Fix: Expected number of columns in feature data

We also expect the iteration number for the sample.

train_model.py

@@ -522,7 +522,7 @@ def validate_jmh_data(jmh_data):
 
 
 def validate_features_data(features_data, number_benchmarks):
-    expected_columns = ['benchmark_id', 'feature', 'feature_type', 'value']
+    expected_columns = ['benchmark_id', 'iteration', 'feature', 'feature_type', 'value']
     assert (features_data.columns == expected_columns).all()
 
     # Set of benchmark IDs matched number of benchmarks