key=‘sentence’, module_path=‘Google | universal-sentence-encoder | Kaggle’, output_key=None, trainable=False), can anyone help me?
Possible to provide the code? Otherwise, it’s difficult to get a sense of what’s going wrong.
yes, sure, how can I send to you?
Ideally please provide a small python script that demonstrates the bug. This shouldn’t be your whole program, instead it should be an example script that reproduces the issue in as little code as possible. That makes it easier to debug.
If it’s under, say, 100 lines, it might be reasonable to post that directly here in the forum. If it’s longer, consider sharing a public GitHub repo or a Colab.
It would also help to include which versions of Tensorflow and TF Hub you are using.