Update space

app.py

@@ -465,7 +465,7 @@ def save_submission(submission_data, csv_file):
 def update_leaderboard_data(submission_data):
     """
     Update leaderboard data with new submission results
-    Only uses model name in the displayed table
+    Only updates the specific dataset submitted, preventing empty rows
     """
     global df_synthesized_full, df_synthesized_10, df_human_generated
     
@@ -477,26 +477,32 @@ def update_leaderboard_data(submission_data):
     }
     
     df_to_update = split_to_df[submission_data['Split']]
+    dataset = submission_data['Dataset'].upper()
     
-    # Prepare new row data
+    # Prepare new row data with only the relevant dataset columns
     new_row = {
-        'Method': submission_data['Method Name'],  # Only use method name in table
-        f'STARK-{submission_data["Dataset"].upper()}_Hit@1': submission_data['results']['hit@1'],
-        f'STARK-{submission_data["Dataset"].upper()}_Hit@5': submission_data['results']['hit@5'],
-        f'STARK-{submission_data["Dataset"].upper()}_R@20': submission_data['results']['recall@20'],
-        f'STARK-{submission_data["Dataset"].upper()}_MRR': submission_data['results']['mrr']
+        'Method': submission_data['Method Name']
     }
+    # Only add metrics for the submitted dataset
+    new_row.update({
+        f'STARK-{dataset}_Hit@1': submission_data['results']['hit@1'],
+        f'STARK-{dataset}_Hit@5': submission_data['results']['hit@5'],
+        f'STARK-{dataset}_R@20': submission_data['results']['recall@20'],
+        f'STARK-{dataset}_MRR': submission_data['results']['mrr']
+    })
     
     # Check if method already exists
     method_mask = df_to_update['Method'] == submission_data['Method Name']
     if method_mask.any():
-        # Update existing row
+        # Update only the columns for the submitted dataset
         for col in new_row:
             df_to_update.loc[method_mask, col] = new_row[col]
     else:
-        # Add new row
+        # For new methods, create a row with only the submitted dataset's values
         df_to_update.loc[len(df_to_update)] = new_row
 
+
+
 # Function to get emails from meta_data
 def get_emails_from_metadata(meta_data):
     """
@@ -689,7 +695,7 @@ def process_submission(
             submission_data = {
                 **meta_data,
                 "results": processed_results,
-                "status": "approved",  # or "pending_review"
+                "status": "pending_review",  # or "approved"
                 "submission_date": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
                 "csv_path": csv_path_in_repo
             }
@@ -711,7 +717,7 @@ def process_submission(
             # 3. Create or update latest.json
             latest_info = {
                 "latest_submission": timestamp,
-                "status": "approved",  # or "pending_review"
+                "status": "pending_review",  # or "approved"
                 "method_name": method_name,
                 "team_name": team_name
             }
@@ -786,10 +792,24 @@ def filter_by_model_type(df, selected_types):
     return df[df['Method'].isin(selected_models)]
 
 def format_dataframe(df, dataset):
+    """
+    Format DataFrame for display, removing rows with no data for the selected dataset
+    """
+    # Select relevant columns
     columns = ['Method'] + [col for col in df.columns if dataset in col]
     filtered_df = df[columns].copy()
+    
+    # Remove rows where all metric columns are empty/NaN for this dataset
+    metric_columns = [col for col in filtered_df.columns if col != 'Method']
+    filtered_df = filtered_df.dropna(subset=metric_columns, how='all')
+    
+    # Rename columns to remove dataset prefix
     filtered_df.columns = [col.split('_')[-1] if '_' in col else col for col in filtered_df.columns]
-    filtered_df = filtered_df.sort_values('MRR', ascending=False)
+    
+    # Sort by MRR
+    if 'MRR' in filtered_df.columns:
+        filtered_df = filtered_df.sort_values('MRR', ascending=False)
+    
     return filtered_df
 
 def update_tables(selected_types):

submissions/debug_submission_none/latest.json

@@ -1,6 +0,0 @@
-{
-    "latest_submission": "20241024_125801",
-    "status": "approved",
-    "method_name": "debug-submission",
-    "team_name": "none"
-}

submissions/debug_submission_none/metadata_20241024_125801.json

@@ -1,21 +0,0 @@
-{
-    "Method Name": "debug-submission",
-    "Team Name": "none",
-    "Dataset": "mag",
-    "Split": "human_generated_eval",
-    "Contact Email(s)": "none",
-    "Code Repository": "none",
-    "Model Description": "none",
-    "Hardware": "none",
-    "(Optional) Paper link": "none",
-    "Model Type": "Others",
-    "results": {
-        "hit@1": 28.57,
-        "hit@5": 41.67,
-        "recall@20": 35.95,
-        "mrr": 35.94
-    },
-    "status": "approved",
-    "submission_date": "2024-10-24 12:58:41",
-    "csv_path": "submissions/debug_submission_none/predictions_20241024_125801.csv"
-}