Update llm_client.py

llm_client.py

@@ -77,9 +77,9 @@ class HybridLLM(LLM):
             try:
                 print(f"🌐 Calling Colab API: {self.api_url}")
                 response = requests.post(
-                    f"{self.api_url}/analyze",
+                    f"{self.api_url}/generate",
                     json={"prompt": prompt, "max_tokens": 512},
-                    timeout=300
+                    timeout=30
                 )
                 if response.status_code == 200:
                     return response.json()["response"]
@@ -125,13 +125,33 @@ class LLMClient:
         self.api_url = os.environ.get("COLAB_API_URL", "")
         self.server_process = None
         self.server_port = 8080
+
+        # 1. Setup Groq (Highest Priority)
+        self.groq_api_key = os.environ.get("GROQ_API_KEY")
+        self.groq_model = "qwen-2.5-32b" # Default, user can override via env if needed
+
+        if self.groq_api_key:
+            try:
+                from langchain_groq import ChatGroq
+                print(f"⚡ Initializing Groq API with model: {self.groq_model}...")
+                self.llm = ChatGroq(
+                    temperature=0.3,
+                    model=self.groq_model,
+                    groq_api_key=self.groq_api_key,
+                    max_tokens=1024,
+                    model_kwargs={"stop": ["<|im_end|>", "Input:", "Context:"]}
+                )
+                print("✅ Groq API ready! Skipping local setup.")
+                return # Exit early, no need for local server
+            except Exception as e:
+                print(f"⚠️ Groq Init Failed: {e}")
         
-        # Setup Local Fallback
+        # 2. Setup Local Fallback (Only if Groq fails/missing)
         try:
-            # 1. Setup Binary
+            # Setup Binary
             self.server_bin, self.lib_path = setup_llama_binaries()
             
-            # 2. Download Model (Qwen3-0.6B)
+            # Download Model (Qwen3-0.6B)
             print("📂 Loading Local Qwen3-0.6B (GGUF)...")
             model_repo = "Qwen/Qwen3-0.6B-GGUF" 
             filename = "Qwen3-0.6B-Q8_0.gguf"
@@ -142,13 +162,13 @@ class LLMClient:
             )
             print(f"✅ Model downloaded to: {self.model_path}")
             
-            # 3. Start Server
+            # Start Server
             self.start_local_server()
             
         except Exception as e:
             print(f"⚠️ Could not setup local fallback: {e}")
 
-        # Create Hybrid LangChain Wrapper
+        # Create Hybrid LangChain Wrapper (Fallback)
         self.llm = HybridLLM(
             api_url=self.api_url,
             local_server_url=f"http://localhost:{self.server_port}"