ML.NET training with floats arrays

Question

I fail to understand the transform component. Basically I have generated 1 million training samples

public class Data{

public float[] Input;

public float[] Output;

}

I want to train a net with 100 layers and then check the output prediction. Input is 1000 in size, and output is 500 in size

The library continue to throw "schema mismatch for feature column"

Answer 1

That error often indicates a discrepancy between the data schema expected by transform component and the actual data provided.

Like this:

[VectorType(1000)] public float[] Input 

And:

[VectorType(500)] public float[] Output