add some more descriptive errors

evaluate.py

@@ -127,14 +127,14 @@ def submit_data(predictions_file: str,
     try:
         results_df = pd.read_csv(file_path)            
     except Exception as e:
-        return f"❌ Error reading results file: {str(e)}"
+        raise gr.Error(f"❌ Error reading results file: {str(e)}")
 
     if results_df.empty:
-        return gr.Error("The uploaded file is empty.")
+        raise gr.Error("The uploaded file is empty.")
     
     missing = set(ENDPOINTS) - set(results_df.columns)
     if missing:
-        return gr.Error(f"The uploaded file must contain all endpoint predictions {ENDPOINTS} as columns.")
+        raise gr.Error(f"The uploaded file must contain all endpoint predictions {ENDPOINTS} as columns, missing: {missing}")
 
     # Save participant record
     try:
@@ -150,8 +150,8 @@ def submit_data(predictions_file: str,
             consent_publication=paper_checkbox
         )
     except ValidationError as e:
-        return f"❌ Error in participant information: {str(e)}"
-    
+        raise gr.Error(f"❌ Error in participant information: {str(e)}")
+
     # Build destination filename in the dataset
     ts = datetime.datetime.now(datetime.timezone.utc).isoformat(timespec="seconds") # should keep default time so can be deserialized correctly
     try:
@@ -163,7 +163,7 @@ def submit_data(predictions_file: str,
             participant=participant_record
         )
     except ValidationError as e:
-        return f"❌ Error in metadata information: {str(e)}"  
+        raise gr.Error(f"❌ Error in metadata information: {str(e)}")
 
     safe_user = _safeify_username(user_state)
     destination_csv = f"submissions/{safe_user}_{ts}.csv"