Ensure the spaces does not fail if no evaluation results are not populated yet

app.py

@@ -24,9 +24,14 @@ with tab1:
     labels = {dialect: dataset[dialect] for dialect in DIALECTS_WITH_LABELS}
 
     # Load the models' predictions
-    model_predictions_rows = datasets.load_dataset(
-        os.environ["PREDICTIONS_DATASET_NAME"]
-    )["train"]
+    try:
+        model_predictions_rows = datasets.load_dataset(
+            os.environ["PREDICTIONS_DATASET_NAME"]
+        )["train"]
+
+    except Exception as e:
+        st.error(f"An error occurred while loading the predictions: {e}")
+        model_predictions_rows = []
 
     evaluation_metrics = []
     for row in model_predictions_rows: