update app.py

app.py

@@ -18,9 +18,9 @@ from src.display.utils import (
     COLS,
     EVAL_COLS,
     EVAL_TYPES,
+    COLUMNS,  # Added this line
     AutoEvalColumn,
     ModelType,
-    fields,
     WeightType,
     Precision
 )
@@ -28,11 +28,10 @@ from src.envs import API, EVAL_REQUESTS_PATH, EVAL_RESULTS_PATH, QUEUE_REPO, REP
 from src.populate import get_evaluation_queue_df, get_leaderboard_df
 from src.submission.submit import add_new_eval
 
-
 def restart_space():
     API.restart_space(repo_id=REPO_ID)
 
-### Space initialisation
+### Space initialization
 try:
     print(EVAL_REQUESTS_PATH)
     snapshot_download(
@@ -62,30 +61,30 @@ def init_leaderboard(dataframe):
         # Instead of raising an error, display an empty leaderboard with a message
         print("Leaderboard DataFrame is empty. No models have been evaluated yet.")
         # Create an empty DataFrame with the necessary columns
-        dataframe = pd.DataFrame(columns=[c.name for c in fields(AutoEvalColumn)])
+        dataframe = pd.DataFrame(columns=[c.name for c in COLUMNS])
         # Optionally, you can add a message to the interface to inform users
     return Leaderboard(
         value=dataframe,
-        datatype=[c.type for c in fields(AutoEvalColumn)],
+        datatype=[c.type for c in COLUMNS],
         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 COLUMNS if c.displayed_by_default],
+            cant_deselect=[c.name for c in COLUMNS if c.never_hidden],
             label="Select Columns to Display:",
         ),
-        search_columns=[AutoEvalColumn.model.name, AutoEvalColumn.license.name],
-        hide_columns=[c.name for c in fields(AutoEvalColumn) if c.hidden],
+        search_columns=[c.name for c in COLUMNS if c.name in ["model", "license"]],
+        hide_columns=[c.name for c in COLUMNS if c.hidden],
         filter_columns=[
-            ColumnFilter(AutoEvalColumn.model_type.name, type="checkboxgroup", label="Model types"),
-            ColumnFilter(AutoEvalColumn.precision.name, type="checkboxgroup", label="Precision"),
+            ColumnFilter("model_type", type="checkboxgroup", label="Model types"),
+            ColumnFilter("precision", type="checkboxgroup", label="Precision"),
             ColumnFilter(
-                AutoEvalColumn.params.name,
+                "params",
                 type="slider",
                 min=0.01,
                 max=150,
                 label="Select the number of parameters (B)",
             ),
             ColumnFilter(
-                AutoEvalColumn.still_on_hub.name, type="boolean", label="Deleted/incomplete", default=True
+                "still_on_hub", type="boolean", label="Deleted/incomplete", default=True
             ),
         ],
         bool_checkboxgroup_label="Hide models",
@@ -100,7 +99,11 @@ with demo:
 
     with gr.Tabs(elem_classes="tab-buttons") as tabs:
         with gr.TabItem("🏅 LLM Benchmark", elem_id="llm-benchmark-tab-table", id=0):
-            leaderboard = init_leaderboard(LEADERBOARD_DF)
+            if LEADERBOARD_DF.empty:
+                gr.Markdown("No evaluations have been performed yet. The leaderboard is currently empty.")
+            else:
+                leaderboard = init_leaderboard(LEADERBOARD_DF)
+                leaderboard.render()  # Ensure the leaderboard is rendered
 
         with gr.TabItem("📝 About", elem_id="llm-benchmark-tab-table", id=2):
             gr.Markdown(LLM_BENCHMARKS_TEXT, elem_classes="markdown-text")
@@ -110,8 +113,7 @@ with demo:
                 with gr.Row():
                     gr.Markdown(EVALUATION_QUEUE_TEXT, elem_classes="markdown-text")
 
-                # Since the evaluation queues are empty, we can remove or hide these sections
-                # Alternatively, display a message
+                # Since the evaluation queues are empty, display a message
                 with gr.Column():
                     gr.Markdown("Evaluations are performed immediately upon submission. There are no pending or running evaluations.")
 
@@ -163,7 +165,7 @@ with demo:
                     precision,
                     weight_type,
                     model_type,
-                    num_examples_input  # Added this line
+                    num_examples_input  # Included this line
                 ],
                 submission_result,
             )
@@ -181,4 +183,4 @@ with demo:
 scheduler = BackgroundScheduler()
 scheduler.add_job(restart_space, "interval", seconds=1800)
 scheduler.start()
-demo.queue(default_concurrency_limit=40).launch()
+demo.queue(default_concurrency_limit=40).launch()