Support scalar results in combined evaluations

src/evaluate/module.py

@@ -982,6 +982,10 @@ def compute(self, predictions=None, references=None, **kwargs):
 
     def _merge_results(self, results):
         merged_results = {}
+        results = [
+            result if isinstance(result, dict) else {module_name: result}
+            for module_name, result in zip(self.evaluation_module_names, results)
+        ]
         results_keys = list(itertools.chain.from_iterable([r.keys() for r in results]))
         duplicate_keys = {item for item, count in collections.Counter(results_keys).items() if count > 1}
 

tests/test_metric.py

@@ -757,3 +757,14 @@ def test_modules_from_string_poslabel(self):
         self.assertDictEqual(
             expected_result, combined_evaluation.compute(predictions=predictions, references=references, pos_label=0)
         )
+
+    def test_combined_evaluation_with_scalar_results(self):
+        predictions = ["this is the prediction", "there is an other sample"]
+        references = ["this is the reference", "there is another one"]
+        expected_result = {"wer": 0.5, "cer": 0.34146341463414637}
+
+        combined_evaluation = combine(["wer", "cer"])
+
+        self.assertDictEqual(
+            expected_result, combined_evaluation.compute(predictions=predictions, references=references)
+        )