update

app.py

@@ -62,10 +62,6 @@ accuracy_df.sort_values(by="⭐ Overall", ascending=False, inplace=True)
 
 
 def load_heatmap(evt: gr.SelectData):
-    print(
-        f"You selected {evt.value} at row {evt.index[0]}, column {evt.index[1]} from the Dataframe."
-    )
-
     heatmap_image = gr.Image(f"results/{evt.value}.jpg")
     return heatmap_image
 
@@ -73,13 +69,20 @@ def load_heatmap(evt: gr.SelectData):
 with gr.Blocks() as demo:
     gr.Markdown("# FSM Benchmark Leaderboard")
     # add link to home page and dataset
+    with gr.Tab("Text-only Benchmark"):
+
+        leader_board = gr.Dataframe(accuracy_df, headers=headers_with_icons)
 
-    leader_board = gr.Dataframe(accuracy_df, headers=headers_with_icons)
+        gr.Markdown("## Heatmap")
 
-    gr.Markdown("## Heatmap")
+        heatamp_image = gr.Image(label="", show_label=False)
 
-    heatamp_image = gr.Image(label="", show_label=False)
+        leader_board.select(fn=load_heatmap, outputs=[heatamp_image])
 
-    leader_board.select(fn=load_heatmap, outputs=[heatamp_image])
+    with gr.Tab("Vision Benchmark"):
+        gr.Markdown("# TBA")
+        leader_board_vision = gr.Dataframe()
+        gr.Markdown("## Heatmap")
+        heatamp_image_vision = gr.Image(label="", show_label=False)
 
     demo.launch()

results/GPT-4-0125-preview.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a7e07fc568a31f3c7c77b212d271f466dfd2d1bd52fb4702761badd75dbc945
+size 18068025