Update app.py

app.py

@@ -74,9 +74,9 @@ def get_arena_table(model_table_df):
         # model display name
         row.append(model_name)
 
-        row.append(
-            model_table_df["Language Model"].values[model_key]
-        )
+        # row.append(
+        #     model_table_df["Language Model"].values[model_key]
+        # )
         row.append(
             model_table_df["Open Source"].values[model_key]
         )
@@ -118,10 +118,56 @@ def get_arena_table(model_table_df):
         values.append(row)
     return values
 
-def build_leaderboard_tab(leaderboard_table_file, show_plot=False):
+def get_cn_table(model_table_df):
+    # sort by rating
+    model_table_df = model_table_df.sort_values(by=["Average Score"], ascending=False)
+    values = []
+    for i in range(len(model_table_df)):
+        row = []
+        model_key = model_table_df.index[i]
+        model_name = model_table_df["Model"].values[model_key]
+        # rank
+        row.append(i + 1)
+        # model display name
+        row.append(model_name)
+
+        row.append(
+            model_table_df["Open Source"].values[model_key]
+        )
+        row.append(
+            model_table_df["Text Recognition"].values[model_key]
+        )
+
+
+        row.append(
+            model_table_df["Relation Extraction"].values[model_key]
+        )
+
+        row.append(
+            model_table_df["Element Parsing"].values[model_key]
+        )
+
+        row.append(
+            model_table_df["Visual Text Understanding"].values[model_key]
+        )
+
+        row.append(
+            model_table_df["Knowledge Reasoning"].values[model_key]
+        )
+        
+        row.append(
+            model_table_df["Average Score"].values[model_key]
+        )
+        values.append(row)
+    return values
+
+def build_leaderboard_tab(leaderboard_table_file, leaderboard_table_file_2, show_plot=False):
     if leaderboard_table_file:
         data = load_leaderboard_table_csv(leaderboard_table_file)
+        data_2 = load_leaderboard_table_csv(leaderboard_table_file_2)
+        
         model_table_df = pd.DataFrame(data)
+        model_table_df_2 = pd.DataFrame(data_2)
         md_head = f"""
         # 🏆 OCRBench v2 Leaderboard
         | [GitHub](https://github.com/Yuliang-Liu/MultimodalOCR) | [Paper](https://arxiv.org/abs/2305.07895) |
@@ -137,7 +183,6 @@ def build_leaderboard_tab(leaderboard_table_file, show_plot=False):
                     headers=[
                         "Rank",
                         "Name",
-                        "Language Model",
                         "Open Source",
                         "Text Recognition",
                         "Text Referring",
@@ -153,7 +198,6 @@ def build_leaderboard_tab(leaderboard_table_file, show_plot=False):
                         "str",
                         "markdown",
                         "str",
-                        "str",
                         "number",
                         "number",
                         "number",
@@ -168,6 +212,39 @@ def build_leaderboard_tab(leaderboard_table_file, show_plot=False):
                     elem_id="arena_leaderboard_dataframe",
                     wrap=False,
                 )
+            with gr.Tab("Text Recognition", id=1):
+                arena_table_vals = get_cn_table(model_table_df_2)
+                md = "OCRBench is a comprehensive evaluation benchmark designed to assess the OCR capabilities of Large Multimodal Models. It comprises five components: Text Recognition, SceneText-Centric VQA, Document-Oriented VQA, Key Information Extraction, and Handwritten Mathematical Expression Recognition. The benchmark includes 1000 question-answer pairs, and all the answers undergo manual verification and correction to ensure a more precise evaluation."
+                gr.Markdown(md, elem_id="leaderboard_markdown")
+                gr.Dataframe(
+                    headers=[
+                        "Rank",
+                        "Name",
+                        "Open Source",
+                        "Text Recognition",
+                        "Relation Extraction",
+                        "Element Parsing",
+                        "Visual Text Understanding",
+                        "Knowledge Reasoning",
+                        "Average Score",
+                    ],
+                    datatype=[
+                        "str",
+                        "markdown",
+                        "str",
+                        "number",
+                        "number",
+                        "number",
+                        "number",
+                        "number",
+                        "number",
+                    ],
+                    value=arena_table_vals,
+                    elem_id="arena_leaderboard_dataframe",
+                    # height=700,
+                    # column_widths=[60, 120,150,100, 100, 100, 100, 100, 80],
+                    wrap=True,
+                )
     else:
         pass
     md_tail = f"""
@@ -176,7 +253,7 @@ def build_leaderboard_tab(leaderboard_table_file, show_plot=False):
     If you would like to include your model in the OCRBench leaderboard, please follow the evaluation instructions provided on [GitHub](https://github.com/Yuliang-Liu/MultimodalOCR), [VLMEvalKit](https://github.com/open-compass/VLMEvalKit) or [lmms-eval](https://github.com/EvolvingLMMs-Lab/lmms-eval) and feel free to contact us via email at [email protected]. We will update the leaderboard in time."""
     gr.Markdown(md_tail, elem_id="leaderboard_markdown")
 
-def build_demo(leaderboard_table_file):
+def build_demo(leaderboard_table_file, leaderboard_table_file_2):
     text_size = gr.themes.sizes.text_lg
 
     with gr.Blocks(
@@ -185,15 +262,16 @@ def build_demo(leaderboard_table_file):
         css=block_css,
     ) as demo:
         leader_components = build_leaderboard_tab(
-            leaderboard_table_file,show_plot=True
+            leaderboard_table_file, leaderboard_table_file_2, show_plot=True
         )
     return demo
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
     parser.add_argument("--share", action="store_true")
-    parser.add_argument("--OCRBench_file", type=str, default="./OCRBench.csv")
+    parser.add_argument("--OCRBench_file", type=str, default="./OCRBench_en.csv")
+    parser.add_argument("--OCRBench_file_2", type=str, default="./OCRBench_cn.csv")
     args = parser.parse_args()
 
-    demo = build_demo(args.OCRBench_file)
+    demo = build_demo(args.OCRBench_file, args.OCRBench_file_2)
     demo.launch()