Fixing generate

app.py

@@ -1,37 +1,77 @@
 import gradio as gr
 import time
 import json
+import traceback
 
 # Import RAG system components
-from rag_system.vector_store import VectorStore
-from rag_system.retriever import SQLRetriever
-from rag_system.prompt_engine import PromptEngine
-from rag_system.sql_generator import SQLGenerator
+print("Starting RAG system initialization...")
+try:
+    from rag_system.vector_store import VectorStore
+    print("✓ VectorStore imported successfully")
+except Exception as e:
+    print(f"✗ VectorStore import failed: {e}")
+    traceback.print_exc()
+
+try:
+    from rag_system.retriever import SQLRetriever
+    print("✓ SQLRetriever imported successfully")
+except Exception as e:
+    print(f"✗ SQLRetriever import failed: {e}")
+    traceback.print_exc()
+
+try:
+    from rag_system.prompt_engine import PromptEngine
+    print("✓ PromptEngine imported successfully")
+except Exception as e:
+    print(f"✗ PromptEngine import failed: {e}")
+    traceback.print_exc()
+
+try:
+    from rag_system.sql_generator import SQLGenerator
+    print("✓ SQLGenerator imported successfully")
+except Exception as e:
+    print(f"✗ SQLGenerator import failed: {e}")
+    traceback.print_exc()
 
 # Initialize RAG system components
-print("Initializing RAG system...")
+print("Initializing RAG system components...")
+sql_generator = None
 try:
     vector_store = VectorStore()
+    print("✓ VectorStore initialized")
+    
     retriever = SQLRetriever(vector_store)
+    print("✓ SQLRetriever initialized")
+    
     prompt_engine = PromptEngine()
+    print("✓ PromptEngine initialized")
+    
     sql_generator = SQLGenerator(retriever, prompt_engine)
-    print("RAG system initialized successfully!")
+    print("✓ SQLGenerator initialized")
+    
+    print("🎉 RAG system initialized successfully!")
 except Exception as e:
-    print(f"Error initializing RAG system: {e}")
+    print(f"❌ Error initializing RAG system: {e}")
+    traceback.print_exc()
     sql_generator = None
 
 def generate_sql(question, table_headers):
     """Generate SQL using the RAG system directly."""
     if sql_generator is None:
-        return "❌ Error: RAG system not initialized"
+        return "❌ Error: RAG system not initialized. Check the logs for initialization errors."
     
     try:
+        print(f"Generating SQL for: {question}")
+        print(f"Table headers: {table_headers}")
+        
         start_time = time.time()
         
         # Generate SQL using RAG system
         result = sql_generator.generate_sql(question, table_headers)
         
         processing_time = time.time() - start_time
+        print(f"SQL generation completed in {processing_time:.2f}s")
+        print(f"Result: {result}")
         
         return f"""
 **Generated SQL:**
@@ -45,12 +85,14 @@ def generate_sql(question, table_headers):
 **Retrieved Examples:** {len(result['retrieved_examples'])} examples used for RAG
 """
     except Exception as e:
-        return f"❌ Error: {str(e)}"
+        error_msg = f"❌ Error: {str(e)}\n\nFull traceback:\n{traceback.format_exc()}"
+        print(error_msg)
+        return error_msg
 
 def batch_generate_sql(questions_text, table_headers):
     """Generate SQL for multiple questions."""
     if sql_generator is None:
-        return "❌ Error: RAG system not initialized"
+        return "❌ Error: RAG system not initialized. Check the logs for initialization errors."
     
     try:
         # Parse questions
@@ -81,16 +123,21 @@ def batch_generate_sql(questions_text, table_headers):
         return output
         
     except Exception as e:
-        return f"❌ Error: {str(e)}"
+        return f"❌ Error: {str(e)}\n\nFull traceback:\n{traceback.format_exc()}"
 
 def check_system_health():
     """Check the health of the RAG system."""
     try:
         if sql_generator is None:
-            return "❌ System Status: RAG system not initialized"
+            return "❌ System Status: RAG system not initialized\n\nCheck the logs above for initialization errors."
         
