XML file describing the run, including user-defined evaluation measures.
A human-readable description of the model that was built.
A serialized description of the model that can be read by the tool that generated it.
ARFF file with instance-level predictions generated by the model.