Update space

README.md

@@ -11,7 +11,6 @@ license: mit
 short_description: 'TRAIL: Trace Reasoning and Agentic Issue Localization'
 sdk_version: 5.19.0
 ---
-
 # Model Performance Leaderboard
 
 This is a Hugging Face Space that hosts a leaderboard for comparing model performances across various metrics.
@@ -23,6 +22,24 @@ This is a Hugging Face Space that hosts a leaderboard for comparing model perfor
 - **Integrated Backend**: Stores all submissions with timestamp and attribution
 - **Customizable Metrics**: Configure which metrics to display and track
 
+## Installation
+
+### Setting Up Your Space
+
+1. Upload all files to your Hugging Face Space
+2. Make sure to make `start.sh` executable:
+   ```bash
+   chmod +x start.sh
+   ```
+3. Configure your Space to use the `start.sh` script as the entry point
+
+### Troubleshooting Installation Issues
+
+If you encounter JSON parsing errors:
+1. Check if `models.json` exists and is a valid JSON file
+2. Run `python setup.py` to regenerate configuration files
+3. If problems persist, delete the `models.json` file and let the setup script create a new one
+
 ## How to Use
 
 ### Viewing the Leaderboard
@@ -66,4 +83,4 @@ This leaderboard is built using:
 
 ## License
 
-This project is open source and available under the MIT license.
+This project is open source and available under the MIT license.

app.py

@@ -3,17 +3,29 @@ import json
 import pandas as pd
 import datetime
 import os
-from database import Database, load_config
+import sys
+from pathlib import Path
 
-# Initialize database
-db = Database()
-config = load_config()
-
-# Set up the title and description from config
-title = config.get("title", "Model Leaderboard")
-description = config.get("description", "Submit and compare model performances")
-metrics = config.get("metrics", ["accuracy"])
-main_metric = config.get("main_metric", metrics[0])
+# Add better error handling for initial setup
+try:
+    from database import Database, load_config
+    
+    # Initialize database
+    db = Database()
+    config = load_config()
+    
+    # Set up the title and description from config
+    title = config.get("title", "Model Leaderboard")
+    description = config.get("description", "Submit and compare model performances")
+    metrics = config.get("metrics", ["accuracy"])
+    main_metric = config.get("main_metric", metrics[0] if metrics else "accuracy")
+except Exception as e:
+    print(f"Error during initialization: {e}")
+    # Set fallback values in case of error
+    title = "Model Leaderboard"
+    description = "Submit and compare model performances"
+    metrics = ["accuracy"]
+    main_metric = "accuracy"
 
 def format_leaderboard_data(submissions):
     """Format submissions data for the leaderboard display"""

database.py

@@ -1,80 +1,129 @@
-import os
+import gradio as gr
 import json
+import pandas as pd
 import datetime
-from pathlib import Path
-import numpy as np
+import os
+from database import Database, load_config
 
-class Database:
-    def __init__(self, submission_dir="submissions"):
-        self.submission_dir = submission_dir
-        os.makedirs(submission_dir, exist_ok=True)
-        
-    def add_submission(self, submission):
-        """Add a new submission to the database"""
-        # Generate a timestamp and ID for the submission
-        timestamp = datetime.datetime.now().isoformat()
-        submission_id = f"{submission['model_name'].replace(' ', '_')}_{timestamp.replace(':', '-')}"
+# Initialize database
+db = Database()
+config = load_config()
+
+# Set up the title and description from config
+title = config.get("title", "Model Leaderboard")
+description = config.get("description", "Submit and compare model performances")
+metrics = config.get("metrics", ["accuracy"])
+main_metric = config.get("main_metric", metrics[0])
+
+def format_leaderboard_data(submissions):
+    """Format submissions data for the leaderboard display"""
+    if not submissions:
+        return pd.DataFrame()
+    
+    # Extract relevant fields for display
+    data = []
+    for sub in submissions:
+        entry = {
+            "Model": sub["model_name"],
+            "Submitted by": sub["submitter_name"],
+            "Date": sub["timestamp"].split("T")[0],
+        }
         
-        # Add timestamp and ID to submission
-        submission['timestamp'] = timestamp
-        submission['id'] = submission_id
+        # Add all metrics from the submission
+        for metric in metrics:
+            if metric in sub:
+                entry[metric.replace("_", " ").title()] = f"{sub[metric]:.4f}"
         
-        # Save submission to a JSON file
-        file_path = os.path.join(self.submission_dir, f"{submission_id}.json")
-        with open(file_path, 'w') as f:
-            json.dump(submission, f, indent=2)
-            
-        return submission_id
+        data.append(entry)
     
-    def get_submission(self, submission_id):
-        """Get a specific submission by ID"""
-        file_path = os.path.join(self.submission_dir, f"{submission_id}.json")
-        if os.path.exists(file_path):
-            with open(file_path, 'r') as f:
-                return json.load(f)
-        return None
+    return pd.DataFrame(data)
+
+def add_submission(model_name, submitter_name, description, **metric_values):
+    """Add a new submission to the leaderboard"""
+    if not model_name or not submitter_name:
+        return "Model name and submitter name are required.", None
     
