filtering by track

app.py

@@ -61,7 +61,7 @@ def init_leaderboard(dataframe, track):
     if dataframe is None or dataframe.empty:
         raise ValueError("Leaderboard DataFrame is empty or None.")
     # filter for correct track
-    dataframe = dataframe.loc[dataframe["eval_name"].endswith(track.lower())]
+    dataframe = dataframe.loc[dataframe["track"].lower() == track.lower()]
     return Leaderboard(
         value=dataframe,
         datatype=[c.type for c in fields(AutoEvalColumn)],

src/display/utils.py

@@ -25,6 +25,7 @@ auto_eval_column_dict = []
 auto_eval_column_dict_multimodal = []
 # Init
 auto_eval_column_dict.append(["model", ColumnContent, ColumnContent("Model", "markdown", True, never_hidden=True)])
+auto_eval_column_dict.append(["track", ColumnContent, ColumnContent("Track", "markdown", False)])
 #Scores
 auto_eval_column_dict.append(["average", ColumnContent, ColumnContent("Average ⬆️", "number", True)])
 for task in Tasks:
@@ -34,6 +35,7 @@ auto_eval_column_dict.append(["still_on_hub", ColumnContent, ColumnContent("Avai
 auto_eval_column_dict.append(["revision", ColumnContent, ColumnContent("Model sha", "str", False, False)])
 
 auto_eval_column_dict_multimodal.append(["model", ColumnContent, ColumnContent("Model", "markdown", True, never_hidden=True)])
+auto_eval_column_dict.append(["track", ColumnContent, ColumnContent("Track", "markdown", False)])
 auto_eval_column_dict_multimodal.append(["average", ColumnContent, ColumnContent("Average ⬆️", "number", True)])
 for task in TasksMultimodal:
     auto_eval_column_dict_multimodal.append([task.name, ColumnContent, ColumnContent(task.value.col_name, "number", True)])

src/populate.py

@@ -14,6 +14,7 @@ def get_leaderboard_df(results_path: str, requests_path: str, cols: list, benchm
     all_data_json = [v.to_dict() for v in raw_data]
     all_data_json_filtered = []
     for item in all_data_json:
+        item["track"] = item["eval_name"].split("_")[-1]
         if "VQA" in benchmark_cols and "VQA" in item:
             all_data_json_filtered.append(item)
         if "VQA" not in benchmark_cols and "VQA" not in item: