Show dataset/token status; mark synthetic fallback

app.py

@@ -232,6 +232,9 @@ def _ensure_local_file(local_path: Path, hub_filename: str) -> Optional[Path]:
         return None
 
 
+USING_SYNTHETIC_DATA = False
+
+
 def load_augmented_samples() -> Tuple[List[Dict[str, object]], bool]:
     moe_path = _ensure_local_file(MOE_DATA_PATH, "demo_data_moe.pt")
     base_path = _ensure_local_file(DEMO_DATA_PATH, "demo_data.pt")
@@ -245,6 +248,8 @@ def load_augmented_samples() -> Tuple[List[Dict[str, object]], bool]:
         return _safe_load_tensor(base_path), False
 
     # Last resort: in-memory synthetic data (keeps app alive, but clearly not the full demo dataset).
+    global USING_SYNTHETIC_DATA
+    USING_SYNTHETIC_DATA = True
     print(
         "[WARN] Falling back to a tiny synthetic dataset (30 samples). "
         "This usually means the real demo_data*.pt could not be downloaded. "
@@ -894,7 +899,12 @@ mapping_info = load_joint_mapping()
 df, has_moe_embeddings = load_data(mapping_info)
 CLASS_LABELS = mapping_info["label_names"]
 
-DATASET_STATUS = f"Dataset loaded: {len(df)} samples | MoE embeddings: {'yes' if has_moe_embeddings else 'no'}"
+DATASET_STATUS = (
+    f"Dataset loaded: {len(df)} samples | "
+    f"MoE embeddings: {'yes' if has_moe_embeddings else 'no'} | "
+    f"HF token detected: {'yes' if HF_TOKEN else 'no'} | "
+    f"Synthetic fallback: {'yes' if USING_SYNTHETIC_DATA else 'no'}"
+)
 
 has_moe_column = df["moe_embedding"].apply(lambda x: x is not None)
 joint_eval_df = df[has_moe_column & df["joint_label_id"].notna()]