Upload folder using huggingface_hub

src/streamlit_app.py

@@ -368,10 +368,33 @@ with tab7:
     ]
     
     df_evaluators = pd.DataFrame(data)
-    
-    st.dataframe(df_evaluators, width="stretch")
-    
-    # Optional: also show totals
+    # Show contributions as charts (by evaluator and aggregated by language)
+    st.markdown("**Contributions by evaluator**")
+    df_evals_sorted = df_evaluators.sort_values('Contributions', ascending=False)
+
+    chart_evaluator = alt.Chart(df_evals_sorted).mark_bar().encode(
+        x=alt.X('Evaluator ID:N', sort=df_evals_sorted['Evaluator ID'].tolist(), title='Evaluator ID'),
+        y=alt.Y('Contributions:Q', title='Contributions'),
+        color=alt.Color('Languages:N', legend=alt.Legend(title='Language')),
+        tooltip=['Evaluator ID', 'Contributions', 'Languages']
+    ).properties(height=300)
+
+    st.altair_chart(chart_evaluator, use_container_width=True)
+
+    st.markdown("**Total contributions by language**")
+    df_lang = df_evaluators.groupby('Languages', as_index=False).sum()
+    chart_lang = alt.Chart(df_lang).mark_bar().encode(
+        x=alt.X('Languages:N', sort='-y', title='Language'),
+        y=alt.Y('Contributions:Q', title='Total Contributions'),
+        tooltip=['Languages', 'Contributions']
+    ).properties(height=300)
+
+    st.altair_chart(chart_lang, use_container_width=True)
+
+    # Optional: also show totals and raw table inside an expander
+    with st.expander("View raw evaluator table"):
+        st.dataframe(df_evaluators, width="stretch")
+
     st.write("### Summary")
     st.write(f"- **Total Evaluators:** {df_evaluators['Evaluator ID'].nunique()}")
     st.write(f"- **Total Contributions:** {df_evaluators['Contributions'].sum()}")
@@ -403,8 +426,56 @@ with tab7:
     ]
     
     df_results = pd.DataFrame(results_data)
-    st.dataframe(df_results, width="stretch")
-    
+    # Visualize results with charts
+    # Grouped bars for Accuracy and Meaning per language
+    df_long = df_results.melt(id_vars=['Language', 'Audios Evaluated'], value_vars=['Accuracy', 'Meaning'],
+                              var_name='Metric', value_name='Score')
+
+    # Use xOffset for grouped bars when available
+    try:
+        chart_metrics = alt.Chart(df_long).mark_bar().encode(
+            x=alt.X('Language:N', sort=df_results['Language'].tolist(), title='Language'),
+            y=alt.Y('Score:Q', title='Score (1-5)'),
+            color=alt.Color('Metric:N', legend=alt.Legend(title='Metric')),
+            tooltip=['Language', 'Metric', 'Score'],
+            xOffset='Metric:N'
+        ).properties(height=360)
+    except Exception:
+        # Fallback when xOffset is not supported: side-by-side by using column
+        chart_metrics = alt.Chart(df_long).mark_bar().encode(
+            x=alt.X('Language:N', sort=df_results['Language'].tolist(), title='Language'),
+            y=alt.Y('Score:Q', title='Score (1-5)'),
+            color=alt.Color('Metric:N', legend=alt.Legend(title='Metric')),
+            tooltip=['Language', 'Metric', 'Score']
+        ).properties(height=360)
+
+    st.altair_chart(chart_metrics, use_container_width=True)
+
+    # Scatter: Accuracy vs Audios Evaluated (size / color by language)
+    chart_scatter = alt.Chart(df_results).mark_circle(size=120).encode(
+        x=alt.X('Audios Evaluated:Q', title='Audios Evaluated'),
+        y=alt.Y('Accuracy:Q', title='Accuracy'),
+        color=alt.Color('Language:N', legend=None),
+        tooltip=['Language', 'Audios Evaluated', 'Accuracy', 'Meaning']
+    ).properties(height=360)
+
+    st.markdown("**Accuracy vs. Dataset Size**")
+    st.altair_chart(chart_scatter, use_container_width=True)
+
+    # Audios evaluated per language (bar)
+    chart_audios = alt.Chart(df_results).mark_bar().encode(
+        x=alt.X('Language:N', sort=df_results['Language'].tolist(), title='Language'),
+        y=alt.Y('Audios Evaluated:Q', title='Audios Evaluated'),
+        tooltip=['Language', 'Audios Evaluated']
+    ).properties(height=320)
+
+    st.markdown("**Number of audios evaluated by language**")
+    st.altair_chart(chart_audios, use_container_width=True)
+
+    # Optional: raw table in expander
+    with st.expander('View raw findings table'):
+        st.dataframe(df_results, width='stretch')
+
     # Narrative summary
     st.markdown("""
     ### Key Takeaways