.

code_eval.py

@@ -145,8 +145,9 @@ class CodeEval(evaluate.Metric):
             # This defines the format of each prediction and reference
             features=datasets.Features(
                 {
-                    "predictions": List[Dict],
-                    "references": List[Dict],
+                    # sequence of dictionaries
+                    "predictions": datasets.Sequence(dict()),
+                    "references": datasets.Sequence(dict()),
                 }
             ),
             homepage="https://github.com/openai/human-eval",