-    def get_all_submissions(self):
-        """Get all submissions"""
-        submissions = []
-        for file_name in os.listdir(self.submission_dir):
-            if file_name.endswith('.json'):
-                file_path = os.path.join(self.submission_dir, file_name)
-                with open(file_path, 'r') as f:
-                    submissions.append(json.load(f))
-        return submissions
+    # Create submission dictionary
+    submission = {
+        "model_name": model_name,
+        "submitter_name": submitter_name,
+        "description": description,
+    }
     
-    def get_leaderboard(self, sort_by="score", ascending=False):
-        """Get submissions sorted for leaderboard display"""
-        submissions = self.get_all_submissions()
-        
-        # Sort submissions
-        if sort_by in submissions[0] if submissions else False:
-            submissions.sort(key=lambda x: x.get(sort_by, 0), reverse=not ascending)
-        
-        return submissions
+    # Add metrics
+    for metric in metrics:
+        metric_key = f"{metric}_input"
+        if metric_key in metric_values and metric_values[metric_key]:
+            try:
+                submission[metric] = float(metric_values[metric_key])
+            except ValueError:
+                return f"Invalid value for {metric}. Please enter a number.", None
+    
+    # Add submission to database
+    submission_id = db.add_submission(submission)
+    
+    # Update leaderboard
+    submissions = db.get_leaderboard(sort_by=main_metric, ascending=False)
+    leaderboard_df = format_leaderboard_data(submissions)
     
-    def delete_submission(self, submission_id):
-        """Delete a submission by ID"""
-        file_path = os.path.join(self.submission_dir, f"{submission_id}.json")
-        if os.path.exists(file_path):
-            os.remove(file_path)
-            return True
-        return False
+    return f"Submission added successfully! ID: {submission_id}", leaderboard_df
 
-# Load leaderboard configuration
-def load_config():
-    if os.path.exists("models.json"):
-        with open("models.json", "r") as f:
-            return json.load(f)
-    else:
-        # Default configuration
-        config = {
-            "title": "Model Leaderboard",
-            "description": "Submit and compare model performances",
-            "metrics": ["accuracy", "f1_score", "precision", "recall"],
-            "main_metric": "accuracy"
-        }
-        with open("models.json", "w") as f:
-            json.dump(config, f, indent=2)
-        return config
+def update_leaderboard(sort_metric=main_metric, ascending=False):
+    """Get the current leaderboard data"""
+    submissions = db.get_leaderboard(sort_by=sort_metric, ascending=ascending)
+    return format_leaderboard_data(submissions)
+
+# Create the Gradio interface
+with gr.Blocks() as app:
+    gr.Markdown(f"# {title}")
+    gr.Markdown(description)
+    
+    with gr.Tabs():
+        with gr.TabItem("Leaderboard"):
+            sort_metric = gr.Dropdown(
+                choices=[m.replace("_", " ").title() for m in metrics],
+                value=main_metric.replace("_", " ").title(),
+                label="Sort by"
+            )
+            sort_order = gr.Checkbox(label="Ascending order")
+            leaderboard = gr.DataFrame(update_leaderboard())
+            refresh_btn = gr.Button("Refresh Leaderboard")
+            
+            def handle_sort(metric, ascending):
+                metric_key = metric.lower().replace(" ", "_")
+                return update_leaderboard(sort_metric=metric_key, ascending=ascending)
+            
+            sort_metric.change(handle_sort, [sort_metric, sort_order], leaderboard)
+            sort_order.change(handle_sort, [sort_metric, sort_order], leaderboard)
+            refresh_btn.click(handle_sort, [sort_metric, sort_order], leaderboard)
+        
+        with gr.TabItem("Submit Model"):
+            with gr.Column():
+                model_name = gr.Textbox(label="Model Name")
+                submitter = gr.Textbox(label="Your Name")
+                model_desc = gr.Textbox(label="Model Description (optional)", lines=3)
+                
+                # Create metric input fields
+                metric_inputs = {}
+                with gr.Column():
+                    for metric in metrics:
+                        metric_inputs[f"{metric}_input"] = gr.Number(
+                            label=f"{metric.replace('_', ' ').title()}", 
+                            min=0, 
+                            max=1 if "accuracy" in metric or "score" in metric else None
+                        )
+                
+                submit_btn = gr.Button("Submit Model")
+                result = gr.Textbox(label="Result")
+                
+                # Connect submission function
+                inputs = [model_name, submitter, model_desc] + list(metric_inputs.values())
+                submit_btn.click(
+                    add_submission, 
+                    inputs=[model_name, submitter, model_desc] + list(metric_inputs.values()),
+                    outputs=[result, leaderboard],
+                    kwargs=metric_inputs
+                )
+
+if __name__ == "__main__":
+    app.launch(debug=True)

model

@@ -0,0 +1,6 @@
+{
+  "title": "Model Performance Leaderboard",
+  "description": "This leaderboard tracks and compares model performance across multiple metrics. Submit your model results to see how they stack up!",
+  "metrics": ["accuracy", "f1_score", "precision", "recall"],
+  "main_metric": "accuracy"
+}

models.json

@@ -0,0 +1,6 @@
+{
+    "title": "Model Performance Leaderboard",
+    "description": "This leaderboard tracks and compares model performance across multiple metrics. Submit your model results to see how they stack up!",
+    "metrics": ["accuracy", "f1_score", "precision", "recall"],
+    "main_metric": "accuracy"
+  }