test

src/leaderboard/read_evals.py

@@ -116,7 +116,6 @@ class EvalResult:
         print(AutoEvalColumn.dummy.name)
         print(AutoEvalColumn.average.name)
         average = sum([v for v in self.results.values() if v is not None]) / len(Tasks)
-        print(self.eval_name)
         data_dict = {
             "eval_name": self.eval_name,  # not a column, just a save name,
             # AutoEvalColumn.precision.name: self.precision.value.name,
@@ -200,12 +199,13 @@ def get_raw_eval_results(results_path: str, requests_path: str) -> list[EvalResu
     # print(eval_results)
     for v in eval_results.values():
         print(f"v : {v}")
-        # try:
-        v.to_dict() # we test if the dict version is complete
-        print("aaa")
-        results.append(v)
-        # except KeyError:  # not all eval values present
-        print("nono")
-        continue
+        try:
+            v.to_dict() # we test if the dict version is complete
+            print("aaa")
+            results.append(v)
+        except KeyError as e:  # not all eval values present
+            print(f"error : {e}")
+            print("nono")
+            continue
 
     return results