Update app.py

app.py

@@ -5,29 +5,31 @@ from apscheduler.schedulers.background import BackgroundScheduler
 
 # --- Make sure these imports work relative to your file structure ---
 # Option 1: If src is a directory in the same folder as your script:
-from src.about import (
-    CITATION_BUTTON_LABEL,
-    CITATION_BUTTON_TEXT,
-    EVALUATION_QUEUE_TEXT, # Keep if used by commented-out submit tab
-    INTRODUCTION_TEXT,
-    LLM_BENCHMARKS_TEXT,
-    TITLE,
-)
-from src.display.css_html_js import custom_css
-from src.envs import REPO_ID # Keep if needed for restart_space or other functions
-from src.submission.submit import add_new_eval # Keep if using the submit tab
-
+try:
+    from src.about import (
+        CITATION_BUTTON_LABEL,
+        CITATION_BUTTON_TEXT,
+        EVALUATION_QUEUE_TEXT, # Keep if used by commented-out submit tab
+        INTRODUCTION_TEXT,
+        LLM_BENCHMARKS_TEXT,
+        TITLE,
+    )
+    from src.display.css_html_js import custom_css # Assuming this exists but might be empty
+    from src.envs import REPO_ID # Keep if needed for restart_space or other functions
+    from src.submission.submit import add_new_eval # Keep if using the submit tab
+    print("Successfully imported from src module.")
 # Option 2: If you don't have these files, define placeholders (REMOVE THIS if using Option 1)
-# print("Warning: Using placeholder values for src module imports.")
-# CITATION_BUTTON_LABEL="Citation"
-# CITATION_BUTTON_TEXT="Please cite us if you use this benchmark..."
-# EVALUATION_QUEUE_TEXT="Current evaluation queue:"
-# INTRODUCTION_TEXT="Welcome to the MLE-Dojo Benchmark Leaderboard."
-# LLM_BENCHMARKS_TEXT="Information about the benchmarks..."
-# TITLE="<h1>🏆 MLE-Dojo Benchmark Leaderboard</h1>"
-# custom_css=""
-# REPO_ID="your/space-id" # Replace with actual ID if needed
-# def add_new_eval(*args): return "Submission placeholder."
+except ImportError:
+    print("Warning: Using placeholder values because src module imports failed.")
+    CITATION_BUTTON_LABEL="Citation"
+    CITATION_BUTTON_TEXT="Please cite us if you use this benchmark..."
+    EVALUATION_QUEUE_TEXT="Current evaluation queue:"
+    INTRODUCTION_TEXT="Welcome to the MLE-Dojo Benchmark Leaderboard."
+    LLM_BENCHMARKS_TEXT="Information about the benchmarks..."
+    TITLE="<h1>🏆 MLE-Dojo Benchmark Leaderboard</h1>"
+    custom_css="" # Start with empty CSS if not imported
+    REPO_ID="your/space-id" # Replace with actual ID if needed
+    def add_new_eval(*args): return "Submission placeholder."
 # --- End Placeholder Definitions ---
 
 
@@ -36,14 +38,14 @@ from src.submission.submit import add_new_eval # Keep if using the submit tab
 # !!! IMPORTANT: Replace placeholder URLs with actual model/project pages. !!!
 # Verify organizer and license information for accuracy.
 data = [
-    {'model_name': 'gpt-4o-mini', 'url': 'https://openai.com/index/hello-gpt-4o/', 'organizer': 'OpenAI', 'license': 'Proprietary', 'MLE-Lite_Elo': 753, 'Tabular_Elo': 839, 'NLP_Elo': 758, 'CV_Elo': 754, 'Overall': 778},
-    {'model_name': 'gpt-4o', 'url': 'https://openai.com/index/hello-gpt-4o/', 'organizer': 'OpenAI', 'license': 'Proprietary', 'MLE-Lite_Elo': 830, 'Tabular_Elo': 861, 'NLP_Elo': 903, 'CV_Elo': 761, 'Overall': 841},
-    {'model_name': 'o3-mini', 'url': 'https://placeholder.url/o3-mini', 'organizer': 'Unknown', 'license': 'Unknown', 'MLE-Lite_Elo': 1108, 'Tabular_Elo': 1019, 'NLP_Elo': 1056, 'CV_Elo': 1207, 'Overall': 1096}, # Fill details later
-    {'model_name': 'deepseek-v3', 'url': 'https://deepseek.com/', 'organizer': 'DeepSeek AI', 'license': 'DeepSeek License', 'MLE-Lite_Elo': 1004, 'Tabular_Elo': 1015, 'NLP_Elo': 1028, 'CV_Elo': 1067, 'Overall': 1023},
-    {'model_name': 'deepseek-r1', 'url': 'https://deepseek.com/', 'organizer': 'DeepSeek AI', 'license': 'DeepSeek License', 'MLE-Lite_Elo': 1137, 'Tabular_Elo': 1053, 'NLP_Elo': 1103, 'CV_Elo': 1083, 'Overall': 1100},
-    {'model_name': 'gemini-2.0-flash', 'url': 'https://deepmind.google/technologies/gemini/flash/', 'organizer': 'Google', 'license': 'Proprietary (API)', 'MLE-Lite_Elo': 847, 'Tabular_Elo': 923, 'NLP_Elo': 860, 'CV_Elo': 978, 'Overall': 895},
-    {'model_name': 'gemini-2.0-pro', 'url': 'https://deepmind.google/technologies/gemini/#introduction', 'organizer': 'Google', 'license': 'Proprietary (API)', 'MLE-Lite_Elo': 1064, 'Tabular_Elo': 1139, 'NLP_Elo': 1028, 'CV_Elo': 973, 'Overall': 1054},
-    {'model_name': 'gemini-2.5-pro', 'url': 'https://deepmind.google/technologies/gemini/2-5-pro/', 'organizer': 'Google', 'license': 'Proprietary (API)', 'MLE-Lite_Elo': 1257, 'Tabular_Elo': 1150, 'NLP_Elo': 1266, 'CV_Elo': 1177, 'Overall': 1214},
+{'model_name': 'gpt-4o-mini', 'url': 'https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/', 'organizer': 'OpenAI', 'license': 'Proprietary', 'MLE-Lite_Elo': 753, 'Tabular_Elo': 839, 'NLP_Elo': 758, 'CV_Elo': 754, 'Overall': 778},
+{'model_name': 'gpt-4o', 'url': 'https://openai.com/index/hello-gpt-4o/', 'organizer': 'OpenAI', 'license': 'Proprietary', 'MLE-Lite_Elo': 830, 'Tabular_Elo': 861, 'NLP_Elo': 903, 'CV_Elo': 761, 'Overall': 841},
+{'model_name': 'o3-mini', 'url': 'https://openai.com/index/openai-o3-mini/', 'organizer': 'OpenAI', 'license': 'Proprietary', 'MLE-Lite_Elo': 1108, 'Tabular_Elo': 1019, 'NLP_Elo': 1056, 'CV_Elo': 1207, 'Overall': 1096}, # Fill details later
+{'model_name': 'deepseek-v3', 'url': 'https://api-docs.deepseek.com/news/news1226', 'organizer': 'DeepSeek', 'license': 'DeepSeek', 'MLE-Lite_Elo': 1004, 'Tabular_Elo': 1015, 'NLP_Elo': 1028, 'CV_Elo': 1067, 'Overall': 1023},
+{'model_name': 'deepseek-r1', 'url': 'https://api-docs.deepseek.com/news/news250120', 'organizer': 'DeepSeek', 'license': 'DeepSeek', 'MLE-Lite_Elo': 1137, 'Tabular_Elo': 1053, 'NLP_Elo': 1103, 'CV_Elo': 1083, 'Overall': 1100},
+{'model_name': 'gemini-2.0-flash', 'url': 'https://ai.google.dev/gemini-api/docs/models#gemini-2.0-flash', 'organizer': 'Google', 'license': 'Proprietary', 'MLE-Lite_Elo': 847, 'Tabular_Elo': 923, 'NLP_Elo': 860, 'CV_Elo': 978, 'Overall': 895},
+{'model_name': 'gemini-2.0-pro', 'url': 'https://blog.google/technology/google-deepmind/gemini-model-updates-february-2025/', 'organizer': 'Google', 'license': 'Proprietary', 'MLE-Lite_Elo': 1064, 'Tabular_Elo': 1139, 'NLP_Elo': 1028, 'CV_Elo': 973, 'Overall': 1054},
+{'model_name': 'gemini-2.5-pro', 'url': 'https://deepmind.google/technologies/gemini/pro/', 'organizer': 'Google', 'license': 'Proprietary', 'MLE-Lite_Elo': 1257, 'Tabular_Elo': 1150, 'NLP_Elo': 1266, 'CV_Elo': 1177, 'Overall': 1214},
 ]
 