         # Get model info
-        model_info = sql_generator.get_model_info()
+        try:
+            model_info = sql_generator.get_model_info()
+            model_status = "Available"
+        except Exception as e:
+            model_info = {"error": str(e)}
+            model_status = f"Error: {e}"
         
         return f"""
 **System Health:**
@@ -98,20 +145,30 @@ def check_system_health():
 - **System Loaded:** ✅ Yes
 - **System Loading:** ❌ No
 - **Error:** None
+- **Model Status:** {model_status}
 - **Timestamp:** {time.strftime('%Y-%m-%d %H:%M:%S')}
 
 **Model Info:**
 {json.dumps(model_info, indent=2) if model_info else 'Not available'}
+
+**Initialization Logs:**
+Check the console/logs above for detailed initialization information.
 """
     except Exception as e:
-        return f"❌ Health check error: {str(e)}"
+        return f"❌ Health check error: {str(e)}\n\nFull traceback:\n{traceback.format_exc()}"
 
 # Create Gradio interface
 with gr.Blocks(title="Text-to-SQL RAG with CodeLlama", theme=gr.themes.Soft()) as demo:
-    gr.Markdown("#Text-to-SQL RAG with CodeLlama")
+    gr.Markdown("# 🚀 Text-to-SQL RAG with CodeLlama")
     gr.Markdown("Generate SQL queries from natural language using **RAG (Retrieval-Augmented Generation)** and **CodeLlama** models.")
     gr.Markdown("**Features:** RAG-enhanced generation, CodeLlama integration, Vector-based retrieval, Advanced prompt engineering")
     
+    # Add initialization status
+    if sql_generator is None:
+        gr.Markdown("⚠️ **Warning:** RAG system failed to initialize. Check the logs for errors.")
+    else:
+        gr.Markdown("✅ **Status:** RAG system initialized successfully!")
+    
     with gr.Tab("Single Query"):
         with gr.Row():
             with gr.Column(scale=1):
@@ -125,7 +182,7 @@ with gr.Blocks(title="Text-to-SQL RAG with CodeLlama", theme=gr.themes.Soft()) a
                     placeholder="e.g., id, name, salary, department",
                     value="id, name, salary, department"
                 )
-                generate_btn = gr.Button("Generate SQL", variant="primary", size="lg")
+                generate_btn = gr.Button("🚀 Generate SQL", variant="primary", size="lg")
             
             with gr.Column(scale=1):
                 output = gr.Markdown(label="Result")
@@ -143,14 +200,14 @@ with gr.Blocks(title="Text-to-SQL RAG with CodeLlama", theme=gr.themes.Soft()) a
                     placeholder="e.g., id, name, salary, department",
                     value="id, name, salary, department"
                 )
-                batch_btn = gr.Button("Generate Batch SQL", variant="primary", size="lg")
+                batch_btn = gr.Button("🚀 Generate Batch SQL", variant="primary", size="lg")
             
             with gr.Column(scale=1):
                 batch_output = gr.Markdown(label="Batch Results")
     
     with gr.Tab("System Health"):
         with gr.Row():
-            health_btn = gr.Button("Check System Health", variant="secondary", size="lg")
+            health_btn = gr.Button("🔍 Check System Health", variant="secondary", size="lg")
             health_output = gr.Markdown(label="Health Status")
     
     # Event handlers
@@ -173,14 +230,14 @@ with gr.Blocks(title="Text-to-SQL RAG with CodeLlama", theme=gr.themes.Soft()) a
     
     gr.Markdown("---")
     gr.Markdown("""
-    ## How It Works
+    ## 🎯 How It Works
     
     1. **RAG System**: Retrieves relevant SQL examples from vector database
     2. **CodeLlama**: Generates SQL using retrieved examples as context
     3. **Vector Search**: Finds similar questions and their SQL solutions
     4. **Enhanced Generation**: Combines retrieval + generation for better accuracy
     
-    ## Technology Stack
+    ## 🛠️ Technology Stack
     
     - **Backend**: Direct RAG system integration
     - **LLM**: CodeLlama-7B-Python-GGUF (primary)
@@ -188,7 +245,7 @@ with gr.Blocks(title="Text-to-SQL RAG with CodeLlama", theme=gr.themes.Soft()) a
     - **Frontend**: Gradio interface
     - **Hosting**: Hugging Face Spaces
     
-    ## Performance
+    ## 📊 Performance
     
     - **Model**: CodeLlama-7B-Python-GGUF
     - **Response Time**: < 5 seconds