NimbusML model loaded in ML.NET cannot create PredictionEngine

[najeeb-kazmi]

Found while investigating #201

Describe the bug Cannot create a PredictionEngine in ML.NET from a model trained in NimbusML. Can only use model.Transform(data) to make predictions, whether single or batch.

To Reproduce Python example on UCI Adult:

train = FileDataStream.read_csv(train_file, sep=',', numeric_dtype=np.float32)

pipe = Pipeline([ColumnConcatenator(columns={'Features': ['age']}),
                AveragedPerceptronBinaryClassifier(feature=['Features'], label='label')])

pipe.fit(train)
pipe.save_model(nimbus_model_file)

ML.NET code:

MLContext mlContext = new MLContext();

ITransformer mlModel = mlContext.Model.Load(GetAbsolutePath(MODEL_FILEPATH), out DataViewSchema inputSchema);
var predEngine = mlContext.Model.CreatePredictionEngine<ModelInput, ModelOutput>(mlModel);

Expected behavior Expected PredictionEngine to be created to make single prediction.

Error message

Unhandled Exception: System.Reflection.TargetInvocationException: Exception has been thrown by the target of an invocation. ---> System.InvalidOperationException: Invalid TValue in GetGetter for column #label: Boolean: 'System.Int64'
   at Microsoft.ML.Data.DataViewConstructionUtils.InputRowBase`1.GetGetter[TValue](Column column)
   at Microsoft.ML.Data.RowToRowMapperTransform.RowImpl.GetGetter[TValue](Column column)
   at Microsoft.ML.Transforms.OptionalColumnTransform.GetSrcGetter[T](DataViewRow input, Int32 iinfo)
   --- End of inner exception stack trace ---
   at System.RuntimeMethodHandle.InvokeMethod(Object target, Object[] arguments, Signature sig, Boolean constructor, Boolean wrapExceptions)
   at System.Reflection.RuntimeMethodInfo.Invoke(Object obj, BindingFlags invokeAttr, Binder binder, Object[] parameters, CultureInfo culture)
   at Microsoft.ML.Transforms.OptionalColumnTransform.CreateGetters(DataViewRow input, IEnumerable`1 activeColumns, Action& disposer)
   at Microsoft.ML.Data.RowToRowMapperTransformBase.Microsoft.ML.Data.IRowToRowMapper.GetRow(DataViewRow input, IEnumerable`1 activeColumns)
   at Microsoft.ML.Data.CompositeRowToRowMapper.Microsoft.ML.Data.IRowToRowMapper.GetRow(DataViewRow input, IEnumerable`1 activeColumns)
   at Microsoft.ML.Data.CompositeRowToRowMapper.Microsoft.ML.Data.IRowToRowMapper.GetRow(DataViewRow input, IEnumerable`1 activeColumns)
   at Microsoft.ML.PredictionEngineBase`2.PredictionEngineCore(IHostEnvironment env, InputRow`1 inputRow, IRowToRowMapper mapper, Boolean ignoreMissingColumns, SchemaDefinition outputSchemaDefinition, Action& disposer, IRowReadableAs`1& outputRow)
   at Microsoft.ML.PredictionEngineBase`2..ctor(IHostEnvironment env, ITransformer transformer, Boolean ignoreMissingColumns, SchemaDefinition inputSchemaDefinition, SchemaDefinition outputSchemaDefinition)
   at Microsoft.ML.PredictionEngineExtensions.CreatePredictionEngine[TSrc,TDst](ITransformer transformer, IHostEnvironment env, Boolean ignoreMissingColumns, SchemaDefinition inputSchemaDefinition, SchemaDefinition outputSchemaDefinition)
   at SampleBinaryClassification.ConsoleApp.Program.Main(String[] args) in D:\Data\UCIAdult\SampleBinaryClassification\SampleBinaryClassification.ConsoleApp\Program.cs:line 32

C:\Program Files\dotnet\dotnet.exe (process 195176) exited with code -532462766.
Press any key to close this window . . .