Lucas Gomez Jimenez

Thanks for your reply Huang. Based on your comment I tried to skip the RaggedTensorToTensor conversion by adding the "ragged=True" parameter to the TextVectorization constructor, however I now get a...

> > Thanks for your reply Huang. > > Based on your comment I tried to skip the RaggedTensorToTensor conversion by adding the "ragged=True" parameter to the TextVectorization constructor, however...

> And the ONNX prediction is wrong: > ``` > [array([[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,...