Fix bug

app.py

@@ -26,7 +26,7 @@ from src.display.utils import (
     Precision
 )
 from src.envs import API, EVAL_REQUESTS_PATH, EVAL_RESULTS_PATH, QUEUE_REPO, REPO_ID, RESULTS_REPO, TOKEN
-from src.populate import get_evaluation_queue_df, get_leaderboard_df
+from src.populate import get_evaluation_queue_df, get_leaderboard_df, TASK_NAME_INVERSE_MAP
 from src.submission.submit import add_new_eval
 
 
@@ -59,20 +59,21 @@ AGENTIC_LEADERBOARD_DF = get_leaderboard_df(EVAL_RESULTS_PATH, EVAL_REQUESTS_PAT
     pending_eval_queue_df,
 ) = get_evaluation_queue_df(EVAL_REQUESTS_PATH, EVAL_COLS)
 
-def init_leaderboard(dataframe):
+def init_leaderboard(dataframe, benchmark_type):
     if dataframe is None or dataframe.empty:
         raise ValueError("Leaderboard DataFrame is empty or None.")
+    AutoEvalColumnSubset = [c for c in fields(AutoEvalColumn) if ((c.name=="Model") or (TASK_NAME_INVERSE_MAP.get(c.name, dict()).get("type", "")==benchmark_type))]
     return Leaderboard(
         value=dataframe,
-        datatype=[c.type for c in fields(AutoEvalColumn)],
+        datatype=[c.type for c in AutoEvalColumnSubset],
         select_columns=SelectColumns(
-            default_selection=[c.name for c in fields(AutoEvalColumn) if c.displayed_by_default],
-            cant_deselect=[c.name for c in fields(AutoEvalColumn) if c.never_hidden],
+            default_selection=[c.name for c in AutoEvalColumnSubset if c.displayed_by_default],
+            cant_deselect=[c.name for c in AutoEvalColumnSubset if c.never_hidden],
             label="Select Columns to Display:",
         ),
         # search_columns=[AutoEvalColumn.model.name, AutoEvalColumn.license.name],
         search_columns=[AutoEvalColumn.model.name,],
-        hide_columns=[c.name for c in fields(AutoEvalColumn) if c.hidden],
+        hide_columns=[c.name for c in AutoEvalColumnSubset if c.hidden],
         # filter_columns=[
         #     ColumnFilter(AutoEvalColumn.model_type.name, type="checkboxgroup", label="Model types"),
         #     ColumnFilter(AutoEvalColumn.precision.name, type="checkboxgroup", label="Precision"),
@@ -100,10 +101,10 @@ with demo:
 
     with gr.Tabs(elem_classes="tab-buttons") as tabs:
         with gr.TabItem("Single-turn Benchmark", elem_id="llm-benchmark-tab-table", id=0):
-            leaderboard = init_leaderboard(ST_LEADERBOARD_DF)
+            leaderboard = init_leaderboard(ST_LEADERBOARD_DF, "single-turn")
 
         with gr.TabItem("Agentic Benchmark", elem_id="llm-benchmark-tab-table", id=1):
-            leaderboard = init_leaderboard(AGENTIC_LEADERBOARD_DF)
+            leaderboard = init_leaderboard(AGENTIC_LEADERBOARD_DF, "agentic")
 
         with gr.TabItem("📝 About", elem_id="llm-benchmark-tab-table", id=2):
             gr.Markdown(LLM_BENCHMARKS_TEXT, elem_classes="markdown-text")