21

app.py

@@ -18,7 +18,7 @@ class ResearchOrchestrator:
         self.analyzer = analyzer
         self.citation_manager = citation_manager
         self.formatter = formatter
-    
+
     def run(self, query):
         """Execute the research pipeline with streaming updates"""
         try:
@@ -41,7 +41,7 @@ class ResearchOrchestrator:
             logging.info(f"Retrieved {len(search_results)} results")
             
             # Step 3: Analyze content
-            yield "🧠 Analyzing search results..."
+            yield "🧠 Analyzing search results...\n\n⏳ The AI model may be initializing. This could take a few minutes if it's the first request..."
             analysis = self.analyzer.analyze(query, search_results)
             logging.info("Analysis completed")
             
@@ -81,16 +81,16 @@ def initialize_modules():
     try:
         if not CONFIG["tavily_api_key"]:
             raise ValueError("TAVILY_API_KEY environment variable is not set")
-        
+
         if not CONFIG["hf_api_key"]:
             raise ValueError("HF_TOKEN environment variable is not set")
-            
+
         input_handler = InputHandler()
         retriever = Retriever(api_key=CONFIG["tavily_api_key"])
         analyzer = Analyzer(base_url=CONFIG["hf_api_base"], api_key=CONFIG["hf_api_key"])
         citation_manager = CitationManager()
         formatter = OutputFormatter()
-        
+
         return ResearchOrchestrator(
             input_handler, 
             retriever, 
@@ -124,7 +124,7 @@ with gr.Blocks(title="Research Assistant") as demo:
                 lines=3
             )
             submit_btn = gr.Button("Research", variant="primary")
-        
+            
         with gr.Column():
             output = gr.Markdown(label="Analysis Results")
     
@@ -150,5 +150,4 @@ with gr.Blocks(title="Research Assistant") as demo:
     )
 
 if __name__ == "__main__":
-    demo.launch()
-
+    demo.launch()

modules/analyzer.py

@@ -1,5 +1,8 @@
+# modules/analyzer.py
 from openai import OpenAI
-import json
+import requests
+import time
+import logging
 
 class Analyzer:
     def __init__(self, base_url, api_key):
@@ -7,7 +10,39 @@ class Analyzer:
             base_url=base_url,
             api_key=api_key
         )
+        self.health_check_url = base_url.rstrip('/') + "/health"
+        self.headers = {"Authorization": f"Bearer {api_key}"}
         
+    def wait_for_server(self, timeout=300, interval=10):
+        """
+        Waits for the server to become available by polling the health endpoint.
+        
+        Parameters:
+            timeout (int): Max time in seconds to wait
+            interval (int): Time between checks
+            
+        Returns:
+            bool: True if server is ready, False if timeout reached
+        """
+        logging.info("⏳ Waiting for the server to initialize...")
+        start_time = time.time()
+        
+        while time.time() - start_time < timeout:
+            try:
+                response = requests.get(self.health_check_url, headers=self.headers, timeout=10)
+                if response.status_code == 200:
+                    logging.info("✅ Server is ready!")
+                    return True
+                else:
+                    logging.info(f"🌐 Server responded with status: {response.status_code} — still initializing...")
+            except requests.exceptions.RequestException as e:
+                logging.info("🔴 Still unreachable — retrying...")
+            
+            time.sleep(interval)
+        
+        logging.warning("⏰ Timeout reached. Server didn't initialize in time.")
+        return False
+
     def analyze(self, query, search_results):
         """
         Analyze search results using the custom LLM
@@ -37,6 +72,10 @@ class Analyzer:
         """
         
         try:
+            # First check if server is ready
+            if not self.wait_for_server(timeout=180):  # 3 minutes timeout
+                return "⚠️ The AI model server is still initializing. Please try again in a few minutes."
+            
             response = self.client.chat.completions.create(
                 model="DavidAU/OpenAi-GPT-oss-20b-abliterated-uncensored-NEO-Imatrix-gguf",
                 messages=[
@@ -51,5 +90,7 @@ class Analyzer:
             return response.choices[0].message.content
             
         except Exception as e:
-            return f"Analysis failed: {str(e)}"
-            
+            error_msg = str(e)
+            if "503" in error_msg or "Service Unavailable" in error_msg:
+                return "⚠️ The AI model server is currently unavailable. It may be initializing. Please try again in a few minutes."
+            return f"Analysis failed: {str(e)}"