some fixes

src/streamlit_app.py

@@ -438,34 +438,6 @@ def main():
     # Main content
     st.header(f"Results for {selected_dataset}")
     
-    # Debug information (can be removed later)
-    with st.expander("Debug Information"):
-        st.write(f"Total rows in filtered_df: {len(filtered_df)}")
-        st.write(f"Total rows in ui_metrics_df: {len(ui_metrics_df)}")
-        if not filtered_df.empty:
-            st.write("Sample data from filtered_df:")
-            st.write(filtered_df[['model', 'base_model', 'is_checkpoint', 'overall_accuracy']].head())
-            
-            # Show UI type results structure
-            st.write("\nUI Type Results Structure:")
-            for idx, row in filtered_df.head(2).iterrows():
-                st.write(f"\nModel: {row['model']}")
-                ui_results = row.get('ui_type_results', {})
-                if ui_results:
-                    st.write("UI Type Keys:", list(ui_results.keys()))
-                    # Show a sample of the structure
-                    for key in list(ui_results.keys())[:2]:
-                        st.write(f"  {key}: {ui_results[key]}")
-                else:
-                    st.write("  No UI type results found")
-                    
-                # Also check dataset_type_results
-                dataset_type_results = row.get('dataset_type_results', {})
-                if dataset_type_results:
-                    st.write("Dataset Type Results Keys:", list(dataset_type_results.keys()))
-                    for key in list(dataset_type_results.keys())[:2]:
-                        st.write(f"  {key}: {dataset_type_results[key]}")
-    
     # Overall metrics
     col1, col2, col3 = st.columns(3)
     with col1: