I’ve trained my model using mxnet_1.3 python HybridBlock. Then I load the trained model in Scala. I didn’t get any errors when I run [demo.Scala,](https://github.com/ustcfd/mxnet_scala/blob/master/scala_model_load/demo.scala) and the output is shown in the following.
 But now I got few problems: - Is possible to get the true value of the NDarray? I followed the examples of the API [https://mxnet.apache.org/api/scala/ndarray.html](url) trying to print the value of NDarray by toarray. But I got an odd output looks like “F@4b553d2d” including the final output. I am very confused about the output value of NDArray.The demo code is shown in demo.Scala. - The API document of Mxnet demonstrates that the type input of the NDArray.array is restricted to Array[float]. Do I have to flatten my n-dimension matrix to one-dimension Array[float] when I want to create my own data？ Are there any other strategies to feed my data just like Numpy or python which directly take a matrix as an input? Maybe I was doing something wrong. Please suggest! Thank you [ Full content available at: https://github.com/apache/incubator-mxnet/issues/12680 ] This message was relayed via gitbox.apache.org for [email protected]
