Update code_eval.py

code_eval.py

@@ -190,8 +190,7 @@ class CodeEval(evaluate.Metric):
         correct = np.array(correct)
 
         ks = k
-        pass_at_k = {f"pass@{k}": estimate_pass_at_k(total, correct, k).mean() for k in ks if (total >= k).all()}
-        pass_at_k = {k: str(v.item()) for k, v in pass_at_k.items()}
+        pass_at_k = {f"pass@{k}": estimate_pass_at_k(total, correct, k).mean().item() for k in ks if (total >= k).all()}
 
         return pass_at_k, results
 
@@ -211,4 +210,4 @@ def estimate_pass_at_k(num_samples, num_correct, k):
         assert len(num_samples) == len(num_correct)
         num_samples_it = iter(num_samples)
 
-    return np.array([estimator(int(n), int(c), k) for n, c in zip(num_samples_it, num_correct)])
+    return np.array([estimator(int(n), int(c), k) for n, c in zip(num_samples_it, num_correct)])