Upload app.py with huggingface_hub

app.py

@@ -270,12 +270,19 @@ def _to_tensor(spec) -> torch.Tensor:
     return t
 
 
+def _normalize_label(val):
+    """Convert labels to hashable, comparable form."""
+    if isinstance(val, (list, tuple)):
+        return tuple(str(v) for v in val)
+    return str(val)
+
+
 def compute_eval(task: str):
     """Compute confusion matrix + macro F1 for the small demo set."""
     predictor = load_predictor()
     y_true, y_pred = [], []
 
-    max_samples = min(len(raw_samples), 500)  # keep eval lightweight in Spaces
+    max_samples = min(len(raw_samples), 200)  # keep eval lightweight in Spaces
     for sample in raw_samples[:max_samples]:
         spec = _to_tensor(sample["data"])
         try:
@@ -287,11 +294,12 @@ def compute_eval(task: str):
 
         if task == "comm":
             routing = res.get("routing") or []
-            pred = routing[0]["comm"] if routing else "Unknown"
-            true = sample["tech"]
+            pred = _normalize_label(routing[0]["comm"]) if routing else "Unknown"
+            true = _normalize_label(sample["tech"])
         else:  # snr_mobility
-            pred = res.get("label", res["predicted_class"])
-            true = (sample["snr"], sample["mob"])
+            pred_raw = res.get("label", res["predicted_class"])
+            pred = _normalize_label(pred_raw)
+            true = _normalize_label((sample["snr"], sample["mob"]))
         y_true.append(true)
         y_pred.append(pred)