 # Create a master DataFrame
@@ -73,15 +75,17 @@ def update_leaderboard(category):
     if score_column is None or score_column not in master_df.columns:
         print(f"Warning: Invalid category '{category}' or column '{score_column}'. Falling back to default.")
         score_column = category_to_column[DEFAULT_CATEGORY]
-        if score_column not in master_df.columns: # Check fallback column too
-             # Return empty df with correct columns if still invalid
-             # Use lowercase keys here consistent with master_df for the empty case
+        # Check fallback column too
+        if score_column not in master_df.columns:
+            # Return empty df with correct columns if still invalid
+            # Use lowercase keys here consistent with master_df for the empty case
+            print(f"Error: Default column '{score_column}' also not found.")
             return pd.DataFrame({
                 "Rank": [],
                 "Model": [],
-                "organizer": [], # lowercase
-                "license": [],   # lowercase
-                "Elo Score": []
+                "Elo Score": [],
+                "Organizer": [], # Changed 'organizer' -> 'Organizer' for consistency in empty case
+                "License": []   # Changed 'license' -> 'License' for consistency in empty case
             })
 
     # Select base columns + the score column for sorting
@@ -106,14 +110,16 @@ def update_leaderboard(category):
     # Rename the score column to 'Elo Score' for consistent display
     df.rename(columns={score_column: 'Elo Score'}, inplace=True)
 
+    # Rename 'organizer' and 'license' to match desired display headers
+    df.rename(columns={'organizer': 'Organizer', 'license': 'License'}, inplace=True)
+
     # Select and reorder columns for final display using the ACTUAL column names in df
-    # Use lowercase 'organizer' and 'license' here because they haven't been renamed.
-    final_columns = ["Rank", "Model", "organizer", "license", "Elo Score"]
+    # Use capitalized 'Organizer' and 'License' here because they have been renamed.
+    final_columns = ["Rank", "Model", "Organizer", "License", "Elo Score"]
     df = df[final_columns]
 
     # Note: The DataFrame returned now has columns:
-    # 'Rank', 'Model', 'organizer', 'license', 'Elo Score'
-
+    # 'Rank', 'Model', 'Organizer', 'License', 'Elo Score'
     return df
 
 # --- Mock/Placeholder functions/data for other tabs ---
@@ -125,23 +131,50 @@ pending_eval_queue_df = pd.DataFrame(columns=["Model", "Status", "Requested", "S
 EVAL_COLS = ["Model", "Status", "Requested", "Started"] # Define for the dataframe headers
 EVAL_TYPES = ["str", "str", "str", "str"] # Define for the dataframe types
 
-
 # --- Keep restart function if relevant ---
 def restart_space():
     # Make sure REPO_ID is correctly defined/imported if this function is used
     print(f"Attempting to restart space: {REPO_ID}")
     # Replace with your actual space restart mechanism if needed (e.g., HfApi().restart_space(REPO_ID))
 
+
 # --- Gradio App Definition ---
-# Add custom CSS rules here or ensure custom_css is imported correctly
-# Example CSS rules you might want in your custom_css:
-# table { width: 100%; border-collapse: collapse; }
-# th, td { padding: 8px 12px; border: 1px solid #ddd; text-align: left; white-space: normal; } /* Allow wrapping */
-# th { background-color: #f2f2f2; font-weight: bold; }
-# tr:nth-child(even) { background-color: #f9f9f9; }
-# tr:hover { background-color: #e9e9e9; }
-# td a { color: #007bff; text-decoration: none; }
-# td a:hover { text-decoration: underline; }
+
+# ***** FONT SIZE INCREASED HERE *****
+# Add CSS rules to make the base font size larger.
+# Adjust the '1.2em' value (e.g., to '1.4em', '16px') to change the size.
+# The !important flag helps override theme defaults.
+# If the imported custom_css already has content, append to it.
+font_size_css = """
+body {
+    font-size: 1.2em !important; /* Increase base font size */
+}
+/* Optional: Target specific elements if needed */
+/*
+#leaderboard-table th, #leaderboard-table td {
+    font-size: 1em !important; /* Adjust table font size relative to new body size */
+    padding: 10px 14px !important; /* Increase padding for better spacing */
+}
+h1, .markdown-text h1 { font-size: 2.2em !important; } /* Make main title larger */
+h2, .markdown-text h2 { font-size: 1.8em !important; } /* Make section titles larger */
+button { font-size: 1.1em !important; padding: 8px 16px !important; } /* Slightly larger buttons */
+.gr-input, .gr-dropdown, .gr-textbox textarea { font-size: 1em !important; } /* Ensure inputs scale too */
+*/
+"""
+# Append the new CSS to any existing custom_css
+custom_css += font_size_css
+
+# Add basic table styling if not already present
+if "table {" not in custom_css:
+    custom_css += """
+table { width: 100%; border-collapse: collapse; margin-top: 10px; margin-bottom: 10px; }
+th, td { padding: 8px 12px; border: 1px solid #ddd; text-align: left; white-space: normal; vertical-align: top; } /* Allow wrapping, top align */
+th { background-color: #f2f2f2; font-weight: bold; }
+tr:nth-child(even) { background-color: #f9f9f9; }
+tr:hover { background-color: #e9e9e9; }
+td a { color: #007bff; text-decoration: none; }
+td a:hover { text-decoration: underline; }
+"""
 
 # Use a theme for better default styling
 demo = gr.Blocks(css=custom_css, theme=gr.themes.Soft())
@@ -149,6 +182,7 @@ demo = gr.Blocks(css=custom_css, theme=gr.themes.Soft())
 with demo:
     # Use the TITLE variable imported or defined above
     gr.HTML(TITLE)
+
     # Use the INTRODUCTION_TEXT variable imported or defined above
     gr.Markdown(INTRODUCTION_TEXT, elem_classes="markdown-text")
 
@@ -165,16 +199,16 @@ with demo:
                 leaderboard_df_component = gr.Dataframe(
                     # Initialize with sorted data for the default category
                     value=update_leaderboard(DEFAULT_CATEGORY),
-                    # Headers for DISPLAY remain capitalized
+                    # Headers for DISPLAY should match the *renamed* columns from update_leaderboard
                     headers=["Rank", "Model", "Organizer", "License", "Elo Score"],
-                    # Datatype maps to the final df columns: Rank, Model, organizer, license, Elo Score
+                    # Datatype maps to the final df columns: Rank, Model, Organizer, License, Elo Score
                     datatype=["number", "html", "str", "str", "number"],
                     interactive=False,
                     # --- FIX APPLIED: Removed unsupported 'height' argument ---
                     # row_count determines the number of rows to display
-                    row_count=(len(master_df), "fixed"),
+                    row_count=(len(master_df), "fixed"), # Display all rows
                     col_count=(5, "fixed"),
-                    wrap=True, # Allow text wrapping
+                    wrap=True, # Allow text wrapping in cells
                     elem_id="leaderboard-table" # CSS hook for custom styling
                 )
                 # Link the radio button change to the update function
@@ -185,7 +219,7 @@ with demo:
                 )
 
         with gr.TabItem("📝 About", elem_id="llm-benchmark-tab-about", id=1):
-             # Use the LLM_BENCHMARKS_TEXT variable imported or defined above
+            # Use the LLM_BENCHMARKS_TEXT variable imported or defined above
             gr.Markdown(LLM_BENCHMARKS_TEXT, elem_classes="markdown-text")
 
         # --- Submit Tab (Commented out as in original request) ---
@@ -227,26 +261,45 @@ with demo:
         #          submission_result,
         #      )
 
-
     # --- Citation Row (at the bottom, outside Tabs) ---
     with gr.Accordion("📙 Citation", open=False):
-         # Use the CITATION_BUTTON_TEXT and CITATION_BUTTON_LABEL variables imported or defined above
-         citation_button = gr.Textbox(
-             value=CITATION_BUTTON_TEXT,
-             label=CITATION_BUTTON_LABEL,
-             lines=10,
-             elem_id="citation-button",
-             show_copy_button=True,
-         )
+        # Use the CITATION_BUTTON_TEXT and CITATION_BUTTON_LABEL variables imported or defined above
+        citation_button = gr.Textbox(
+            value=CITATION_BUTTON_TEXT,
+            label=CITATION_BUTTON_LABEL,
+            lines=10, # Adjust lines if needed for new font size
+            elem_id="citation-button",
+            show_copy_button=True,
+        )
+
+# IGNORE_WHEN_COPYING_START
+# content_copy  download
+# Use code with caution.
+# IGNORE_WHEN_COPYING_END
 
 # --- Keep scheduler if relevant ---
-# scheduler = BackgroundScheduler()
-# scheduler.add_job(restart_space, "interval", seconds=1800) # Restart every 30 mins
-# scheduler.start()
+# Only start scheduler if the script is run directly
+if __name__ == "__main__":
+    try:
+        scheduler = BackgroundScheduler()
+        # Add job only if restart_space is callable (i.e., not a placeholder or failed import)
+        if callable(restart_space):
+             # Check if REPO_ID seems valid before scheduling
+             if REPO_ID and REPO_ID != "your/space-id":
+                 scheduler.add_job(restart_space, "interval", seconds=1800) # Restart every 30 mins
+                 scheduler.start()
+             else:
+                 print("Warning: REPO_ID not set or is placeholder; space restart job not scheduled.")
+        else:
+             print("Warning: restart_space function not available; space restart job not scheduled.")
+    except Exception as e:
+        print(f"Failed to initialize or start scheduler: {e}")
+
 
 # --- Launch the app ---
 # Ensures the app launches only when the script is run directly
 if __name__ == "__main__":
     # Ensure you have installed necessary libraries: pip install gradio pandas apscheduler
     # Make sure your src module files (about.py etc.) are accessible OR use the placeholder definitions above.
+    print("Launching Gradio App...")
     demo.launch()