Update app.py

app.py

@@ -65,38 +65,59 @@ def main():
     st.markdown("Leaderboard made with [🧐 LLM AutoEval](https://github.com/mlabonne/llm-autoeval) using [Nous](https://huggingface.co/NousResearch) benchmark suite. It's a collection of my own evaluations.")
 
     content = create_yall()
-    if content:
-        try:
-            score_columns = ['Average', 'AGIEval', 'GPT4All', 'TruthfulQA', 'Bigbench']
-
-            # Display dataframe
-            df = convert_markdown_table_to_dataframe(content)
-            for col in score_columns:
-                df[col] = pd.to_numeric(df[col].str.strip(), errors='coerce')
-            st.dataframe(df, use_container_width=True)
-
-            # Full-width plot for the first category
-            create_bar_chart(df, score_columns[0])
-
-            # Next two plots in two columns
-            col1, col2 = st.columns(2)
-            with col1:
-                create_bar_chart(df, score_columns[1])
-            with col2:
-                create_bar_chart(df, score_columns[2])
-
-            # Last two plots in two columns
-            col3, col4 = st.columns(2)
-            with col3:
-                create_bar_chart(df, score_columns[3])
-            with col4:
-                create_bar_chart(df, score_columns[4])
-
-        except Exception as e:
-            st.error("An error occurred while processing the markdown table.")
-            st.error(str(e))
-    else:
-        st.error("Failed to download the content from the URL provided.")
+    tab1, tab2 = st.tabs(["🏆 Leaderboard", "📝 About"])
+
+    # Leaderboard tab
+    with tab1:
+        if content:
+            try:
+                score_columns = ['Average', 'AGIEval', 'GPT4All', 'TruthfulQA', 'Bigbench']
+    
+                # Display dataframe
+                df = convert_markdown_table_to_dataframe(content)
+                for col in score_columns:
+                    df[col] = pd.to_numeric(df[col].str.strip(), errors='coerce')
+                st.dataframe(df, use_container_width=True)
+    
+                # Full-width plot for the first category
+                create_bar_chart(df, score_columns[0])
+    
+                # Next two plots in two columns
+                col1, col2 = st.columns(2)
+                with col1:
+                    create_bar_chart(df, score_columns[1])
+                with col2:
+                    create_bar_chart(df, score_columns[2])
+    
+                # Last two plots in two columns
+                col3, col4 = st.columns(2)
+                with col3:
+                    create_bar_chart(df, score_columns[3])
+                with col4:
+                    create_bar_chart(df, score_columns[4])
+    
+            except Exception as e:
+                st.error("An error occurred while processing the markdown table.")
+                st.error(str(e))
+        else:
+            st.error("Failed to download the content from the URL provided.")
+
+     # About tab
+    with tab2:
+        st.markdown('''
+            ## Nous benchmark suite
+
+            Popularized by [Teknium](https://huggingface.co/teknium) and [NousResearch](https://huggingface.co/NousResearch), this benchmark suite aggregates four benchmarks:
+            
+            * [**AGIEval**](https://arxiv.org/abs/2304.06364) (0-shot): `agieval_aqua_rat,agieval_logiqa_en,agieval_lsat_ar,agieval_lsat_lr,agieval_lsat_rc,agieval_sat_en,agieval_sat_en_without_passage,agieval_sat_math`
+            * **GPT4ALL** (0-shot): `hellaswag,openbookqa,winogrande,arc_easy,arc_challenge,boolq,piqa`
+            * [**TruthfulQA**](https://arxiv.org/abs/2109.07958) (0-shot): `truthfulqa_mc`
+            * [**Bigbench**](https://arxiv.org/abs/2206.04615) (0-shot): `bigbench_causal_judgement,bigbench_date_understanding,bigbench_disambiguation_qa,bigbench_geometric_shapes,bigbench_logical_deduction_five_objects,bigbench_logical_deduction_seven_objects,bigbench_logical_deduction_three_objects,bigbench_movie_recommendation,bigbench_navigate,bigbench_reasoning_about_colored_objects,bigbench_ruin_names,bigbench_salient_translation_error_detection,bigbench_snarks,bigbench_sports_understanding,bigbench_temporal_sequences,bigbench_tracking_shuffled_objects_five_objects,bigbench_tracking_shuffled_objects_seven_objects,bigbench_tracking_shuffled_objects_three_objects`
+
+            ## Reproducibility
+
+            You can easily reproduce these results using [🧐 LLM AutoEval](https://github.com/mlabonne/llm-autoeval/tree/master), a colab notebook that automates the evaluation process (benchmark: `nous`). This will upload the results to GitHub as gists. You can find the entire table with the links to the detailed results [here](https://gist.github.com/mlabonne/90294929a2dbcb8877f9696f28105fdf).
+        ''')
 
 if __name__ == "__main__":
     main()