Nan is not captured properly

src/display_models/read_results.py

@@ -1,4 +1,5 @@
 import json
+import math
 import os
 from dataclasses import dataclass
 from typing import Dict, List, Tuple
@@ -96,7 +97,7 @@ def parse_eval_result(json_filepath: str) -> Tuple[str, list[dict]]:
     eval_results = []
     for benchmark, metric in zip(BENCHMARKS, METRICS):
         accs = np.array([v.get(metric, None) for k, v in data["results"].items() if benchmark in k])
-        if accs.size == 0 or any([acc is None for acc in accs]):
+        if accs.size == 0 or any([acc is None or math.isnan(acc) for acc in accs]):
             continue
         mean_acc = np.mean(accs) * 100.0
         eval_results.append(