Write metric names for csv mode, and don't check params if not in mod…

…el_info

src/spikeinterface/curation/model_based_curation.py

@@ -109,7 +109,8 @@ def predict_labels(
         probabilities = np.max(probabilities, axis=1)
 
         if isinstance(label_conversion, dict):
-            if set(predictions) != set(label_conversion.keys()):
+
+            if set(predictions).issubset(set(label_conversion.keys())) is False:
                 raise ValueError("Labels in predictions do not match those in label_conversion")
             predictions = [label_conversion[label] for label in predictions]
 
@@ -161,9 +162,10 @@ def _check_params_for_classification(self, enforce_metric_params=False, model_in
 
             # remove the 's' at the end of the extension name
             extension_name = extension_name[:-1]
-            if metric_extension is not None:
+            model_metric_params = model_info["metric_params"].get(extension_name + "_params")
+
+            if metric_extension is not None and model_metric_params is not None:
 
-                model_metric_params = model_info["metric_params"][extension_name + "_params"]
                 metric_params = metric_extension.params["metric_params"]
 
                 inconsistent_metrics = []

src/spikeinterface/curation/train_manual_curation.py

@@ -301,6 +301,9 @@ def _check_metrics_parameters(self, analyzers, enforce_metric_params):
     def load_and_preprocess_csv(self, paths):
         self._load_data_files(paths)
         self.process_test_data_for_classification()
+        self.metrics_params = {}
+        for metric_name in self.metric_names:
+            self.metrics_params[metric_name] = {}
 
     def process_test_data_for_classification(self):
         """